This is primarily useful for the upcoming pg_xlogdump --stats feature,
but also allows to remove some duplicated code in the rmgr_desc
routines.
Due to the separation and harmonization, the output of dipsplayed
records changes somewhat. But since this isn't enduser oriented
content that's ok.
It's potentially desirable to further change pg_xlogdump's display of
records. It previously wasn't possible to show the record type
separately from the description forcing it to be in the last
column. But that's better done in a separate commit.
Author: Abhijit Menon-Sen, slightly editorialized by me
Reviewed-By: Álvaro Herrera, Andres Freund, and Heikki Linnakangas
Discussion: 20140604104716.GA3989@toroid.org
Every redo routine uses the same idiom to determine what to do to a page:
check if there's a backup block for it, and if not read, the buffer if the
block exists, and check its LSN. Refactor that into a common function,
XLogReadBufferForRedo, making all the redo routines shorter and more
readable.
This has no user-visible effect, and makes no changes to the WAL format.
Reviewed by Andres Freund, Alvaro Herrera, Michael Paquier.
This reverts commit e23014f3d4.
As the side effect of the reverted commit, when the unit is
specified, the reloption was stored in the catalog with the unit.
This broke pg_dump (specifically, it prevented pg_dump from
outputting restorable backup regarding the reloption) and
turned the buildfarm red. Revert the commit until the fixed
version is ready.
This introduces an infrastructure which allows us to specify the units
like ms (milliseconds) in integer relation option, like GUC parameter.
Currently only autovacuum_vacuum_cost_delay reloption can accept
the units.
Reviewed by Michael Paquier
log_newpage is used by many indexams, in addition to heap, but for
historical reasons it's always been part of the heapam rmgr. Starting with
9.3, we have another WAL record type for logging an image of a page,
XLOG_FPI. Simplify things by moving log_newpage and log_newpage_buffer to
xlog.c, and switch to using the XLOG_FPI record type.
Bump the WAL version number because the code to replay the old HEAP_NEWPAGE
records is removed.
Commit 0ac5ad5134 removed an optimization in multixact.c that skipped
fetching members of MultiXactId that were older than our
OldestVisibleMXactId value. The reason this was removed is that it is
possible for multixacts that contain updates to be older than that
value. However, if the caller is certain that the multi does not
contain an update (because the infomask bits say so), it can pass this
info down to GetMultiXactIdMembers, enabling it to use the old
optimization.
Pointed out by Andres Freund in 20131121200517.GM7240@alap2.anarazel.de
There were several oversights in recovery code where COMMIT/ABORT PREPARED
records were ignored:
* pg_last_xact_replay_timestamp() (wasn't updated for 2PC commits)
* recovery_min_apply_delay (2PC commits were applied immediately)
* recovery_target_xid (recovery would not stop if the XID used 2PC)
The first of those was reported by Sergiy Zuban in bug #11032, analyzed by
Tom Lane and Andres Freund. The bug was always there, but was masked before
commit d19bd29f07, because COMMIT PREPARED
always created an extra regular transaction that was WAL-logged.
Backpatch to all supported versions (older versions didn't have all the
features and therefore didn't have all of the above bugs).
Commit 1b86c81d2d fixed the decoding of toasted columns for the rows
contained in one xl_heap_multi_insert record. But that's not actually
enough, because heap_multi_insert() will actually first toast all
passed in rows and then emit several *_multi_insert records; one for
each page it fills with tuples.
Add a XLOG_HEAP_LAST_MULTI_INSERT flag which is set in
xl_heap_multi_insert->flag denoting that this multi_insert record is
the last emitted by one heap_multi_insert() call. Then use that flag
in decode.c to only set clear_toast_afterwards in the right situation.
Expand the number of rows inserted via COPY in the corresponding
regression test to make sure that more than one heap page is filled
with tuples by one heap_multi_insert() call.
Backpatch to 9.4 like the previous commit.
The previous code, perhaps out of concern for avoid memory leaks, formed
the tuple in one memory context and then copied it to another memory
context. However, this doesn't appear to be necessary, since
index_form_tuple and the functions it calls take precautions against
leaking memory. In my testing, building the tuple directly inside the
sort context shaves several percent off the index build time.
Rearrange things so we do that.
Patch by me. Review by Amit Kapila, Tom Lane, Andres Freund.
Instead of truncating pg_multixact at vacuum time, do it only at
checkpoint time. The reason for doing it this way is twofold: first, we
want it to delete only segments that we're certain will not be required
if there's a crash immediately after the removal; and second, we want to
do it relatively often so that older files are not left behind if
there's an untimely crash.
Per my proposal in
http://www.postgresql.org/message-id/20140626044519.GJ7340@eldon.alvh.no-ip.org
we now execute the truncation in the checkpointer process rather than as
part of vacuum. Vacuum is in only charge of maintaining in shared
memory the value to which it's possible to truncate the files; that
value is stored as part of checkpoints also, and so upon recovery we can
reuse the same value to re-execute truncate and reset the
oldest-value-still-safe-to-use to one known to remain after truncation.
Per bug reported by Jeff Janes in the course of his tests involving
bug #8673.
While at it, update some comments that hadn't been updated since
multixacts were changed.
Backpatch to 9.3, where persistency of pg_multixact files was
introduced by commit 0ac5ad5134.
In a50d976254 I already changed this, but got it wrong for the case
where the number of members is larger than the number of entries that
fit in the last page of the last segment.
As reported by Serge Negodyuck in a followup to bug #8673.
The way the code was written, the padding was copied from uninitialized
memory areas.. Because the structs are local variables in the code where
the WAL records are constructed, making them larger and zeroing the padding
bytes would not make the code very pretty, so rather than fixing this
directly by zeroing out the padding bytes, it seems more clear to not try to
align the tuples in the WAL records. The redo functions are taught to copy
the tuple header to a local variable to avoid unaligned access.
Stable-branches have the same problem, but we can't change the WAL format
there, so fix in master only. Reading a few random extra bytes at the stack
is harmless in practice, so it's not worth crafting a different
back-patchable fix.
Per reports from Kevin Grittner and Andres Freund, using clang static
analyzer and Valgrind, respectively.
To lock a prepared transaction's shared memory entry, we used to mark it
with the XID of the backend. When the XID was no longer active according
to the proc array, the entry was implicitly considered as not locked
anymore. However, when preparing a transaction, the backend's proc array
entry was cleared before transfering the locks (and some other state) to
the prepared transaction's dummy PGPROC entry, so there was a window where
another backend could finish the transaction before it was in fact fully
prepared.
To fix, rewrite the locking mechanism of global transaction entries. Instead
of an XID, just have simple locked-or-not flag in each entry (we store the
locking backend's backend id rather than a simple boolean, but that's just
for debugging purposes). The backend is responsible for explicitly unlocking
the entry, and to make sure that that happens, install a callback to unlock
it on abort or process exit.
Backpatch to all supported versions.
If we have an array of records stored on disk, the individual record fields
cannot contain out-of-line TOAST pointers: the tuptoaster.c mechanisms are
only prepared to deal with TOAST pointers appearing in top-level fields of
a stored row. The same applies for ranges over composite types, nested
composites, etc. However, the existing code only took care of expanding
sub-field TOAST pointers for the case of nested composites, not for other
structured types containing composites. For example, given a command such
as
UPDATE tab SET arraycol = ARRAY[(ROW(x,42)::mycompositetype] ...
where x is a direct reference to a field of an on-disk tuple, if that field
is long enough to be toasted out-of-line then the TOAST pointer would be
inserted as-is into the array column. If the source record for x is later
deleted, the array field value would become a dangling pointer, leading
to errors along the line of "missing chunk number 0 for toast value ..."
when the value is referenced. A reproducible test case for this was
provided by Jan Pecek, but it seems likely that some of the "missing chunk
number" reports we've heard in the past were caused by similar issues.
Code-wise, the problem is that PG_DETOAST_DATUM() is not adequate to
produce a self-contained Datum value if the Datum is of composite type.
Seen in this light, the problem is not just confined to arrays and ranges,
but could also affect some other places where detoasting is done in that
way, for example form_index_tuple().
I tried teaching the array code to apply toast_flatten_tuple_attribute()
along with PG_DETOAST_DATUM() when the array element type is composite,
but this was messy and imposed extra cache lookup costs whether or not any
TOAST pointers were present, indeed sometimes when the array element type
isn't even composite (since sometimes it takes a typcache lookup to find
that out). The idea of extending that approach to all the places that
currently use PG_DETOAST_DATUM() wasn't attractive at all.
This patch instead solves the problem by decreeing that composite Datum
values must not contain any out-of-line TOAST pointers in the first place;
that is, we expand out-of-line fields at the point of constructing a
composite Datum, not at the point where we're about to insert it into a
larger tuple. This rule is applied only to true composite Datums, not
to tuples that are being passed around the system as tuples, so it's not
as invasive as it might sound at first. With this approach, the amount
of code that has to be touched for a full solution is greatly reduced,
and added cache lookup costs are avoided except when there actually is
a TOAST pointer that needs to be inlined.
The main drawback of this approach is that we might sometimes dereference
a TOAST pointer that will never actually be used by the query, imposing a
rather large cost that wasn't there before. On the other side of the coin,
if the field value is used multiple times then we'll come out ahead by
avoiding repeat detoastings. Experimentation suggests that common SQL
coding patterns are unaffected either way, though. Applications that are
very negatively affected could be advised to modify their code to not fetch
columns they won't be using.
In future, we might consider reverting this solution in favor of detoasting
only at the point where data is about to be stored to disk, using some
method that can drill down into multiple levels of nested structured types.
That will require defining new APIs for structured types, though, so it
doesn't seem feasible as a back-patchable fix.
Note that this patch changes HeapTupleGetDatum() from a macro to a function
call; this means that any third-party code using that macro will not get
protection against creating TOAST-pointer-containing Datums until it's
recompiled. The same applies to any uses of PG_RETURN_HEAPTUPLEHEADER().
It seems likely that this is not a big problem in practice: most of the
tuple-returning functions in core and contrib produce outputs that could
not possibly be toasted anyway, and the same probably holds for third-party
extensions.
This bug has existed since TOAST was invented, so back-patch to all
supported branches.
If a tuple is locked, and this lock is later upgraded either to an
update or to a stronger lock, and in the meantime some other process
tries to lock, update or delete the same tuple, it (the tuple) could end
up being updated twice, or having conflicting locks held.
The reason for this is that the second updater checks for a change in
Xmax value, or in the HEAP_XMAX_IS_MULTI infomask bit, after noticing
the first lock; and if there's a change, it restarts and re-evaluates
its ability to update the tuple. But it neglected to check for changes
in lock strength or in lock-vs-update status when those two properties
stayed the same. This would lead it to take the wrong decision and
continue with its own update, when in reality it shouldn't do so but
instead restart from the top.
This could lead to either an assertion failure much later (when a
multixact containing multiple updates is detected), or duplicate copies
of tuples.
To fix, make sure to compare the other relevant infomask bits alongside
the Xmax value and HEAP_XMAX_IS_MULTI bit, and restart from the top if
necessary.
Also, in the belt-and-suspenders spirit, add a check to
MultiXactCreateFromMembers that a multixact being created does not have
two or more members that are claimed to be updates. This should protect
against other bugs that might cause similar bogus situations.
Backpatch to 9.3, where the possibility of multixacts containing updates
was introduced. (In prior versions it was possible to have the tuple
lock upgraded from shared to exclusive, and an update would not restart
from the top; yet we're protected against a bug there because there's
always a sleep to wait for the locking transaction to complete before
continuing to do anything. Really, the fact that tuple locks always
conflicted with concurrent updates is what protected against bugs here.)
Per report from Andrew Dunstan and Josh Berkus in thread at
http://www.postgresql.org/message-id/534C8B33.9050807@pgexperts.com
Bug analysis by Andres Freund.
When marking a branch as half-dead, a pointer to the top of the branch is
stored in the leaf block's hi-key. During normal operation, the high key
was left in place, and the block number was just stored in the ctid field
of the high key tuple, but in WAL replay, the high key was recreated as a
truncated tuple with zero columns. For the sake of easier debugging, also
truncate the tuple in normal operation, so that the page is identical
after WAL replay. Also, rename the 'downlink' field in the WAL record to
'topparent', as that seems like a more descriptive name. And make sure
it's set to invalid when unlinking the leaf page.
This allows squeezing out the unused space in full-page writes. And more
importantly, it can be a useful debugging aid.
In hindsight we should've done this back when GIN was added - we wouldn't
need the 'maxoff' field in the page opaque struct if we had used pd_lower
and pd_upper like on normal pages. But as long as there can be pages in the
index that have been binary-upgraded from pre-9.4 versions, we can't rely
on that, and have to continue using 'maxoff'.
Most of the code churn comes from renaming some macros, now that they're
used on internal pages, too.
This change is completely backwards-compatible, no effect on pg_upgrade.
Memory allocation can fail if you run out of memory, and inside a critical
section that will lead to a PANIC. Use conservatively-sized arrays in stack
instead.
There was previously no explicit limit on the number of pages a GiST split
can produce, it was only limited by the number of LWLocks that can be held
simultaneously (100 at the moment). This patch adds an explicit limit of 75
pages. That should be plenty, a typical split shouldn't produce more than
2-3 page halves.
The bug has been there forever, but only backpatch down to 9.1. The code
was changed significantly in 9.1, and it doesn't seem worth the risk or
trouble to adapt this for 9.0 and 8.4.
Inserting (in retail) into the new 9.4 format GIN posting tree created much
larger WAL records than in 9.3. The previous strategy to WAL logging was
basically to log the whole page on each change, with the exception of
completely unmodified segments up to the first modified one. That was not
too bad when appending to the end of the page, as only the last segment had
to be WAL-logged, but per Fujii Masao's testing, even that produced 2x the
WAL volume that 9.3 did.
The new strategy is to keep track of changes to the posting lists in a more
fine-grained fashion, and also make the repacking" code smarter to avoid
decoding and re-encoding segments unnecessarily.
The special feature the XLogInsert slots had over regular LWLocks is the
insertingAt value that was updated atomically with releasing backends
waiting on it. Add new functions to the LWLock API to do that, and replace
the slots with LWLocks. This reduces the amount of duplicated code.
(There's still some duplication, but at least it's all in lwlock.c now.)
Reviewed by Andres Freund.
It is no longer used, none of the resource managers have multi-record
actions that would make it unsafe to perform a restartpoint.
Also don't allow rm_cleanup to write WAL records, it's also no longer
required. Move the call to rm_cleanup routines to make it more symmetric
with rm_startup.
Splitting a page consists of two separate steps: splitting the child page,
and inserting the downlink for the new right page to the parent. Previously,
we handled the case that you crash in between those steps with a cleanup
routine after the WAL recovery had finished, which finished the incomplete
split. However, that doesn't help if the page split is interrupted but the
database doesn't crash, so that you don't perform WAL recovery. That could
happen for example if you run out of disk space.
Remove the end-of-recovery cleanup step. Instead, when a page is split, the
left page is marked with a new INCOMPLETE_SPLIT flag, and when the downlink
is inserted to the parent, the flag is cleared again. If an insertion sees
a page with the flag set, it knows that the split was interrupted for some
reason, and inserts the missing downlink before proceeding.
I used the same approach to fix GIN and GiST split algorithms earlier. This
was the last WAL cleanup routine, so we could get rid of that whole
machinery now, but I'll leave that for a separate patch.
Reviewed by Peter Geoghegan.
In short, we don't allow a page to be deleted if it's the rightmost child
of its parent, but that situation can change after we check for it.
Problem
-------
We check that the page to be deleted is not the rightmost child of its
parent, and then lock its left sibling, the page itself, its right sibling,
and the parent, in that order. However, if the parent page is split after
the check but before acquiring the locks, the target page might become the
rightmost child, if the split happens at the right place. That leads to an
error in vacuum (I reproduced this by setting a breakpoint in debugger):
ERROR: failed to delete rightmost child 41 of block 3 in index "foo_pkey"
We currently re-check that the page is still the rightmost child, and throw
the above error if it's not. We could easily just give up rather than throw
an error, but that approach doesn't scale to half-dead pages. To recap,
although we don't normally allow deleting the rightmost child, if the page
is the *only* child of its parent, we delete the child page and mark the
parent page as half-dead in one atomic operation. But before we do that, we
check that the parent can later be deleted, by checking that it in turn is
not the rightmost child of the grandparent (potentially recursing all the
way up to the root). But the same situation can arise there - the
grandparent can be split while we're not holding the locks. We end up with
a half-dead page that we cannot delete.
To make things worse, the keyspace of the deleted page has already been
transferred to its right sibling. As the README points out, the keyspace at
the grandparent level is "out-of-whack" until the half-dead page is deleted,
and if enough tuples with keys in the transferred keyspace are inserted, the
page might get split and a downlink might be inserted into the grandparent
that is out-of-order. That might not cause any serious problem if it's
transient (as the README ponders), but is surely bad if it stays that way.
Solution
--------
This patch changes the page deletion algorithm to avoid that problem. After
checking that the topmost page in the chain of to-be-deleted pages is not
the rightmost child of its parent, and then deleting the pages from bottom
up, unlink the pages from top to bottom. This way, the intermediate stages
are similar to the intermediate stages in page splitting, and there is no
transient stage where the keyspace is "out-of-whack". The topmost page in
the to-be-deleted chain doesn't have a downlink pointing to it, like a page
split before the downlink has been inserted.
This also allows us to get rid of the cleanup step after WAL recovery, if we
crash during page deletion. The deletion will be continued at next VACUUM,
but the tree is consistent for searches and insertions at every step.
This bug is old, all supported versions are affected, but this patch is too
big to back-patch (and changes the WAL record formats of related records).
We have not heard any reports of the bug from users, so clearly it's not
easy to bump into. Maybe backpatch later, after this has had some field
testing.
Reviewed by Kevin Grittner and Peter Geoghegan.
When a row is updated, and the new tuple version is put on the same page as
the old one, only WAL-log the part of the new tuple that's not identical to
the old. This saves significantly on the amount of WAL that needs to be
written, in the common case that most fields are not modified.
Amit Kapila, with a lot of back and forth with me, Robert Haas, and others.
With the GIN "fast scan" feature, GIN can skip items without fetching all
the keys for them, if it can prove that they don't match regardless of
those keys. So far, it has done the proving by calling the boolean
consistent function with all combinations of TRUE/FALSE for the unfetched
keys, but since that's O(n^2), it becomes unfeasible with more than a few
keys. We can avoid calling consistent with all the combinations, if we can
tell the operator class implementation directly which keys are unknown.
This commit includes a triConsistent function for the built-in array and
tsvector opclasses.
Alexander Korotkov, with some changes by me.
This feature, building on previous commits, allows the write-ahead log
stream to be decoded into a series of logical changes; that is,
inserts, updates, and deletes and the transactions which contain them.
It is capable of handling decoding even across changes to the schema
of the effected tables. The output format is controlled by a
so-called "output plugin"; an example is included. To make use of
this in a real replication system, the output plugin will need to be
modified to produce output in the format appropriate to that system,
and to perform filtering.
Currently, information can be extracted from the logical decoding
system only via SQL; future commits will add the ability to stream
changes via walsender.
Andres Freund, with review and other contributions from many other
people, including Álvaro Herrera, Abhijit Menon-Sen, Peter Gheogegan,
Kevin Grittner, Robert Haas, Heikki Linnakangas, Fujii Masao, Abhijit
Menon-Sen, Michael Paquier, Simon Riggs, Craig Ringer, and Steve
Singer.
If you have a GIN query like "rare & frequent", we currently fetch all the
items that match either rare or frequent, call the consistent function for
each item, and let the consistent function filter out items that only match
one of the terms. However, if we can deduce that "rare" must be present for
the overall qual to be true, we can scan all the rare items, and for each
rare item, skip over to the next frequent item with the same or greater TID.
That greatly speeds up "rare & frequent" type queries.
To implement that, introduce the concept of a tri-state consistent function,
where the 3rd value is MAYBE, indicating that we don't know if that term is
present. Operator classes only provide a boolean consistent function, so we
simulate the tri-state consistent function by calling the boolean function
several times, with the MAYBE arguments set to all combinations of TRUE and
FALSE. Testing all combinations is only feasible for a small number of MAYBE
arguments, but it is envisioned that we'll provide a way for operator
classes to provide a native tri-state consistent function, which can be much
more efficient. But that is not included in this patch.
We were already using that trick to for lossy pages, calling the consistent
function with the lossy entry set to TRUE and FALSE. Now that we have the
tri-state consistent function, use it for lossy pages too.
Alexander Korotkov, with fair amount of refactoring by me.
Replication slots are a crash-safe data structure which can be created
on either a master or a standby to prevent premature removal of
write-ahead log segments needed by a standby, as well as (with
hot_standby_feedback=on) pruning of tuples whose removal would cause
replication conflicts. Slots have some advantages over existing
techniques, as explained in the documentation.
In a few places, we refer to the type of replication slots introduced
by this patch as "physical" slots, because forthcoming patches for
logical decoding will also have slots, but with somewhat different
properties.
Andres Freund and Robert Haas
When skipping over some items in a posting tree, re-find the new location
by descending the tree from root, rather than walking the right links.
This can save a lot of I/O.
Heavily modified from Alexander Korotkov's fast scan patch.
If we're skipping past a certain TID, avoid decoding posting list segments
that only contain smaller TIDs.
Extracted from Alexander Korotkov's fast scan patch, heavily modified.
This makes it possible to store lwlocks as part of some other data
structure in the main shared memory segment, or in a dynamic shared
memory segment. There is still a main LWLock array and this patch does
not move anything out of it, but it provides necessary infrastructure
for doing that in the future.
This change is likely to increase the size of LWLockPadded on some
platforms, especially 32-bit platforms where it was previously only
16 bytes.
Patch by me. Review by Andres Freund and KaiGai Kohei.
This allows ending recovery as a consistent state has been reached. Without
this, there was no easy way to e.g restore an online backup, without
replaying any extra WAL after the backup ended.
MauMau and me.
GIN posting lists are now encoded using varbyte-encoding, which allows them
to fit in much smaller space than the straight ItemPointer array format used
before. The new encoding is used for both the lists stored in-line in entry
tree items, and in posting tree leaf pages.
To maintain backwards-compatibility and keep pg_upgrade working, the code
can still read old-style pages and tuples. Posting tree leaf pages in the
new format are flagged with GIN_COMPRESSED flag, to distinguish old and new
format pages. Likewise, entry tree tuples in the new format have a
GIN_ITUP_COMPRESSED flag set in a bit that was previously unused.
This patch bumps GIN_CURRENT_VERSION from 1 to 2. New indexes created with
version 9.4 will therefore have version number 2 in the metapage, while old
pg_upgraded indexes will have version 1. The code treats them the same, but
it might be come handy in the future, if we want to drop support for the
uncompressed format.
Alexander Korotkov and me. Reviewed by Tomas Vondra and Amit Langote.
In ordinary operation, VACUUM must be careful to take a cleanup lock on
each leaf page of a btree index; this ensures that no indexscans could
still be "in flight" to heap tuples due to be deleted. (Because of
possible index-tuple motion due to concurrent page splits, it's not enough
to lock only the pages we're deleting index tuples from.) In Hot Standby,
the WAL replay process must likewise lock every leaf page. There were
several bugs in the code for that:
* The replay scan might come across unused, all-zero pages in the index.
While btree_xlog_vacuum itself did the right thing (ie, nothing) with
such pages, xlogutils.c supposed that such pages must be corrupt and
would throw an error. This accounts for various reports of replication
failures with "PANIC: WAL contains references to invalid pages". To
fix, add a ReadBufferMode value that instructs XLogReadBufferExtended
not to complain when we're doing this.
* btree_xlog_vacuum performed the extra locking if standbyState ==
STANDBY_SNAPSHOT_READY, but that's not the correct test: we won't open up
for hot standby queries until the database has reached consistency, and
we don't want to do the extra locking till then either, for fear of reading
corrupted pages (which bufmgr.c would complain about). Fix by exporting a
new function from xlog.c that will report whether we're actually in hot
standby replay mode.
* To ensure full coverage of the index in the replay scan, btvacuumscan
would emit a dummy WAL record for the last page of the index, if no
vacuuming work had been done on that page. However, if the last page
of the index is all-zero, that would result in corruption of said page,
since the functions called on it weren't prepared to handle that case.
There's no need to lock any such pages, so change the logic to target
the last normal leaf page instead.
The first two of these bugs were diagnosed by Andres Freund, the other one
by me. Fixes based on ideas from Heikki Linnakangas and myself.
This has been wrong since Hot Standby was introduced, so back-patch to 9.0.
In pg_multixact/members, relying on modulo-2^32 arithmetic for
wraparound handling doesn't work all that well. Because we don't
explicitely track wraparound of the allocation counter for members, it
is possible that the "live" area exceeds 2^31 entries; trying to remove
SLRU segments that are "old" according to the original logic might lead
to removal of segments still in use. To fix, have the truncation
routine use a tailored SlruScanDirectory callback that keeps track of
the live area in actual use; that way, when the live range exceeds 2^31
entries, the oldest segments still live will not get removed untimely.
This new SlruScanDir callback needs to take care not to remove segments
that are "in the future": if new SLRU segments appear while the
truncation is ongoing, make sure we don't remove them. This requires
examination of shared memory state to recheck for false positives, but
testing suggests that this doesn't cause a problem. The original coding
didn't suffer from this pitfall because segments created when truncation
is running are never considered to be removable.
Per Andres Freund's investigation of bug #8673 reported by Serge
Negodyuck.
Instead of changing the tuple xmin to FrozenTransactionId, the combination
of HEAP_XMIN_COMMITTED and HEAP_XMIN_INVALID, which were previously never
set together, is now defined as HEAP_XMIN_FROZEN. A variety of previous
proposals to freeze tuples opportunistically before vacuum_freeze_min_age
is reached have foundered on the objection that replacing xmin by
FrozenTransactionId might hinder debugging efforts when things in this
area go awry; this patch is intended to solve that problem by keeping
the XID around (but largely ignoring the value to which it is set).
Third-party code that checks for HEAP_XMIN_INVALID on tuples where
HEAP_XMIN_COMMITTED might be set will be broken by this change. To fix,
use the new accessor macros in htup_details.h rather than consulting the
bits directly. HeapTupleHeaderGetXmin has been modified to return
FrozenTransactionId when the infomask bits indicate that the tuple is
frozen; use HeapTupleHeaderGetRawXmin when you already know that the
tuple isn't marked commited or frozen, or want the raw value anyway.
We currently do this in routines that display the xmin for user consumption,
in tqual.c where it's known to be safe and important for the avoidance of
extra cycles, and in the function-caching code for various procedural
languages, which shouldn't invalidate the cache just because the tuple
gets frozen.
Robert Haas and Andres Freund
If a tuple was locked by transaction A, and transaction B updated it,
the new version of the tuple created by B would be locked by A, yet
visible only to B; due to an oversight in HeapTupleSatisfiesUpdate, the
lock held by A wouldn't get checked if transaction B later deleted (or
key-updated) the new version of the tuple. This might cause referential
integrity checks to give false positives (that is, allow deletes that
should have been rejected).
This is an easy oversight to have made, because prior to improved tuple
locks in commit 0ac5ad5134 it wasn't possible to have tuples created by
our own transaction that were also locked by remote transactions, and so
locks weren't even considered in that code path.
It is recommended that foreign keys be rechecked manually in bulk after
installing this update, in case some referenced rows are missing with
some referencing row remaining.
Per bug reported by Daniel Wood in
CAPweHKe5QQ1747X2c0tA=5zf4YnS2xcvGf13Opd-1Mq24rF1cQ@mail.gmail.com
Tuple freezing was broken in connection to MultiXactIds; commit
8e53ae025d tried to fix it, but didn't go far enough. As noted by
Noah Misch, freezing a tuple whose Xmax is a multi containing an aborted
update might cause locks in the multi to go ignored by later
transactions. This is because the code depended on a multixact above
their cutoff point not having any lock-only member older than the cutoff
point for Xids, which is easily defeated in READ COMMITTED transactions.
The fix for this involves creating a new MultiXactId when necessary.
But this cannot be done during WAL replay, and moreover multixact
examination requires using CLOG access routines which are not supposed
to be used during WAL replay either; so tuple freezing cannot be done
with the old freeze WAL record. Therefore, separate the freezing
computation from its execution, and change the WAL record to carry all
necessary information. At WAL replay time, it's easy to re-execute
freezing because we don't need to re-compute the new infomask/Xmax
values but just take them from the WAL record.
While at it, restructure the coding to ensure all page changes occur in
a single critical section without much room for failures. The previous
coding wasn't using a critical section, without any explanation as to
why this was acceptable.
In replication scenarios using the 9.3 branch, standby servers must be
upgraded before their master, so that they are prepared to deal with the
new WAL record once the master is upgraded; failure to do so will cause
WAL replay to die with a PANIC message. Later upgrade of the standby
will allow the process to continue where it left off, so there's no
disruption of the data in the standby in any case. Standbys know how to
deal with the old WAL record, so it's okay to keep the master running
the old code for a while.
In master, the old freeze WAL record is gone, for cleanliness' sake;
there's no compatibility concern there.
Backpatch to 9.3, where the original bug was introduced and where the
previous fix was backpatched.
Álvaro Herrera and Andres Freund
WAL records of hint bit updates is useful to tools that want to examine
which pages have been modified. In particular, this is required to make
the pg_rewind tool safe (without checksums).
This can also be used to test how much extra WAL-logging would occur if
you enabled checksums, without actually enabling them (which you can't
currently do without re-initdb'ing).
Sawada Masahiko, docs by Samrat Revagade. Reviewed by Dilip Kumar, with
further changes by me.
When wal_level=logical, we'll log columns from the old tuple as
configured by the REPLICA IDENTITY facility added in commit
07cacba983. This makes it possible
a properly-configured logical replication solution to correctly
follow table updates even if they change the chosen key columns,
or, with REPLICA IDENTITY FULL, even if the table has no key at
all. Note that updates which do not modify the replica identity
column won't log anything extra, making the choice of a good key
(i.e. one that will rarely be changed) important to performance
when wal_level=logical is configured.
Each insert, update, or delete to a catalog table will also log
the CMIN and/or CMAX values of stamped by the current transaction.
This is necessary because logical decoding will require access to
historical snapshots of the catalog in order to decode some data
types, and the CMIN/CMAX values that we may need in order to judge
row visibility may have been overwritten by the time we need them.
Andres Freund, reviewed in various versions by myself, Heikki
Linnakangas, KONDO Mitsumasa, and many others.
Commit 9dc842f08 of 8.2 era prevented MultiXact truncation during crash
recovery, because there was no guarantee that enough state had been
setup, and because it wasn't deemed to be a good idea to remove data
during crash recovery anyway. Since then, due to Hot-Standby, streaming
replication and PITR, the amount of time a cluster can spend doing crash
recovery has increased significantly, to the point that a cluster may
even never come out of it. This has made not truncating the content of
pg_multixact/ not defensible anymore.
To fix, take care to setup enough state for multixact truncation before
crash recovery starts (easy since checkpoints contain the required
information), and move the current end-of-recovery actions to a new
TrimMultiXact() function, analogous to TrimCLOG().
At some later point, this should probably done similarly to the way
clog.c is doing it, which is to just WAL log truncations, but we can't
do that for the back branches.
Back-patch to 9.0. 8.4 also has the problem, but since there's no hot
standby there, it's much less pressing. In 9.2 and earlier, this patch
is simpler than in newer branches, because multixact access during
recovery isn't required. Add appropriate checks to make sure that's not
happening.
Andres Freund
While autovacuum dutifully launched anti-multixact-wraparound vacuums
when the multixact "age" was reached, the vacuum code was not aware that
it needed to make them be full table vacuums. As the resulting
partial-table vacuums aren't capable of actually increasing relminmxid,
autovacuum continued to launch anti-wraparound vacuums that didn't have
the intended effect, until age of relfrozenxid caused the vacuum to
finally be a full table one via vacuum_freeze_table_age.
To fix, introduce logic for multixacts similar to that for plain
TransactionIds, using the same GUCs.
Backpatch to 9.3, where permanent MultiXactIds were introduced.
Andres Freund, some cleanup by Álvaro
Replace it with an approach similar to what GiST uses: when a page is split,
the left sibling is marked with a flag indicating that the parent hasn't been
updated yet. When the parent is updated, the flag is cleared. If an insertion
steps on a page with the flag set, it will finish split before proceeding
with the insertion.
The post-recovery cleanup mechanism was never totally reliable, as insertion
to the parent could fail e.g because of running out of memory or disk space,
leaving the tree in an inconsistent state.
This also divides the responsibility of WAL-logging more clearly between
the generic ginbtree.c code, and the parts specific to entry and posting
trees. There is now a common WAL record format for insertions and deletions,
which is written by ginbtree.c, followed by tree-specific payload, which is
returned by the placetopage- and split- callbacks.
Separate the insertion payload from the more static portions of GinBtree.
GinBtree now only contains information related to searching the tree, and
the information of what to insert is passed separately.
Add root block number to GinBtree, instead of passing it around all the
functions as argument.
Split off ginFinishSplit() from ginInsertValue(). ginFinishSplit is
responsible for finding the parent and inserting the downlink to it.
Change SET LOCAL/CONSTRAINTS/TRANSACTION behavior outside of a
transaction block from error (post-9.3) to warning. (Was nothing in <=
9.3.) Also change ABORT outside of a transaction block from notice to
warning.
This patch adds the ability to write TABLE( function1(), function2(), ...)
as a single FROM-clause entry. The result is the concatenation of the
first row from each function, followed by the second row from each
function, etc; with NULLs inserted if any function produces fewer rows than
others. This is believed to be a much more useful behavior than what
Postgres currently does with multiple SRFs in a SELECT list.
This syntax also provides a reasonable way to combine use of column
definition lists with WITH ORDINALITY: put the column definition list
inside TABLE(), where it's clear that it doesn't control the ordinality
column as well.
Also implement SQL-compliant multiple-argument UNNEST(), by turning
UNNEST(a,b,c) into TABLE(unnest(a), unnest(b), unnest(c)).
The SQL standard specifies TABLE() with only a single function, not
multiple functions, and it seems to require an implicit UNNEST() which is
not what this patch does. There may be something wrong with that reading
of the spec, though, because if it's right then the spec's TABLE() is just
a pointless alternative spelling of UNNEST(). After further review of
that, we might choose to adopt a different syntax for what this patch does,
but in any case this functionality seems clearly worthwhile.
Andrew Gierth, reviewed by Zoltán Böszörményi and Heikki Linnakangas, and
significantly revised by me
This creates a new gin-btree callback function for creating a downlink for
a page. Previously, ginxlog.c duplicated the logic used during normal
operation.
If a page is deleted, and reused for something else, just as a search is
following a rightlink to it from its left sibling, the search would continue
scanning whatever the new contents of the page are. That could lead to
incorrect query results, or even something more curious if the page is
reused for a different kind of a page.
To fix, modify the search algorithm to lock the next page before releasing
the previous one, and refrain from deleting pages from the leftmost branch
of the tree.
Add a new Concurrency section to the README, explaining why this works.
There is a lot more one could say about concurrency in GIN, but that's for
another patch.
Backpatch to all supported versions.
Merge the isEnoughSpace and placeToPage functions in the b-tree interface
into one function that tries to put a tuple on page, and returns false if
it doesn't fit.
Move createPostingTree function to gindatapage.c, and change its contract
so that it can be passed more items than fit on the root page. It's in a
better position than the callers to know how many items fit.
Move ginMergeItemPointers out of gindatapage.c, into a separate file.
These changes make no difference now, but reduce the footprint of Alexander
Korotkov's upcoming patch to pack item pointers more tightly.
This avoids an assumption about the signed number representation. It is
anticipated to have no functional changes on supported configurations;
many two's complement assumptions remain elsewhere.
Per a suggestion from Andres Freund.
It makes for cleaner code to have separate Get/Add functions for PostingItems
and ItemPointers. A few callsites that have to deal with both types need to
be duplicated because of this, but all the callers have to know which one
they're dealing with anyway. Overall, this reduces the amount of casting
required.
Extracted from Alexander Korotkov's larger patch to change the data page
format.
It seems to make more sense to use "cutoff multixact" terminology
throughout the backend code; "freeze" is associated with replacing of an
Xid with FrozenTransactionId, which is not what we do for MultiXactIds.
Andres Freund
Some adjustments by Álvaro Herrera
When upgrading from servers of versions 9.2 and older, and MultiXactIds
have been used in the old server beyond the first page (that is, 2048
multis or more in the default 8kB-page build), pg_upgrade would set the
next multixact offset to use beyond what has been allocated in the new
cluster. This would cause a failure the first time the new cluster
needs to use this value, because the pg_multixact/offsets/ file wouldn't
exist or wouldn't be large enough. To fix, ensure that the transient
server instances launched by pg_upgrade extend the file as necessary.
Per report from Jesse Denardo in
CANiVXAj4c88YqipsyFQPboqMudnjcNTdB3pqe8ReXqAFQ=HXyA@mail.gmail.com
This patch replaces WALInsertLock with a number of WAL insertion slots,
allowing multiple backends to insert WAL records to the WAL buffers
concurrently. This is particularly useful for parallel loading large amounts
of data on a system with many CPUs.
This has one user-visible change: switching to a new WAL segment with
pg_switch_xlog() now fills the remaining unused portion of the segment with
zeros. This potentially adds some overhead, but it has been a very common
practice by DBA's to clear the "tail" of the segment with an external
pg_clearxlogtail utility anyway, to make the WAL files compress better.
With this patch, it's no longer necessary to do that.
This patch adds a new GUC, xloginsert_slots, to tune the number of WAL
insertion slots. Performance testing suggests that the default, 8, works
pretty well for all kinds of worklods, but I left the GUC in place to allow
others with different hardware to test that easily. We might want to remove
that before release.
Reviewed by Andres Freund.
In 9.3, there's no particular limit on the number of bgworkers;
instead, we just count up the number that are actually registered,
and use that to set MaxBackends. However, that approach causes
problems for Hot Standby, which needs both MaxBackends and the
size of the lock table to be the same on the standby as on the
master, yet it may not be desirable to run the same bgworkers in
both places. 9.3 handles that by failing to notice the problem,
which will probably work fine in nearly all cases anyway, but is
not theoretically sound.
A further problem with simply counting the number of registered
workers is that new workers can't be registered without a
postmaster restart. This is inconvenient for administrators,
since bouncing the postmaster causes an interruption of service.
Moreover, there are a number of applications for background
processes where, by necessity, the background process must be
started on the fly (e.g. parallel query). While this patch
doesn't actually make it possible to register new background
workers after startup time, it's a necessary prerequisite.
Patch by me. Review by Michael Paquier.
Treat TOAST index just the same as normal one and get the OID
of TOAST index from pg_index but not pg_class.reltoastidxid.
This change allows us to handle multiple TOAST indexes, and
which is required infrastructure for upcoming
REINDEX CONCURRENTLY feature.
Patch by Michael Paquier, reviewed by Andres Freund and me.
To that end, support tags rather than lengths for external datums.
As an example of how this can be used, add support or "indirect"
tuples which point to some externally allocated memory containing
a toast tuple. Similar infrastructure could be used for other
purposes, including, perhaps, support for alternative compression
algorithms.
Andres Freund, reviewed by Hitoshi Harada and myself
SnapshotNow scans have the undesirable property that, in the face of
concurrent updates, the scan can fail to see either the old or the new
versions of the row. In many cases, we work around this by requiring
DDL operations to hold AccessExclusiveLock on the object being
modified; in some cases, the existing locking is inadequate and random
failures occur as a result. This commit doesn't change anything
related to locking, but will hopefully pave the way to allowing lock
strength reductions in the future.
The major issue has held us back from making this change in the past
is that taking an MVCC snapshot is significantly more expensive than
using a static special snapshot such as SnapshotNow. However, testing
of various worst-case scenarios reveals that this problem is not
severe except under fairly extreme workloads. To mitigate those
problems, we avoid retaking the MVCC snapshot for each new scan;
instead, we take a new snapshot only when invalidation messages have
been processed. The catcache machinery already requires that
invalidation messages be sent before releasing the related heavyweight
lock; else other backends might rely on locally-cached data rather
than scanning the catalog at all. Thus, making snapshot reuse
dependent on the same guarantees shouldn't break anything that wasn't
already subtly broken.
Patch by me. Review by Michael Paquier and Andres Freund.
MarkBufferDirtyHint() writes WAL, and should know if it's got a
standard buffer or not. Currently, the only callers where buffer_std
is false are related to the FSM.
In passing, rename XLOG_HINT to XLOG_FPI, which is more descriptive.
Back-patch to 9.3.
SP-GiST's original scheme for avoiding deadlocks during concurrent index
insertions doesn't work, as per report from Hailong Li, and there isn't any
evident way to make it work completely. We could possibly lock individual
inner tuples instead of their whole pages, but preliminary experimentation
suggests that the performance penalty would be huge. Instead, if we fail
to get a buffer lock while descending the tree, just restart the tree
descent altogether. We keep the old tuple positioning rules, though, in
hopes of reducing the number of cases where this can happen.
Teodor Sigaev, somewhat edited by Tom Lane
Checksums are set immediately prior to flush out of shared buffers
and checked when pages are read in again. Hint bit setting will
require full page write when block is dirtied, which causes various
infrastructure changes. Extensive comments, docs and README.
WARNING message thrown if checksum fails on non-all zeroes page;
ERROR thrown but can be disabled with ignore_checksum_failure = on.
Feature enabled by an initdb option, since transition from option off
to option on is long and complex and has not yet been implemented.
Default is not to use checksums.
Checksum used is WAL CRC-32 truncated to 16-bits.
Simon Riggs, Jeff Davis, Greg Smith
Wide input and assistance from many community members. Thank you.
If we were about to enter archive recovery after crash recovery, we scanned
the archive for the latest tli history file, and set the recovery target
timeline to that. However, when we actually tried to read the history file,
we would not fetch the file from the archive, because we were not in archive
recovery yet.
To fix, make readTimeLineHistory and existsTimeLineHistory to always fetch
the file from archive if archive recovery is requested, even if we're not in
archive recovery yet.
Backpatch to 9.2. Mitsumasa KONDO
fmgr_sql had been designed on the assumption that the FmgrInfo it's called
with has only query lifespan. This is demonstrably unsafe in connection
with range types, as shown in bug #7881 from Andrew Gierth. Fix things
so that we re-generate the function's cache data if the (sub)transaction
it was made in is no longer active.
Back-patch to 9.2. This might be needed further back, but it's not clear
whether the case can realistically arise without range types, so for now
I'll desist from back-patching further.
This is a forward-patch of commit 6f4b8a4f4f,
applied to 9.2 back in August. The plan was to do something else in master,
but it looks like it's not going to happen, so let's just apply the 9.2
solution to master as well.
Fujii Masao
Currently it's only possible for loadable modules to get control during
post-commit cleanup of a transaction. That doesn't work too well if they
want to do something that could throw an error; for example, an FDW might
need to issue a remote commit, which could well fail. To improve matters,
extend the existing APIs for XactCallback and SubXactCallback functions
to provide new pre-commit events for this purpose.
The release notes will need to mention that existing callback functions
should be checked to make sure they don't do something unwanted when one
of the new event types occurs. In the examples within our source tree,
contrib/sepgsql was fine but plpgsql had been a bit too cute.
The reason this wasn't supported before was that GiST indexes need an
increasing sequence to detect concurrent page-splits. In a regular WAL-
logged GiST index, the LSN of the page-split record is used for that
purpose, and in a temporary index, we can get away with a backend-local
counter. Neither of those methods works for an unlogged relation.
To provide such an increasing sequence of numbers, create a "fake LSN"
counter that is saved and restored across shutdowns. On recovery, unlogged
relations are blown away, so the counter doesn't need to survive that
either.
Jeevan Chalke, based on discussions with Robert Haas, Tom Lane and me.
This isn't used for anything but a sanity check at the moment, but it could
be highly valuable for debugging purposes. It could also be used to recreate
timeline history by traversing WAL, which seems useful.
Improve comments, rename some variables and functions, slightly simplify
a couple of APIs, in an attempt to make this code readable by people other
than its original author.
Even though this is essentially just cosmetic, back-patch to all active
branches, because otherwise it's going to make back-patching future fixes
in this file very painful.
When considering a non-last column in a multi-column GiST index,
gistsplit.c tries to improve on the split chosen by the opclass-specific
pickSplit function by considering penalties for the next column. However,
there were two bugs in this code: it failed to recompute the union keys for
the leftmost index columns, even though these might well change after
reassigning tuples; and it included the old union keys in the recomputation
for the columns it did recompute, so that those keys couldn't get smaller
even if they should. The first problem could result in an invalid index
in which searches wouldn't find index entries that are in fact present;
the second would make the index less efficient to search.
Both of these errors were caused by misuse of gistMakeUnionItVec, whose
API was designed in a way that just begged such errors to be made. There
is no situation in which it's safe or useful to compute the union keys for
a subset of the index columns, and there is no caller that wants any
previous union keys to be included in the computation; so the undocumented
choice to treat the union keys as in/out rather than pure output parameters
is a waste of code as well as being dangerous.
Hence, rather than just making a minimal patch, I've changed the API of
gistMakeUnionItVec to remove the "startkey" parameter (it now always
processes all index columns) and treat the attr/isnull arrays as purely
output parameters.
In passing, also get rid of a couple of unnecessary and dangerous uses
of static variables in gistutil.c. It's remarkable that the one in
gistMakeUnionKey hasn't given us portability troubles before now, because
in addition to posing a re-entrancy hazard, it was unsafely assuming that
a static char[] array would have at least Datum alignment.
Per investigation of a trouble report from Tomas Vondra. (There are also
some bugs in contrib/btree_gist to be fixed, but that seems like material
for a separate patch.) Back-patch to all supported branches.
The new rmgrlist.h header, containing all necessary data
about built-in resource managers, allows other pieces of code to
access them.
In particular, this allows a future pg_xlogdump program to extract
rm_desc function pointers, without having to keep a duplicate list of
them.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
pg_ctl promote -m fast will skip the checkpoint at end of recovery so that we
can achieve very fast failover when the apply delay is low. Write new WAL record
XLOG_END_OF_RECOVERY to allow us to switch timeline correctly for downstream log
readers. If we skip synchronous end of recovery checkpoint we request a normal
spread checkpoint so that the window of re-recovery is low.
Simon Riggs and Kyotaro Horiguchi, with input from Fujii Masao.
Review by Heikki Linnakangas
Tuples marked SELECT FOR UPDATE in a cluster that's later processed by
pg_upgrade would have a different infomask bit pattern than those
produced by 9.3dev; that bit pattern was being seen as "dead" by HEAD
(because they would fail the "is this tuple locked" test, and so the
visibility rules would thing they're updated, even though there's no
HEAP_UPDATED version of them). In other words, some rows could silently
disappear after pg_upgrade.
With this new definition, those tuples become visible again.
This is breakage resulting from my commit 0ac5ad5134.
This patch introduces two additional lock modes for tuples: "SELECT FOR
KEY SHARE" and "SELECT FOR NO KEY UPDATE". These don't block each
other, in contrast with already existing "SELECT FOR SHARE" and "SELECT
FOR UPDATE". UPDATE commands that do not modify the values stored in
the columns that are part of the key of the tuple now grab a SELECT FOR
NO KEY UPDATE lock on the tuple, allowing them to proceed concurrently
with tuple locks of the FOR KEY SHARE variety.
Foreign key triggers now use FOR KEY SHARE instead of FOR SHARE; this
means the concurrency improvement applies to them, which is the whole
point of this patch.
The added tuple lock semantics require some rejiggering of the multixact
module, so that the locking level that each transaction is holding can
be stored alongside its Xid. Also, multixacts now need to persist
across server restarts and crashes, because they can now represent not
only tuple locks, but also tuple updates. This means we need more
careful tracking of lifetime of pg_multixact SLRU files; since they now
persist longer, we require more infrastructure to figure out when they
can be removed. pg_upgrade also needs to be careful to copy
pg_multixact files over from the old server to the new, or at least part
of multixact.c state, depending on the versions of the old and new
servers.
Tuple time qualification rules (HeapTupleSatisfies routines) need to be
careful not to consider tuples with the "is multi" infomask bit set as
being only locked; they might need to look up MultiXact values (i.e.
possibly do pg_multixact I/O) to find out the Xid that updated a tuple,
whereas they previously were assured to only use information readily
available from the tuple header. This is considered acceptable, because
the extra I/O would involve cases that would previously cause some
commands to block waiting for concurrent transactions to finish.
Another important change is the fact that locking tuples that have
previously been updated causes the future versions to be marked as
locked, too; this is essential for correctness of foreign key checks.
This causes additional WAL-logging, also (there was previously a single
WAL record for a locked tuple; now there are as many as updated copies
of the tuple there exist.)
With all this in place, contention related to tuples being checked by
foreign key rules should be much reduced.
As a bonus, the old behavior that a subtransaction grabbing a stronger
tuple lock than the parent (sub)transaction held on a given tuple and
later aborting caused the weaker lock to be lost, has been fixed.
Many new spec files were added for isolation tester framework, to ensure
overall behavior is sane. There's probably room for several more tests.
There were several reviewers of this patch; in particular, Noah Misch
and Andres Freund spent considerable time in it. Original idea for the
patch came from Simon Riggs, after a problem report by Joel Jacobson.
Most code is from me, with contributions from Marti Raudsepp, Alexander
Shulgin, Noah Misch and Andres Freund.
This patch was discussed in several pgsql-hackers threads; the most
important start at the following message-ids:
AANLkTimo9XVcEzfiBR-ut3KVNDkjm2Vxh+t8kAmWjPuv@mail.gmail.com1290721684-sup-3951@alvh.no-ip.org1294953201-sup-2099@alvh.no-ip.org1320343602-sup-2290@alvh.no-ip.org1339690386-sup-8927@alvh.no-ip.org4FE5FF020200002500048A3D@gw.wicourts.gov4FEAB90A0200002500048B7D@gw.wicourts.gov
When a standby server follows the master using WAL archive, and it chooses
a new timeline (recovery_target_timeline='latest'), it only fetches the
timeline history file for the chosen target timeline, not any other history
files that might be missing from pg_xlog. For example, if the current
timeline is 2, and we choose 4 as the new recovery target timeline, the
history file for timeline 3 is not fetched, even if it's part of this
server's history. That's enough for the standby itself - the history file
for timeline 4 includes timeline 3 as well - but if a cascading standby
server wants to recover to timeline 3, it needs the history file. To fix,
when a new recovery target timeline is chosen, try to copy any missing
history files from the archive to pg_xlog between the old and new target
timeline.
A second similar issue was with the WAL files. When a standby recovers from
archive, and it reaches a segment that contains a switch to a new timeline,
recovery fetches only the WAL file labelled with the new timeline's ID. The
file from the new timeline contains a copy of the WAL from the old timeline
up to the point where the switch happened, and recovery recovers it from the
new file. But in streaming replication, walsender only tries to read it
from the old timeline's file. To fix, change walsender to read it from the
new file, so that it behaves the same as recovery in that sense, and doesn't
try to open the possibly nonexistent file with the old timeline's ID.
The xlogreader refactoring broke the logic to decide which timeline to start
streaming from. XLogPageRead() uses the timeline history to check which
timeline the requested WAL position falls into. However, after the
refactoring, XLogPageRead() is always first called with the first page in
the segment, to verify the segment header, and only then with the actual WAL
position we're interested in. That first read of the segment's header made
XLogPageRead() to always start streaming from the old timeline containing
the segment header, not the timeline containing the actual record, if there
was a timeline switch within the segment.
I thought I fixed this yesterday, but that fix was too narrow and only fixed
this for the corner-case that the timeline switch happened in the first page
of the segment. To fix this more robustly, pass explicitly the position of
the record we're actually interested in to XLogPageRead, and use that to
decide which timeline to read from, rather than deduce it from the page and
offset.
Per report from Fujii Masao.
This mirrors the changes done earlier to the server in standby mode. When
receivelog reaches the end of a timeline, as reported by the server, it
fetches the timeline history file of the next timeline, and restarts
streaming from the new timeline by issuing a new START_STREAMING command.
When pg_receivexlog crosses a timeline, it leaves the .partial suffix on the
last segment on the old timeline. This helps you to tell apart a partial
segment left in the directory because of a timeline switch, and a completed
segment. If you just follow a single server, it won't make a difference, but
it can be significant in more complicated scenarios where new WAL is still
generated on the old timeline.
This includes two small changes to the streaming replication protocol:
First, when you reach the end of timeline while streaming, the server now
sends the TLI of the next timeline in the server's history to the client.
pg_receivexlog uses that as the next timeline, so that it doesn't need to
parse the timeline history file like a standby server does. Second, when
BASE_BACKUP command sends the begin and end WAL positions, it now also sends
the timeline IDs corresponding the positions.
The patch that turned XLogRecPtr into a uint64 inadvertently changed the
on-disk format of GiST indexes, because the NSN field in the GiST page
opaque is an XLogRecPtr. That breaks pg_upgrade. Revert the format of that
field back to the two-field struct that XLogRecPtr was before. This is the
same we did to LSNs in the page header to avoid changing on-disk format.
Bump catversion, as this invalidates any existing GiST indexes built on
9.3devel.
This new facility can not only be used by xlog.c to carry out crash
recovery, but also by external programs. By supplying a function to
read XLog pages from somewhere, all the WAL reading can be used for
completely different purposes.
For the standard backend use, the behavior should be pretty much the
same as previously. As for non-backend programs, an hypothetical
pg_xlogdump program is now closer to reality, but some more backend
support is still necessary.
This patch was originally submitted by Andres Freund in a different
form, but Heikki Linnakangas opted for and authored another design of
the concept. Andres has advanced the patch since Heikki's initial
version. Review and some (mostly cosmetics) changes by me.
Historically we've used a couple of very ad-hoc fudge factors to try to
get the right results when indexes of different sizes would satisfy a
query with the same number of index leaf tuples being visited. In
commit 21a39de580 I tweaked one of these
fudge factors, with results that proved disastrous for larger indexes.
Commit bf01e34b55 fudged it some more,
but still with not a lot of principle behind it.
What seems like a better way to address these issues is to explicitly model
index-descent costs, since that's what's really at stake when considering
diferent indexes with similar leaf-page-level costs. We tried that once
long ago, and found that charging random_page_cost per page descended
through was way too much, because upper btree levels tend to stay in cache
in real-world workloads. However, there's still CPU costs to think about,
and the previous fudge factors can be seen as a crude attempt to account
for those costs. So this patch replaces those fudge factors with explicit
charges for the number of tuple comparisons needed to descend the index
tree, plus a small charge per page touched in the descent. The cost
multipliers are chosen so that the resulting charges are in the vicinity of
the historical (pre-9.2) fudge factors for indexes of up to about a million
tuples, while not ballooning unreasonably beyond that, as the old fudge
factor did (even more so in 9.2).
To make this work accurately for btree indexes, add some code that allows
extraction of the known root-page height from a btree. There's no
equivalent number readily available for other index types, but we can use
the log of the number of index pages as an approximate substitute.
This seems like too much of a behavioral change to risk back-patching,
but it should improve matters going forward. In 9.2 I'll just revert
the fudge-factor change.
If you take a base backup from a standby server with "pg_basebackup -X
fetch", and the timeline switches while the backup is being taken, the
backup used to fail with an error "requested WAL segment %s has already
been removed". This is because the server-side code that sends over the
required WAL files would not construct the WAL filename with the correct
timeline after a switch.
Fix that by using readdir() to scan pg_xlog for all the WAL segments in the
range, regardless of timeline.
Also, include all timeline history files in the backup, if taken with
"-X fetch". That fixes another related bug: If a timeline switch happened
just before the backup was initiated in a standby, the WAL segment
containing the initial checkpoint record contains WAL from the older
timeline too. Recovery will not accept that without a timeline history file
that lists the older timeline.
Backpatch to 9.2. Versions prior to that were not affected as you could not
take a base backup from a standby before 9.2.
The cascading standby patch in 9.2 changed the way WAL files are treated
when restored from the archive. Before, they were restored under a temporary
filename, and not kept in pg_xlog, but after the patch, they were copied
under pg_xlog. This is necessary for a cascading standby to find them, but
it also means that if the archive goes offline and a standby is restarted,
it can recover back to where it was using the files in pg_xlog. It also
means that if you take an offline backup from a standby server, it includes
all the required WAL files in pg_xlog.
However, the same change was not made to timeline history files, so if the
WAL segment containing the checkpoint record contains a timeline switch, you
will still get an error if you try to restart recovery without the archive,
or recover from an offline backup taken from the standby.
With this patch, timeline history files restored from archive are copied
into pg_xlog like WAL files are, so that pg_xlog contains all the files
required to recover. This is a corner-case pre-existing issue in 9.2, but
even more important in master where it's possible for a standby to follow a
timeline switch through streaming replication. To make that possible, the
timeline history files must be present in pg_xlog.
This gets rid of XLByteLT, XLByteLE, XLByteEQ and XLByteAdvance.
These were useful for brevity when XLogRecPtrs were split in
xlogid/xrecoff; but now that they are simple uint64's, they are just
clutter. The only downside to making this change would be ease of
backporting patches, but that has been negated by other substantive
changes to the involved code anyway. The clarity of simpler expressions
makes the change worthwhile.
Most of the changes are mechanical, but in a couple of places, the patch
author chose to invert the operator sense, making the code flow more
logical (and more in line with preceding comments).
Author: Andres Freund
Eyeballed by Dimitri Fontaine and Alvaro Herrera
Most of the time, the last replayed record comes from the recovery target
timeline, but there is a corner case where it makes a difference. When
the startup process scans for a new timeline, and decides to change recovery
target timeline, there is a window where the recovery target TLI has already
been bumped, but there are no WAL segments from the new timeline in pg_xlog
yet. For example, if we have just replayed up to point 0/30002D8, on
timeline 1, there is a WAL file called 000000010000000000000003 in pg_xlog
that contains the WAL up to that point. When recovery switches recovery
target timeline to 2, a walsender can immediately try to read WAL from
0/30002D8, from timeline 2, so it will try to open WAL file
000000020000000000000003. However, that doesn't exist yet - the startup
process hasn't copied that file from the archive yet nor has the walreceiver
streamed it yet, so walsender fails with error "requested WAL segment
000000020000000000000003 has already been removed". That's harmless, in that
the standby will try to reconnect later and by that time the segment is
already created, but error messages that should be ignored are not good.
To fix that, have walsender track the TLI of the last replayed record,
instead of the recovery target timeline. That way walsender will not try to
read anything from timeline 2, until the WAL segment has been created and at
least one record has been replayed from it. The recovery target timeline is
now xlog.c's internal affair, it doesn't need to be exposed in shared memory
anymore.
This fixes the error reported by Thom Brown. depesz the same error message,
but I'm not sure if this fixes his scenario.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
This allows us to do some more rigorous sanity checking for various
incorrect point-in-time recovery scenarios, and provides more information
for debugging purposes. It will also come handy in the upcoming patch to
allow timeline switches to be replicated by streaming replication.
Allow support only for freezing tuples by explicit
command. Previous coding mistakenly extended
slightly beyond what was agreed as correct on -hackers.
So essentially a partial revoke of earlier work,
leaving just the COPY FREEZE command.
When a relfilenode is created in this subtransaction or
a committed child transaction and it cannot otherwise
be seen by our own process, mark tuples committed ahead
of transaction commit for all COPY commands in same
transaction. If FREEZE specified on COPY
and pre-conditions met then rows will also be frozen.
Both options designed to avoid revisiting rows after commit,
increasing performance of subsequent commands after
data load and upgrade. pg_restore changes later.
Simon Riggs, review comments from Heikki Linnakangas, Noah Misch and design
input from Tom Lane, Robert Haas and Kevin Grittner
When I moved ExecuteRecoveryCommand() from xlog.c to xlogarchive.c, I didn't
realize that it's called from the checkpoint process, not the startup
process. I tried to use InRedo variable to decide whether or not to attempt
cleaning up the archive (must not do so before we have read the initial
checkpoint record), but that variable is only valid within the startup
process.
Instead, let ExecuteRecoveryCommand() always clean up the archive, and add
an explicit argument to RestoreArchivedFile() to say whether that's allowed
or not. The caller knows better.
Reported by Erik Rijkers, diagnosis by Fujii Masao. Only 9.3devel is
affected.
Most of the replay functions for WAL record types that modify more than
one page failed to ensure that those pages were locked correctly to ensure
that concurrent queries could not see inconsistent page states. This is
a hangover from coding decisions made long before Hot Standby was added,
when it was hardly necessary to acquire buffer locks during WAL replay
at all, let alone hold them for carefully-chosen periods.
The key problem was that RestoreBkpBlocks was written to hold lock on each
page restored from a full-page image for only as long as it took to update
that page. This was guaranteed to break any WAL replay function in which
there was any update-ordering constraint between pages, because even if the
nominal order of the pages is the right one, any mixture of full-page and
non-full-page updates in the same record would result in out-of-order
updates. Moreover, it wouldn't work for situations where there's a
requirement to maintain lock on one page while updating another. Failure
to honor an update ordering constraint in this way is thought to be the
cause of bug #7648 from Daniel Farina: what seems to have happened there
is that a btree page being split was rewritten from a full-page image
before the new right sibling page was written, and because lock on the
original page was not maintained it was possible for hot standby queries to
try to traverse the page's right-link to the not-yet-existing sibling page.
To fix, get rid of RestoreBkpBlocks as such, and instead create a new
function RestoreBackupBlock that restores just one full-page image at a
time. This function can be invoked by WAL replay functions at the points
where they would otherwise perform non-full-page updates; in this way, the
physical order of page updates remains the same no matter which pages are
replaced by full-page images. We can then further adjust the logic in
individual replay functions if it is necessary to hold buffer locks
for overlapping periods. A side benefit is that we can simplify the
handling of concurrency conflict resolution by moving that code into the
record-type-specfic functions; there's no more need to contort the code
layout to keep conflict resolution in front of the RestoreBkpBlocks call.
In connection with that, standardize on zero-based numbering rather than
one-based numbering for referencing the full-page images. In HEAD, I
removed the macros XLR_BKP_BLOCK_1 through XLR_BKP_BLOCK_4. They are
still there in the header files in previous branches, but are no longer
used by the code.
In addition, fix some other bugs identified in the course of making these
changes:
spgRedoAddNode could fail to update the parent downlink at all, if the
parent tuple is in the same page as either the old or new split tuple and
we're not doing a full-page image: it would get fooled by the LSN having
been advanced already. This would result in permanent index corruption,
not just transient failure of concurrent queries.
Also, ginHeapTupleFastInsert's "merge lists" case failed to mark the old
tail page as a candidate for a full-page image; in the worst case this
could result in torn-page corruption.
heap_xlog_freeze() was inconsistent about using a cleanup lock or plain
exclusive lock: it did the former in the normal path but the latter for a
full-page image. A plain exclusive lock seems sufficient, so change to
that.
Also, remove gistRedoPageDeleteRecord(), which has been dead code since
VACUUM FULL was rewritten.
Back-patch to 9.0, where hot standby was introduced. Note however that 9.0
had a significantly different WAL-logging scheme for GIST index updates,
and it doesn't appear possible to make that scheme safe for concurrent hot
standby queries, because it can leave inconsistent states in the index even
between WAL records. Given the lack of complaints from the field, we won't
work too hard on fixing that branch.
This prevents surprising behavior when a FOR EACH ROW trigger
BEFORE UPDATE or BEFORE DELETE directly or indirectly updates or
deletes the the old row. Prior to this patch the requested action
on the row could be silently ignored while all triggered actions
based on the occurence of the requested action could be committed.
One example of how this could happen is if the BEFORE DELETE
trigger for a "parent" row deleted "children" which had trigger
functions to update summary or status data on the parent.
This also prevents similar surprising problems if the query has a
volatile function which updates a target row while it is already
being updated.
There are related issues present in FOR UPDATE cursors and READ
COMMITTED queries which are not handled by this patch. These
issues need further evalution to determine what change, if any, is
needed.
Where the new error messages are generated, in most cases the best
fix will be to move code from the BEFORE trigger to an AFTER
trigger. Where this is not feasible, the trigger can avoid the
error by re-issuing the triggering statement and returning NULL.
Documentation changes will be submitted in a separate patch.
Kevin Grittner and Tom Lane with input from Florian Pflug and
Robert Haas, based on problems encountered during conversion of
Wisconsin Circuit Court trigger logic to plpgsql triggers.
This is just refactoring, to make the functions accessible outside xlog.c.
A followup patch will make use of that, to allow fetching timeline history
files over streaming replication.
This fixes another error in commit 9e8da0f757.
I neglected to make the mark/restore functionality save and restore the
current set of array key values, which led to strange behavior if an
IndexScan with ScalarArrayOpExpr quals was used as the inner side of a
mergejoin. Per bug #7570 from Melese Tesfaye.
The cascading replication code assumed that the current RecoveryTargetTLI
never changes, but that's not true with recovery_target_timeline='latest'.
The obvious upshot of that is that RecoveryTargetTLI in shared memory needs
to be protected by a lock. A less obvious consequence is that when a
cascading standby is connected, and the standby switches to a new target
timeline after scanning the archive, it will continue to stream WAL to the
cascading standby, but from a wrong file, ie. the file of the previous
timeline. For example, if the standby is currently streaming from the middle
of file 000000010000000000000005, and the timeline changes, the standby
will continue to stream from that file. However, the WAL on the new
timeline is in file 000000020000000000000005, so the standby sends garbage
from 000000010000000000000005 to the cascading standby, instead of the
correct WAL from file 000000020000000000000005.
This also fixes a related bug where a partial WAL segment is restored from
the archive and streamed to a cascading standby. The code assumed that when
a WAL segment is copied from the archive, it can immediately be fully
streamed to a cascading standby. However, if the segment is only partially
filled, ie. has the right size, but only N first bytes contain valid WAL,
that's not safe. That can happen if a partial WAL segment is manually copied
to the archive, or if a partial WAL segment is archived because a server is
started up on a new timeline within that segment. The cascading standby will
get confused if the WAL it received is not valid, and will get stuck until
it's restarted. This patch fixes that problem by not allowing WAL restored
from the archive to be streamed to a cascading standby until it's been
replayed, and thus validated.
This reduces unnecessary exposure of other headers through htup.h, which
is very widely included by many files.
I have chosen to move the function prototypes to the new file as well,
because that means htup.h no longer needs to include tupdesc.h. In
itself this doesn't have much effect in indirect inclusion of tupdesc.h
throughout the tree, because it's also required by execnodes.h; but it's
something to explore in the future, and it seemed best to do the htup.h
change now while I'm busy with it.
The heapam XLog functions are used by other modules, not all of which
are interested in the rest of the heapam API. With this, we let them
get just the XLog stuff in which they are interested and not pollute
them with unrelated includes.
Also, since heapam.h no longer requires xlog.h, many files that do
include heapam.h no longer get xlog.h automatically, including a few
headers. This is useful because heapam.h is getting pulled in by
execnodes.h, which is in turn included by a lot of files.
We use a hash table to track the parents of inner pages, but when inserting
to a leaf page, the caller of gistbufferinginserttuples() must pass a
correct block number of the leaf's parent page. Before gistProcessItup()
descends to a child page, it checks if the downlink needs to be adjusted to
accommodate the new tuple, and updates the downlink if necessary. However,
updating the downlink might require splitting the page, which might move the
downlink to a page to the right. gistProcessItup() doesn't realize that, so
when it descends to the leaf page, it might pass an out-of-date parent block
number as a result. Fix that by returning the block a tuple was inserted to
from gistbufferinginserttuples().
This fixes the bug reported by Zdeněk Jílovec.
The correct test for whether a redirection tuple is removable is whether
tuple's xid < RecentGlobalXmin, not OldestXmin; the previous coding
failed to protect index searches being done in concurrent transactions that
have no XID. This mirrors the recent fix in btree's page recycling logic
made in commit d3abbbebe5.
Also, WAL-log the newest XID of any removed redirection tuple on an index
page, and apply ResolveRecoveryConflictWithSnapshot during InHotStandby WAL
replay. This protects against concurrent Hot Standby transactions possibly
needing to see the redirection tuple(s).
Per my query of 2012-03-12 and subsequent discussion.
Commit 061e7efb1b changed the rules
for splitting xlog records across pages, but neglected to update this
test. It's possible that there's some better action here than just
removing the test completely, but this at least appears to get some
of the things that are currently broken (like initdb on MacOS X)
working again.
The latter was already the dominant use, and it's preferable because
in C the convention is that intXX means XX bits. Therefore, allowing
mixed use of int2, int4, int8, int16, int32 is obviously confusing.
Remove the typedefs for int2 and int4 for now. They don't seem to be
widely used outside of the PostgreSQL source tree, and the few uses
can probably be cleaned up by the time this ships.
This simplifies code that needs to do arithmetic on XLogRecPtrs.
To avoid changing on-disk format of data pages, the LSN on data pages is
still stored in the old format. That should keep pg_upgrade happy. However,
we have XLogRecPtrs embedded in the control file, and in the structs that
are sent over the replication protocol, so this changes breaks compatibility
of pg_basebackup and server. I didn't do anything about this in this patch,
per discussion on -hackers, the right thing to do would to be to change the
replication protocol to be architecture-independent, so that you could use
a newer version of pg_receivexlog, for example, against an older server
version.
This saves a few bytes of WAL space, but the real motivation is to make it
predictable how much WAL space a record requires, as it no longer depends
on whether we need to waste the last few bytes at end of WAL page because
the header doesn't fit.
The total length field of WAL record, xl_tot_len, is moved to the beginning
of the WAL record header, so that it is still always found on the first page
where a WAL record begins.
Bump WAL version number again as this is an incompatible change.
The continuation record only contained one field, xl_rem_len, so it makes
things simpler to just include it in the WAL page header. This wastes four
bytes on pages that don't begin with a continuation from previos page, plus
four bytes on every page, because of padding.
The motivation of this is to make it easier to calculate how much space a
WAL record needs. Before this patch, it depended on how many page boundaries
the record crosses. The motivation of that, in turn, is to separate the
allocation of space in the WAL from the copying of the record data to the
allocated space. Keeping the calculation of space required simple helps to
keep the critical section of allocating the space from WAL short. But that's
not included in this patch yet.
Bump WAL version number again, as this is an incompatible change.
The comments claimed that wasting the last segment made it easier to do
calculations with XLogRecPtrs, because you don't have problems representing
last-byte-position-plus-1 that way. In my experience, however, it only made
things more complicated, because the there was two ways to represent the
boundary at the beginning of a logical log file: logid = n+1 and xrecoff = 0,
or as xlogid = n and xrecoff = 4GB - XLOG_SEG_SIZE. Some functions were
picky about which representation was used.
Also, use a 64-bit segment number instead of the log/seg combination, to
point to a certain WAL segment. We assume that all platforms have a working
64-bit integer type nowadays.
This is an incompatible change in WAL format, so bumping WAL version number.
When I implemented the ginbuildempty() function as part of
implementing unlogged tables, I falsified the note in the header
comment for log_newpage. Although we could fix that up by changing
the comment, it seems cleaner to add a new function which is
specifically intended to handle this case. So do that.
We used to mimic the way a stack is constructed when descending the tree
during normal GiST inserts, but that was quite complicated during a buffered
build. It was also wrong: in GiST, the left-to-right relationships on
different levels might not match each other, so that when you know the
parent of a child page, you won't necessarily find the parent of the page to
the right of the child page by following the rightlinks at the parent level.
This sometimes led to "could not re-find parent" errors while building a
GiST index.
We now use a simple hash table to track the parent of every internal page.
Whenever a page is split, and downlinks are moved from one page to another,
we update the hash table accordingly. This is also better for performance
than the old method, as we never need to move right to re-find the parent
page, which could take a significant amount of time for buffers that were
created much earlier in the index build.
When we create a temporary copy of the old node buffer, in stack, we mustn't
leak that into any of the long-lived data structures. Before this patch,
when we called gistPopItupFromNodeBuffer(), it got added to the array of
"loaded buffers". After gistRelocateBuildBuffersOnSplit() exits, the
pointer added to the loaded buffers array points to garbage. Often that goes
unnotied, because when we go through the array of loaded buffers to unload
them, buffers with a NULL pageBuffer are ignored, which can often happen by
accident even if the pointer points to garbage.
This patch fixes that by marking the temporary copy in stack explicitly as
temporary, and refrain from adding buffers marked as temporary to the array
of loaded buffers.
While we're at it, initialize nodeBuffer->pageBlocknum to InvalidBlockNumber
and improve comments a bit. This isn't strictly necessary, but makes
debugging easier.
Users of asynchronous-commit mode expect there to be a guaranteed maximum
delay before an async commit's WAL records get flushed to disk. The
original version of the walwriter hibernation patch broke that. Add an
extra shared-memory flag to allow async commits to kick the walwriter out
of hibernation mode, without adding any noticeable overhead in cases where
no action is needed.
This patch modifies the walwriter process so that, when it has not found
anything useful to do for many consecutive wakeup cycles, it extends its
sleep time to reduce the server's idle power consumption. It reverts to
normal as soon as it's done any successful flushes. It's still true that
during any async commit, backends check for completed, unflushed pages of
WAL and signal the walwriter if there are any; so that in practice the
walwriter can get awakened and returned to normal operation sooner than the
sleep time might suggest.
Also, improve the checkpointer so that it uses a latch and a computed delay
time to not wake up at all except when it has something to do, replacing a
previous hardcoded 0.5 sec wakeup cycle. This also is primarily useful for
reducing the server's power consumption when idle.
In passing, get rid of the dedicated latch for signaling the walwriter in
favor of using its procLatch, since that comports better with possible
generic signal handlers using that latch. Also, fix a pre-existing bug
with failure to save/restore errno in walwriter's signal handlers.
Peter Geoghegan, somewhat simplified by Tom
The alternative of disallowing index-only scans in HS operation was
discussed, but the consensus was that it was better to treat marking
a page all-visible as a recovery conflict for snapshots that could still
fail to see XIDs on that page. We may in the future try to soften this,
so that we simply force index scans to do heap fetches in cases where
this may be an issue, rather than throwing a hard conflict.
The original syntax wasn't universally loved, and it didn't allow its
usage in CREATE TABLE, only ALTER TABLE. It now works everywhere, and
it also allows using ALTER TABLE ONLY to add an uninherited CHECK
constraint, per discussion.
The pg_constraint column has accordingly been renamed connoinherit.
This commit partly reverts some of the changes in
61d81bd28d, particularly some pg_dump and
psql bits, because now pg_get_constraintdef includes the necessary NO
INHERIT within the constraint definition.
Author: Nikhil Sontakke
Some tweaks by me
It used to be case that lazy vacuum could call this function with only
a shared lock on the buffer, but neither lazy vacuum nor any other
code path does that any more. Simplify the code accordingly and clean
up some related, obsolete comments.
This patch fixes the other major compatibility-breaking limitation of
SPGiST, that it didn't store anything for null values of the indexed
column, and so could not support whole-index scans or "x IS NULL"
tests. The approach is to create a wholly separate search tree for
the null entries, and use fixed "allTheSame" insertion and search
rules when processing this tree, instead of calling the index opclass
methods. This way the opclass methods do not need to worry about
dealing with nulls.
Catversion bump is for pg_am updates as well as the change in on-disk
format of SPGiST indexes; there are some tweaks in SPGiST WAL records
as well.
Heavily rewritten version of a patch by Oleg Bartunov and Teodor Sigaev.
(The original also stored nulls separately, but it reused GIN code to do
so; which required undesirable compromises in the on-disk format, and
would likely lead to bugs due to the GIN code being required to work in
two very different contexts.)
The original API definition was incapable of supporting whole-index scans
because there was no way to invoke leaf-value reconstruction without
checking any qual conditions. Also, it was inefficient for
multiple-qual-condition scans because value reconstruction got done over
again for each qual condition, and because other internal work in the
consistent functions likewise had to be done for each qual. To fix these
issues, pass the whole scankey array to the opclass consistent functions,
instead of only letting them see one item at a time. (Essentially, the
loop over scankey entries is now inside the consistent functions not
outside them. This makes the consistent functions a bit more complicated,
but not unreasonably so.)
In itself this commit does nothing except save a few cycles in
multiple-qual-condition index scans, since we can't support whole-index
scans on SPGiST indexes until nulls are included in the index. However,
I consider this a must-fix for 9.2 because once we release it will get
very much harder to change the opclass API definition.
Comparing two xlog locations are useful for example when calculating
replication lag.
Euler Taveira de Oliveira, reviewed by Fujii Masao, and some cleanups
from me
Base backup follows recommended procedure, plus goes to great
lengths to ensure that partial page writes are avoided.
Jun Ishizuka and Fujii Masao, with minor modifications
Replication occurs only to memory on standby, not to disk,
so provides additional performance if user wishes to
reduce durability level slightly. Adds concept of multiple
independent sync rep queues.
Fujii Masao and Simon Riggs
Previously, this was hardcoded: we always had 8. Performance testing
shows that isn't enough, especially on big SMP systems, so we allow it
to scale up as high as 32 when there's adequate memory. On the flip
side, when shared_buffers is very small, drop the number of CLOG buffers
down to as little as 4, so that we can start the postmaster even
when very little shared memory is available.
Per extensive discussion with Simon Riggs, Tom Lane, and others on
pgsql-hackers.
When a view is marked as a security barrier, it will not be pulled up
into the containing query, and no quals will be pushed down into it,
so that no function or operator chosen by the user can be applied to
rows not exposed by the view. Views not configured with this
option cannot provide robust row-level security, but will perform far
better.
Patch by KaiGai Kohei; original problem report by Heikki Linnakangas
(in October 2009!). Review (in earlier versions) by Noah Misch and
others. Design advice by Tom Lane and myself. Further review and
cleanup by me.
On reflection, the original name seems way too generic for a global
symbol. A quick check shows this is the only exported function name
in SP-GiST that doesn't begin with "spg" or contain "SpGist", so the
rest of them seem all right.
This makes them enforceable only on the parent table, not on children
tables. This is useful in various situations, per discussion involving
people bitten by the restrictive behavior introduced in 8.4.
Message-Id:
8762mp93iw.fsf@comcast.netCAFaPBrSMMpubkGf4zcRL_YL-AERUbYF_-ZNNYfb3CVwwEqc9TQ@mail.gmail.com
Authors: Nikhil Sontakke, Alex Hunsaker
Reviewed by Robert Haas and myself
Operator classes can specify whether or not they support this; this
preserves the flexibility to use lossy representations within an index.
In passing, move constant data about a given index into the rd_amcache
cache area, instead of doing fresh lookups each time we start an index
operation. This is mainly to try to make sure that spgcanreturn() has
insignificant cost; I still don't have any proof that it matters for
actual index accesses. Also, get rid of useless copying of FmgrInfo
pointers; we can perfectly well use the relcache's versions in-place.
The need for this was debated when we put in the index-only-scan feature,
but at the time we had no near-term expectation of having AMs that could
support such scans for only some indexes; so we kept it simple. However,
the SP-GiST AM forces the issue, so let's fix it.
This patch only installs the new API; no behavior actually changes.
SP-GiST is comparable to GiST in flexibility, but supports non-balanced
partitioned search structures rather than balanced trees. As described at
PGCon 2011, this new indexing structure can beat GiST in both index build
time and query speed for search problems that it is well matched to.
There are a number of areas that could still use improvement, but at this
point the code seems committable.
Teodor Sigaev and Oleg Bartunov, with considerable revisions by Tom Lane
Heikki Linnakangas had the idea of rearranging GetSnapshotData to
avoid checking for sub-XIDs when no top-level XID is present. This
patch does that plus further a bit of further, related rearrangement.
Benchmarking show a significant improvement on unlogged tables at
higher concurrency levels, and mostly indifferent result on permanent
tables (which are presumably bottlenecked elsewhere). Most of the
benefit seems to come from using the new NormalTransactionIdPrecedes()
macro rather than the function call TransactionIdPrecedes().
Removing this bit from xl_info allows us to restore the old limit of four
(not three) separate pages touched by a WAL record, which is needed for the
upcoming SP-GiST feature, and will likely be useful elsewhere in future.
When we implemented XLR_BKP_REMOVABLE in 2007, we had to do it like that
because no special WAL-visible action was taken when starting a backup.
However, now we force a segment switch when starting a backup, so a
compressing WAL archiver (such as pglesslog) that uses the state shown in
the current page header will not be fooled as to removability of backup
blocks. The only downside is that the archiver will not return to
compressing mode for up to one WAL page after the backup is over, which is
a small price to pay for getting back the extra xl_info bit. In any case
the archiver could look for XLOG_BACKUP_END records if it thought it was
worth the trouble to do so.
Bump XLOG_PAGE_MAGIC since this is effectively a change in WAL format.
we don't reach consistency before replaying all of the WAL. Rename the
variable to reachedConsistency, to make its intention clearer.
In master, that was an active bug because of the recent patch to
immediately PANIC if a reference to a missing page is found in WAL after
reaching consistency, as Tom Lane's test case demonstrated. In 9.1 and 9.0,
the only consequence was a misleading "consistent recovery state reached at
%X/%X" message in the log at the beginning of crash recovery (the database
is not consistent at that point yet). In 8.4, the log message was not
printed in crash recovery, even though there was a similar
reachedMinRecoveryPoint local variable that was also set early. So,
backpatch to 9.1 and 9.0.
This patch creates an API whereby a btree index opclass can optionally
provide non-SQL-callable support functions for sorting. In the initial
patch, we only use this to provide a directly-callable comparator function,
which can be invoked with a bit less overhead than the traditional
SQL-callable comparator. While that should be of value in itself, the real
reason for doing this is to provide a datatype-extensible framework for
more aggressive optimizations, as in Peter Geoghegan's recent work.
Robert Haas and Tom Lane
invalid-page hash table, PANIC immediately. Immediate PANIC is much better
than waiting for end-of-recovery, which is what we did before, because the
end-of-recovery might not come until months later if this is a standby
server.
Also refrain from creating a restartpoint if there are invalid-page entries
in the hash table. Restarting recovery from such a restartpoint would not
see the invalid references, and wouldn't be able to cross-check them when
consistency is reached. That wouldn't matter when things are going smoothly,
but the more sanity checks you have the better.
Fujii Masao
Previously we waited for wal_writer_delay before flushing WAL. Now
we also wake WALWriter as soon as a WAL buffer page has filled.
Significant effect observed on performance of asynchronous commits
by Robert Haas, attributed to the ability to set hint bits on tuples
earlier and so reducing contention caused by clog lookups.
This greatly reduces the WAL volume, especially when the table is narrow.
The overhead of locking the heap page is also reduced. Reduced WAL traffic
also makes it scale a lot better, if you run multiple COPY processes at
the same time.
In a regular VACUUM, it's OK to skip pages for which a cleanup lock
isn't immediately available; the next VACUUM will deal with them. If
we're scanning the entire relation to advance relfrozenxid, we might
need to wait, but only if there are tuples on the page that actually
require freezing. These changes should greatly reduce the incidence
of of vacuum processes getting "stuck".
Simon Riggs and Robert Haas
If a tuple in a syscache contains an out-of-line toasted field, and we
try to fetch that field shortly after some other transaction has committed
an update or deletion of the tuple, there is a race condition: vacuum
could come along and remove the toast tuples before we can fetch them.
This leads to transient failures like "missing chunk number 0 for toast
value NNNNN in pg_toast_2619", as seen in recent reports from Andrew
Hammond and Tim Uckun.
The design idea of syscache is that access to stale syscache entries
should be prevented by relation-level locks, but that fails for at least
two cases where toasted fields are possible: ANALYZE updates pg_statistic
rows without locking out sessions that might want to plan queries on the
same table, and CREATE OR REPLACE FUNCTION updates pg_proc rows without
any meaningful lock at all.
The least risky fix seems to be an idea that Heikki suggested when we
were dealing with a related problem back in August: forcibly detoast any
out-of-line fields before putting a tuple into syscache in the first place.
This avoids the problem because at the time we fetch the parent tuple from
the catalog, we should be holding an MVCC snapshot that will prevent
removal of the toast tuples, even if the parent tuple is outdated
immediately after we fetch it. (Note: I'm not convinced that this
statement holds true at every instant where we could be fetching a syscache
entry at all, but it does appear to hold true at the times where we could
fetch an entry that could have a toasted field. We will need to be a bit
wary of adding toast tables to low-level catalogs that don't have them
already.) An additional benefit is that subsequent uses of the syscache
entry should be faster, since they won't have to detoast the field.
Back-patch to all supported versions. The problem is significantly harder
to reproduce in pre-9.0 releases, because of their willingness to flush
every entry in a syscache whenever the underlying catalog is vacuumed
(cf CatalogCacheFlushRelation); but there is still a window for trouble.
bgwriter is now a much less important process, responsible for page
cleaning duties only. checkpointer is now responsible for checkpoints
and so has a key role in shutdown. Later patches will correct doc
references to the now old idea that bgwriter performs checkpoints.
Has beneficial effect on performance at high write rates, but mainly
refactoring to more easily allow changes for power reduction by
simplifying previously tortuous code around required to allow page
cleaning and checkpointing to time slice in the same process.
Patch by me, Review by Dickson Guedes
There's no particular value in doing AssertMacro((tup) != NULL) in front
of code that's certain to crash anyway if tup is NULL. And if "tup" is
actually the address of a local variable, gcc 4.6 whinges about it. That's
arguably pretty broken on gcc's part, but we might as well remove the
useless test to silence the warnings. This gets rid of all the -Waddress
warnings in the backend; there are some in libpq and psql that are a bit
harder to avoid.
In general the data returned by an index-only scan should have the
datatypes originally computed by FormIndexDatum. If the index opclasses
use "storage" datatypes different from their input datatypes, the scan
tuple will not have the same rowtype attributed to the index; but we had
a hard-wired assumption that that was true in nodeIndexonlyscan.c. We'd
already hacked around the issue for the one case where the types are
different in btree indexes (btree name_ops), but this would definitely
come back to bite us if we ever implement index-only scans in GiST.
To fix, require the index AM to explicitly provide the tupdesc for the
tuple it is returning. btree can just pass back the index's tupdesc, but
GiST will have to work harder when and if it supports index-only scans.
I had previously proposed fixing this by allowing the index AM to fill the
scan tuple slot directly; but on reflection that seemed like a module
layering violation, since TupleTableSlots are creatures of the executor.
At least in the btree case, it would also be less efficient, since the
tuple deconstruction work would occur even for rows later found to be
invisible to the scan's snapshot.
Add a column pg_class.relallvisible to remember the number of pages that
were all-visible according to the visibility map as of the last VACUUM
(or ANALYZE, or some other operations that update pg_class.relpages).
Use relallvisible/relpages, instead of an arbitrary constant, to estimate
how many heap page fetches can be avoided during an index-only scan.
This is pretty primitive and will no doubt see refinements once we've
acquired more field experience with the index-only scan mechanism, but
it's way better than using a constant.
Note: I had to adjust an underspecified query in the window.sql regression
test, because it was changing answers when the plan changed to use an
index-only scan. Some of the adjacent tests perhaps should be adjusted
as well, but I didn't do that here.
We copy all the matched tuples off the page during _bt_readpage, instead of
expensively re-locking the page during each subsequent tuple fetch. This
costs a bit more local storage, but not more than 2*BLCKSZ worth, and the
reduction in LWLock traffic is certainly worth that. What's more, this
lets us get rid of the API wart in the original patch that said an index AM
could randomly decline to supply an index tuple despite having asserted
pg_am.amcanreturn. That will be important for future improvements in the
index-only-scan feature, since the executor will now be able to rely on
having the index data available.
When a btree index contains all columns required by the query, and the
visibility map shows that all tuples on a target heap page are
visible-to-all, we don't need to fetch that heap page. This patch depends
on the previous patches that made the visibility map reliable.
There's a fair amount left to do here, notably trying to figure out a less
chintzy way of estimating the cost of an index-only scan, but the core
functionality seems ready to commit.
Robert Haas and Ibrar Ahmed, with some previous work by Heikki Linnakangas.
Previously, the code assumed that the only possible action to take was
to delete files behind a certain cutoff point. The async notify code
was already a crock: it used a different "pagePrecedes" function for
truncation than for regular operation. By allowing it to pass a
callback to SlruScanDirectory it can do cleanly exactly what it needs to
do.
The clog.c code also had its own use for SlruScanDirectory, which is
made a bit simpler with this.
pg_trgm was already doing this unofficially, but the implementation hadn't
been thought through very well and leaked memory. Restructure the core
GiST code so that it actually works, and document it. Ordinarily this
would have required an extra memory context creation/destruction for each
GiST index search, but I was able to avoid that in the normal case of a
non-rescanned search by finessing the handling of the RBTree. It used to
have its own context always, but now shares a context with the
scan-lifespan data structures, unless there is more than one rescan call.
This should make the added overhead unnoticeable in typical cases.
As per my recent proposal, this refactors things so that these typedefs and
macros are available in a header that can be included in frontend-ish code.
I also changed various headers that were undesirably including
utils/timestamp.h to include datatype/timestamp.h instead. Unsurprisingly,
this showed that half the system was getting utils/timestamp.h by way of
xlog.h.
No actual code changes here, just header refactoring.
When building a GiST index that doesn't fit in cache, buffers are attached
to some internal nodes in the index. This speeds up the build by avoiding
random I/O that would otherwise be needed to traverse all the way down the
tree to the find right leaf page for tuple.
Alexander Korotkov
walsender.h should depend on xlog.h, not vice versa. (Actually, the
inclusion was circular until a couple hours ago, which was even sillier;
but Bruce broke it in the expedient rather than logically correct
direction.) Because of that poor decision, plus blind application of
pgrminclude, we had a situation where half the system was depending on
xlog.h to include such unrelated stuff as array.h and guc.h. Clean up
the header inclusion, and manually revert a lot of what pgrminclude had
done so things build again.
This episode reinforces my feeling that pgrminclude should not be run
without adult supervision. Inclusion changes in header files in particular
need to be reviewed with great care. More generally, it'd be good if we
had a clearer notion of module layering to dictate which headers can sanely
include which others ... but that's a big task for another day.
Don't try to allocate the default value for a string relopt in the same
palloc chunk as the relopt_string struct. That didn't work too well if you
added a built-in string relopt in the stringRelOpts array, as it's not
possible to have an initializer for a variable length struct in C. This
makes the code slightly simpler too.
While we're at it, move the call to validator function in
add_string_reloption to before the allocation, so that if someone does pass
a bogus default value, we don't leak memory.
Standby servers can now have WALSender processes, which can work with
either WALReceiver or archive_commands to pass data. Fully updated
docs, including new conceptual terms of sending server, upstream and
downstream servers. WALSenders terminated when promote to master.
Fujii Masao, review, rework and doc rewrite by Simon Riggs
GISTInsertStack.childoffnum used to mean "offset of the downlink in this
node, pointing to the child node in the stack". It's now replaced with
downlinkoffnum, which means "offset of the downlink in the parent of this
node". gistFindPath() already used childoffnum with this new meaning, and
had an extra step at the end to pull all the childoffnum values down one
node in the stack, to adjust the stack for the meaning that childoffnum had
elsewhere. That's no longer required.
The reason to do this now is this new representation is more convenient for
the GiST fast build patch that Alexander Korotkov is working on.
While we're at it, replace the linked list used in gistFindPath with a
standard List, and make gistFindPath() static.
Alexander Korotkov, with some changes by me.
This means that they can initially be added to a large existing table
without checking its initial contents, but new tuples must comply to
them; a separate pass invoked by ALTER TABLE / VALIDATE can verify
existing data and ensure it complies with the constraint, at which point
it is marked validated and becomes a normal part of the table ecosystem.
An non-validated CHECK constraint is ignored in the planner for
constraint_exclusion purposes; when validated, cached plans are
recomputed so that partitioning starts working right away.
This patch also enables domains to have unvalidated CHECK constraints
attached to them as well by way of ALTER DOMAIN / ADD CONSTRAINT / NOT
VALID, which can later be validated with ALTER DOMAIN / VALIDATE
CONSTRAINT.
Thanks to Thom Brown, Dean Rasheed and Jaime Casanova for the various
reviews, and Robert Hass for documentation wording improvement
suggestions.
This patch was sponsored by Enova Financial.
more consistent that way, since all the other PredicateLock* calls are
made in various heapam.c and index AM functions. The call in nodeSeqscan.c
was unnecessarily aggressive anyway, there's no need to try to lock the
relation every time a tuple is fetched, it's enough to do it once.
This has the user-visible effect that if a seq scan is initialized in the
executor, but never executed, we now acquire the predicate lock on the heap
relation anyway. We could avoid that by taking the lock on the first
heap_getnext() call instead, but it doesn't seem worth the trouble given
that it feels more natural to do it in heap_beginscan().
Also, remove the retail PredicateLockTuple() calls from heap_getnext(). In
a seqscan, started with heap_begin(), we're holding a whole-relation
predicate lock on the heap so there's no need to lock the tuples
individually.
Kevin Grittner and me
XLOG_XACT_COMMIT_COMPACT leaves out invalidation messages and relfilenodes,
saving considerable space for the vast majority of transaction commits.
XLOG_XACT_COMMIT keeps same definition as XLOG_PAGE_MAGIC 0xD067 and earlier.
Leonardo Francalanci and Simon Riggs
Since the names try_relation_openrv() and try_heap_openrv() don't seem
quite appropriate, rename the functions to relation_openrv_extended()
and heap_openrv_extended(). This is also more general, if we have a
future need for additional parameters that are of interest to only a
few callers.
This is infrastructure for a forthcoming patch to allow
get_object_address() to take a missing_ok argument as well.
Patch by me, review by Noah Misch.
It's been like this since HOT was originally introduced, but the logic
is complex enough that this is a recipe for bugs, as we've already
found out with SSI. So refactor heap_hot_search_buffer() so that it
can satisfy the needs of index_getnext(), and make index_getnext() use
that rather than duplicating the logic.
This change was originally proposed by Heikki Linnakangas as part of a
larger refactoring oriented towards allowing index-only scans. I
extracted and adjusted this part, since it seems to have independent
merit. Review by Jeff Davis.
This involves two main changes from the previous behavior. First,
when we set a bit in the visibility map, emit a new WAL record of type
XLOG_HEAP2_VISIBLE. Replay sets the page-level PD_ALL_VISIBLE bit and
the visibility map bit. Second, when inserting, updating, or deleting
a tuple, we can no longer get away with clearing the visibility map
bit after releasing the lock on the corresponding heap page, because
an intervening crash might leave the visibility map bit set and the
page-level bit clear. Making this work requires a bit of interface
refactoring.
In passing, a few minor but related cleanups: change the test in
visibilitymap_set and visibilitymap_clear to throw an error if the
wrong page (or no page) is pinned, rather than silently doing nothing;
this case should never occur. Also, remove duplicate definitions of
InvalidXLogRecPtr.
Patch by me, review by Noah Misch.
Flexible array members are a C99 feature that avoids "cheating" in the
declaration of variable-length arrays at the end of structs. With
Autoconf support, this should be transparent for older compilers.
We start with one use in gist.h because gcc 4.6 started to raise a
warning there. Over time, it can be expanded to other places in the
source, but they will likely need some review of sizeof and offsetof
usage. The current change in gist.h appears to be safe in this
regard.
is added to the end, and existing resource managers keep their old ids.
We're not going to guarantee on-disk compatibility for 2PC state files over
major releases, but it seems better to avoid changing the ids them anyway.
It will help anyone who might want to write external tools to inspect the
state files to work with files from different versions, if nothing else.
Per complaint from Tom Lane.
ReadRecord's habit of using both direct references to tmpRecPtr and
references to *RecPtr (which is pointing at tmpRecPtr) triggers an
optimization bug in gcc 4.6.0, which apparently has forgotten about
aliasing rules. Avoid the compiler bug, and make the code more readable
to boot, by getting rid of the direct references. Improve the comments
while at it.
Back-patch to all supported versions, in case they get built with 4.6.0.
Tom Lane, with some cosmetic suggestions from Alex Hunsaker
Experimentation with contrib/btree_gist shows that the majority of the GIST
support functions potentially need collation information. Safest policy
seems to be to pass it to all of them, instead of making assumptions about
which ones could possibly need it.
Since collation is effectively an argument, not a property of the function,
FmgrInfo is really the wrong place for it; and this becomes critical in
cases where a cached FmgrInfo is used for varying purposes that might need
different collation settings. Fix by passing it in FunctionCallInfoData
instead. In particular this allows a clean fix for bug #5970 (record_cmp
not working). This requires touching a bit more code than the original
method, but nobody ever thought that collations would not be an invasive
patch...
Also avoid hardcoding the current default state by giving it the name
"on" and replace with a meaningful name that reflects its behaviour.
Coding only, no change in behaviour.
This means one less thing to configure when setting up synchronous
replication, and also avoids some ambiguity around what the behavior
should be when the settings of these variables conflict.
Fujii Masao, with additional hacking by me.
Standby optionally sends back information about oldestXmin of queries
which is then checked and applied to the WALSender's proc->xmin.
GetOldestXmin() is modified slightly to agree with GetSnapshotData(),
so that all backends on primary include WALSender within their snapshots.
Note this does nothing to change the snapshot xmin on either master or
standby. Feedback piggybacks on the standby reply message.
vacuum_defer_cleanup_age is no longer used on standby, though parameter
still exists on primary, since some use cases still exist.
Simon Riggs, review comments from Fujii Masao, Heikki Linnakangas, Robert Haas
the standby has written, flushed, and applied the WAL. At the moment, this
is for informational purposes only, the values are only shown in
pg_stat_replication system view, but in the future they will also be needed
for synchronous replication.
Extracted from Simon riggs' synchronous replication patch by Robert Haas, with
some tweaking by me.
Specifying this option makes the server not wait for the
xlog to be archived, or emit a warning that it can't,
instead leaving the responsibility with the client.
This is useful when the log is being streamed using
the streaming protocol in parallel with the backup,
without having log archiving enabled.
This adds collation support for columns and domains, a COLLATE clause
to override it per expression, and B-tree index support.
Peter Eisentraut
reviewed by Pavel Stehule, Itagaki Takahiro, Robert Haas, Noah Misch
new recovery.conf parameter recovery_target_name allows PITR to
specify named points as recovery targets.
Jaime Casanova, reviewed by Euler Taveira de Oliveira, plus minor edits
Until now, our Serializable mode has in fact been what's called Snapshot
Isolation, which allows some anomalies that could not occur in any
serialized ordering of the transactions. This patch fixes that using a
method called Serializable Snapshot Isolation, based on research papers by
Michael J. Cahill (see README-SSI for full references). In Serializable
Snapshot Isolation, transactions run like they do in Snapshot Isolation,
but a predicate lock manager observes the reads and writes performed and
aborts transactions if it detects that an anomaly might occur. This method
produces some false positives, ie. it sometimes aborts transactions even
though there is no anomaly.
To track reads we implement predicate locking, see storage/lmgr/predicate.c.
Whenever a tuple is read, a predicate lock is acquired on the tuple. Shared
memory is finite, so when a transaction takes many tuple-level locks on a
page, the locks are promoted to a single page-level lock, and further to a
single relation level lock if necessary. To lock key values with no matching
tuple, a sequential scan always takes a relation-level lock, and an index
scan acquires a page-level lock that covers the search key, whether or not
there are any matching keys at the moment.
A predicate lock doesn't conflict with any regular locks or with another
predicate locks in the normal sense. They're only used by the predicate lock
manager to detect the danger of anomalies. Only serializable transactions
participate in predicate locking, so there should be no extra overhead for
for other transactions.
Predicate locks can't be released at commit, but must be remembered until
all the transactions that overlapped with it have completed. That means that
we need to remember an unbounded amount of predicate locks, so we apply a
lossy but conservative method of tracking locks for committed transactions.
If we run short of shared memory, we overflow to a new "pg_serial" SLRU
pool.
We don't currently allow Serializable transactions in Hot Standby mode.
That would be hard, because even read-only transactions can cause anomalies
that wouldn't otherwise occur.
Serializable isolation mode now means the new fully serializable level.
Repeatable Read gives you the old Snapshot Isolation level that we have
always had.
Kevin Grittner and Dan Ports, reviewed by Jeff Davis, Heikki Linnakangas and
Anssi Kääriäinen
With this patch, pg_basebackup doesn't write a backup_label file in the
data directory, so it doesn't interfere with a pg_start/stop_backup() based
backup anymore. backup_label is still included in the backup, but it is
injected directly into the tar stream.
Heikki Linnakangas, reviewed by Fujii Masao and Magnus Hagander.
Move the actual functionality into a separate function that's
easier to call internally, and change the SQL-callable function
to be a wrapper calling this.
Also create a pg_abort_backup() function, only callable internally,
that does only the most vital parts of pg_stop_backup(), making it
safe(r) to call from error handlers.
The original coding could combine duplicate entries only when they
originated from the same qual condition. In particular it could not
combine cases where multiple qual conditions all give rise to full-index
scan requests, which is an expensive case well worth optimizing. Refactor
so that duplicates are recognized across all the quals.
Per my recent proposal(s). Null key datums can now be returned by
extractValue and extractQuery functions, and will be stored in the index.
Also, placeholder entries are made for indexable items that are NULL or
contain no keys according to extractValue. This means that the index is
now always complete, having at least one entry for every indexed heap TID,
and so we can get rid of the prohibition on full-index scans. A full-index
scan is implemented much the same way as partial-match scans were already:
we build a bitmap representing all the TIDs found in the index, and then
drive the results off that.
Also, introduce a concept of a "search mode" that can be requested by
extractQuery when the operator requires matching to empty items (this is
just as cheap as matching to a single key) or requires a full index scan
(which is not so cheap, but it sure beats failing or giving wrong answers).
The behavior remains backward compatible for opclasses that don't return
any null keys or request a non-default search mode.
Using these features, we can now make the GIN index opclass for anyarray
behave in a way that matches the actual anyarray operators for &&, <@, @>,
and = ... which it failed to do before in assorted corner cases.
This commit fixes the core GIN code and ginarrayprocs.c, updates the
documentation, and adds some simple regression test cases for the new
behaviors using the array operators. The tsearch and contrib GIN opclass
support functions still need to be looked over and probably fixed.
Another thing I intend to fix separately is that this is pretty inefficient
for cases where more than one scan condition needs a full-index search:
we'll run duplicate GinScanEntrys, each one of which builds a large bitmap.
There is some existing logic to merge duplicate GinScanEntrys but it needs
refactoring to make it work for entries belonging to different scan keys.
Note that most of gin.h has been split out into a new file gin_private.h,
so that gin.h doesn't export anything that's not supposed to be used by GIN
opclasses or the rest of the backend. I did quite a bit of other code
beautification work as well, mostly fixing comments and choosing more
appropriate names for things.
This is advantageous first because it allows us to hash the smaller table
regardless of the outer-join type, and second because hash join can be more
flexible than merge join in dealing with arbitrary join quals in a FULL
join. For merge join all the join quals have to be mergejoinable, but hash
join will work so long as there's at least one hashjoinable qual --- the
others can be any condition. (This is true essentially because we don't
keep per-inner-tuple match flags in merge join, while hash join can do so.)
To do this, we need a has-it-been-matched flag for each tuple in the
hashtable, not just one for the current outer tuple. The key idea that
makes this practical is that we can store the match flag in the tuple's
infomask, since there are lots of bits there that are of no interest for a
MinimalTuple. So we aren't increasing the size of the hashtable at all for
the feature.
To write this without turning the hash code into even more of a pile of
spaghetti than it already was, I rewrote ExecHashJoin in a state-machine
style, similar to ExecMergeJoin. Other than that decision, it was pretty
straightforward.
Instead, declare a public wrapper of the sole function using it for
external callers, so that they don't have to always pass a NULL
argument.
Author: Kevin Grittner
The contents of an unlogged table are WAL-logged; thus, they are not
available on standby servers and are truncated whenever the database
system enters recovery. Indexes on unlogged tables are also unlogged.
Unlogged GiST indexes are not currently supported.
cleanup stage to finish incomplete inserts or splits anymore. There was two
reasons for the cleanup step:
1. When a new tuple was inserted to a leaf page, the downlink in the parent
needed to be updated to contain (ie. to be consistent with) the new key.
Updating the parent in turn might require recursively updating the parent of
the parent. We now handle that by updating the parent while traversing down
the tree, so that when we insert the leaf tuple, all the parents are already
consistent with the new key, and the tree is consistent at every step.
2. When a page is split, we need to insert the downlink for the new right
page(s), and update the downlink for the original page to not include keys
that moved to the right page(s). We now handle that by setting a new flag,
F_FOLLOW_RIGHT, on the non-rightmost pages in the split. When that flag is
set, scans always follow the rightlink, regardless of the NSN mechanism used
to detect concurrent page splits. That way the tree is consistent right after
split, even though the downlink is still missing. This is very similar to the
way B-tree splits are handled. When the downlink is inserted in the parent,
the flag is cleared. To keep the insertion algorithm simple, when an
insertion sees an incomplete split, indicated by the F_FOLLOW_RIGHT flag, it
finishes the split before doing anything else.
These changes allow removing the whole "invalid tuple" mechanism, but I
retained the scan code to still follow invalid tuples correctly. While we
don't create any such tuples anymore, we want to handle them gracefully in
case you pg_upgrade a GiST index that has them. If we encounter any on an
insert, though, we just throw an error saying that you need to REINDEX.
The issue that got me into doing this is that if you did a checkpoint while
an insert or split was in progress, and the checkpoint finishes quickly so
that there is no WAL record related to the insert between RedoRecPtr and the
checkpoint record, recovery from that checkpoint would not know to finish
the incomplete insert. IOW, we have the same issue we solved with the
rm_safe_restartpoint mechanism during normal operation too. It's highly
unlikely to happen in practice, and this fix is far too large to backpatch,
so we're just going to live with in previous versions, but this refactoring
fixes it going forward.
With this patch, you don't get the annoying
'index "FOO" needs VACUUM or REINDEX to finish crash recovery' notices
anymore if you crash at an unfortunate moment.
Recent versions of the Linux system header files cause xlogdefs.h to
believe that open_datasync should be the default sync method, whereas
formerly fdatasync was the default on Linux. open_datasync is a bad
choice, first because it doesn't actually outperform fdatasync (in fact
the reverse), and second because we try to use O_DIRECT with it, causing
failures on certain filesystems (e.g., ext4 with data=journal option).
This part of the patch is largely per a proposal from Marti Raudsepp.
More extensive changes are likely to follow in HEAD, but this is as much
change as we want to back-patch.
Also clean up confusing code and incorrect documentation surrounding the
fsync_writethrough option. Those changes shouldn't result in any actual
behavioral change, but I chose to back-patch them anyway to keep the
branches looking similar in this area.
In 9.0 and HEAD, also do some copy-editing on the WAL Reliability
documentation section.
Back-patch to all supported branches, since any of them might get used
on modern Linux versions.
This commit represents a rather heavily editorialized version of
Teodor's builtin_knngist_itself-0.8.2 and builtin_knngist_proc-0.8.1
patches. I redid the opclass API to add a separate Distance method
instead of turning the Consistent method into an illogical mess,
fixed some bit-rot in the rbtree interfaces, and generally worked over
the code style and comments.
There's still no non-code documentation to speak of, but I'll work on
that separately. Some contrib-module changes are also yet to come
(right now, point <-> point is the only KNN-ified operator).
Teodor Sigaev and Tom Lane
This is a heavily revised version of builtin_knngist_core-0.9. The
ordering operators are no longer mixed in with actual quals, which would
have confused not only humans but significant parts of the planner.
Instead, ordering operators are carried separately throughout planning and
execution.
Since the API for ambeginscan and amrescan functions had to be changed
anyway, this commit takes the opportunity to rationalize that a bit.
RelationGetIndexScan no longer forces a premature index_rescan call;
instead, callers of index_beginscan must call index_rescan too. Aside from
making the AM-side initialization logic a bit less peculiar, this has the
advantage that we do not make a useless extra am_rescan call when there are
runtime key values. AMs formerly could not assume that the key values
passed to amrescan were actually valid; now they can.
Teodor Sigaev and Tom Lane
temporary indexes are not WAL-logged. We used a constant LSN for temporary
indexes, on the assumption that we don't need to worry about concurrent page
splits in temporary indexes because they're only visible to the current
session. But that assumption is wrong, it's possible to insert rows and
split pages in the same session, while a scan is in progress. For example,
by opening a cursor and fetching some rows, and INSERTing new rows before
fetching some more.
Fix by generating fake increasing LSNs, used in place of real LSNs in
temporary GiST indexes.
The GIN code has absolutely no business exporting GIN-specific functions
with names as generic as compareItemPointers() or newScanKey(); that's
just trouble waiting to happen. I got annoyed about this again just now
and decided to fix it. This commit ensures that all global symbols
defined in access/gin/ have names including "gin" or "Gin". There were a
couple of cases, like names involving "PostingItem", where arguably the
names were already sufficiently nongeneric; but I figured as long as I was
risking creating merge problems for unapplied GIN patches I might as well
impose a uniform policy.
I didn't touch any static symbol names. There might be some places
where it'd be appropriate to rename some static functions to match
siblings that are exported, but I'll leave that for another time.
The better estimate requires more statistics than we previously stored:
in particular, counts of "entry" versus "data" pages within the index,
as well as knowledge of the number of distinct key values. We collect
this information during initial index build and update it during VACUUM,
storing the info in new fields on the index metapage. No initdb is
required because these fields will read as zeroes in a pre-existing
index, and the new gincostestimate code is coded to behave (reasonably)
sanely if they are zeroes.
Teodor Sigaev, reviewed by Jan Urbanski, Tom Lane, and Itagaki Takahiro.
This patch resurrects some of the information that could be logged by the
old, now-dead implementation of VACUUM FULL, in particular counts of live
and dead tuples and the time taken for the table rebuild proper. There's
still no logging about the ensuing index rebuilds, though.
Itagaki Takahiro
new WAL arrives via streaming replication. This reduces the latency, and
also allows us to use a longer polling interval, which is good for energy
efficiency.
We still need to poll to check for the appearance of a trigger file, but
the interval is now 5 seconds (instead of 100ms), like when waiting for
a new WAL segment to appear in WAL archive.
transaction snapshots, i.e. a snapshot registered at the beginning of
a transaction. Change variable naming and comments to reflect this reality
in preparation for a future, truly serializable mode, e.g.
Serializable Snapshot Isolation (SSI).
For the moment transaction snapshots are still used to implement
SERIALIZABLE, but hopefully not for too much longer. Patch by Kevin
Grittner and Dan Ports with review and some minor wording changes by me.
This allows us to reliably remove all leftover temporary relation
files on cluster startup without reference to system catalogs or WAL;
therefore, we no longer include temporary relations in XLOG_XACT_COMMIT
and XLOG_XACT_ABORT WAL records.
Since these changes require including a backend ID in each
SharedInvalSmgrMsg, the size of the SharedInvalidationMessage.id
field has been reduced from two bytes to one, and the maximum number
of connections has been reduced from INT_MAX / 4 to 2^23-1. It would
be possible to remove these restrictions by increasing the size of
SharedInvalidationMessage by 4 bytes, but right now that doesn't seem
like a good trade-off.
Review by Jaime Casanova and Tom Lane.
Although the key-combining code claimed to work correctly if its input
contained both lossy and exact pointers for a single page in a single TID
stream, in fact this did not work, and could not work without pretty
fundamental redesign. Modify keyGetItem so that it will not return such a
stream, by handling lossy-pointer cases a bit more explicitly than we did
before.
Per followup investigation of a gripe from Artur Dabrowski.
An example of a query that failed given his data set is
select count(*) from search_tab where
(to_tsvector('german', keywords ) @@ to_tsquery('german', 'ee:* | dd:*')) and
(to_tsvector('german', keywords ) @@ to_tsquery('german', 'aa:*'));
Back-patch to 8.4 where the lossy pointer code was introduced.
struct representing a tree entry, rather than being a separately allocated
piece of storage. This API is at least as clean as the old one (if not
more so --- there were some bizarre choices in there) and it permits a
very substantial memory savings, on the order of 2X in ginbulk.c's usage.
Also, fix minor memory leaks in code called by ginEntryInsert, in
particular in ginInsertValue and entryFillRoot, as well as ginEntryInsert
itself. These leaks resulted in the GIN index build context continuing
to bloat even after we'd filled it to maintenance_work_mem and started
to dump data out to the index.
In combination these fixes restore the GIN index build code to honoring
the maintenance_work_mem limit about as well as it did in 8.4. Speed
seems on par with 8.4 too, maybe even a bit faster, for a non-pathological
case in which HEAD was formerly slower.
Back-patch to 9.0 so we don't have a performance regression from 8.4.
routines to make them behave better in the presence of "lossy" index pointers.
The previous coding was outright incorrect for some cases, as recently
reported by Artur Dabrowski: scanGetItem would fail to return index entries in
cases where one index key had multiple exact pointers on the same page as
another key had a lossy pointer. Also, keyGetItem was extremely inefficient
for cases where a single index key generates multiple "entry" streams, such as
an @@ operator with a multiple-clause tsquery. The presence of a lossy page
pointer in any one stream defeated its ability to use the opclass
consistentFn, resulting in probing many heap pages that didn't really need to
be visited. In Artur's example case, a query like
WHERE tsvector @@ to_tsquery('a & b')
was about 50X slower than the theoretically equivalent
WHERE tsvector @@ to_tsquery('a') AND tsvector @@ to_tsquery('b')
The way that I chose to fix this was to have GIN call the consistentFn
twice with both TRUE and FALSE values for the in-doubt entry stream,
returning a hit if either call produces TRUE, but not if they both return
FALSE. The code handles this for the case of a single in-doubt entry stream,
but punts (falling back to the stupid behavior) if there's more than one lossy
reference to the same page. The idea could be scaled up to deal with multiple
lossy references, but I think that would probably be wasted complexity. At
least to judge by Artur's example, such cases don't occur often enough to be
worth trying to optimize.
Back-patch to 8.4. 8.3 did not have lossy GIN index pointers, so not
subject to these problems.
Transaction aborts now record their LSN to avoid corner case
behaviour in SR/HS, hence change of name of variables and functions.
As pointed out by Fujii Masao. Cosmetic changes only.
max_standby_streaming_delay, and revise the implementation to avoid assuming
that timestamps found in WAL records can meaningfully be compared to clock
time on the standby server. Instead, the delay limits are compared to the
elapsed time since we last obtained a new WAL segment from archive or since
we were last "caught up" to WAL data arriving via streaming replication.
This avoids problems with clock skew between primary and standby, as well
as other corner cases that the original coding would misbehave in, such
as the primary server having significant idle time between transactions.
Per my complaint some time ago and considerable ensuing discussion.
Do some desultory editing on the hot standby documentation, too.
master. Otherwise a subsequent crash could cause the master to lose WAL that
has already been applied on the slave, resulting in the slave being out of
sync and soon corrupt. Per recent discussion and an example from Robert Haas.
Fujii Masao
confusion with streaming-replication settings. Also, change its default
value to "off", because of concern about executing new and poorly-tested
code during ordinary non-replicating operation. Per discussion.
In passing do some minor editing of related documentation.
archival or hot standby should be WAL-logged, instead of deducing that from
other options like archive_mode. This replaces recovery_connections GUC in
the primary, where it now has no effect, but it's still used in the standby
to enable/disable hot standby.
Remove the WAL-logging of "unlogged operations", like creating an index
without WAL-logging and fsyncing it at the end. Instead, we keep a copy of
the wal_mode setting and the settings that affect how much shared memory a
hot standby server needs to track master transactions (max_connections,
max_prepared_xacts, max_locks_per_xact) in pg_control. Whenever the settings
change, at server restart, write a WAL record noting the new settings and
update pg_control. This allows us to notice the change in those settings in
the standby at the right moment, they used to be included in checkpoint
records, but that meant that a changed value was not reflected in the
standby until the first checkpoint after the change.
Bump PG_CONTROL_VERSION and XLOG_PAGE_MAGIC. Whack XLOG_PAGE_MAGIC back to
the sequence it used to follow, before hot standby and subsequent patches
changed it to 0x9003.
vacuum_log_cleanup_info() now generates log records with a valid
latestRemovedXid set in all cases. Also be careful not to zero the
value when we do a round of vacuuming part-way through lazy_scan_heap().
Incidentally, this reduces frequency of conflicts in Hot Standby.
Also, make the name of the GUC and the name of the backing variable match.
Alnong the way, clean up a couple of slight typographical errors in the
related docs.
through normal backends. Makes code clearer also, since we
avoid various Assert()s. Performance of snapshots taken
during recovery no longer depends upon number of read-only
backends.
after actually removing one, so that if we can't remove segments because
WAL archiving is lagging behind, we don't unnecessarily forbid streaming
the old not-yet-archived segments that are still perfectly valid. Per
suggestion from Fujii Masao.
doesn't take into account how far the WAL senders are. This way a hung
WAL sender doesn't prevent old WAL segments from being recycled/removed
in the primary, ultimately causing the disk to fill up. Instead add
standby_keep_segments setting to control how many old WAL segments are
kept in the primary. This also makes it more reliable to use streaming
replication without WAL archiving, assuming that you set
standby_keep_segments high enough.
The error message now makes explicit reference to the GUC that must be changed
to fix the problem, using wording suggested by Tom Lane. Along the way,
rename the GUC from MaxWalSenders to max_wal_senders for consistency and
grep-ability.
WAL record for btree delete contains a list of tids, even when backup
blocks are present. We follow the tids to their heap tuples, taking
care to follow LP_REDIRECT tuples. We ignore LP_DEAD tuples on the
understanding that they will always have xmin/xmax earlier than any
LP_NORMAL tuples referred to by killed index tuples. Iff all tuples
are LP_DEAD we return InvalidTransactionId. The heap relfilenode is
added to the WAL record, requiring API changes to pass down the heap
Relation. XLOG_PAGE_MAGIC updated.
present since 8.0 was never fully meaningful, since two recovery targets
cannot be specified. Refactor recovery target type to make this change
and associated code easier to understand. No change in function.
Bug report arising from internal support question.
field into WAL record and reset it from there, rather than using
FrozenTransactionId which can lead to some corner case bugs.
Problem report and suggested route to a fix from Heikki, details by me.
enabled. Bypassing the kernel cache is counter-productive in that case,
because the archiver/walsender process will read from the WAL file
soon after it's written, and if it's not cached the read will cause
a physical read, eating I/O bandwidth available on the WAL drive.
Also, walreceiver process does unaligned writes, so disable O_DIRECT
in walreceiver process for that reason too.
In addition, add support for a "payload" string to be passed along with
each notify event.
This implementation should be significantly more efficient than the old one,
and is also more compatible with Hot Standby usage. There is not yet any
facility for HS slaves to receive notifications generated on the master,
although such a thing is possible in future.
Joachim Wieland, reviewed by Jeff Davis; also hacked on by me.
where a database has a non-default tablespaceid. Pass thru MyDatabaseId
and MyDatabaseTableSpace to allow file path to be re-created in
standby and correct invalidation to take place in all cases.
Update and rework xact_commit_desc() debug messages.
Bug report from Tom by code inspection. Fix by me.
several places, but for now only GIN uses it during index creation.
Using self-balanced tree greatly speeds up index creation in corner cases
with preordered data.
VACUUM FULL INPLACE), along with a boatload of subsidiary code and complexity.
Per discussion, the use case for this method of vacuuming is no longer large
enough to justify maintaining it; not to mention that we don't wish to invest
the work that would be needed to make it play nicely with Hot Standby.
Aside from the code directly related to old-style VACUUM FULL, this commit
removes support for certain WAL record types that could only be generated
within VACUUM FULL, redirect-pointer removal in heap_page_prune, and
nontransactional generation of cache invalidation sinval messages (the last
being the sticking point for Hot Standby).
We still have to retain all code that copes with finding HEAP_MOVED_OFF and
HEAP_MOVED_IN flag bits on existing tuples. This can't be removed as long
as we want to support in-place update from pre-9.0 databases.
of shared or nailed system catalogs. This has two key benefits:
* The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs.
* We no longer have to use an unsafe reindex-in-place approach for reindexing
shared catalogs.
CLUSTER on nailed catalogs now works too, although I left it disabled on
shared catalogs because the resulting pg_index.indisclustered update would
only be visible in one database.
Since reindexing shared system catalogs is now fully transactional and
crash-safe, the former special cases in REINDEX behavior have been removed;
shared catalogs are treated the same as non-shared.
This commit does not do anything about the recently-discussed problem of
deadlocks between VACUUM FULL/CLUSTER on a system catalog and other
concurrent queries; will address that in a separate patch. As a stopgap,
parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid
such failures during the regression tests.
false positives during Hot Standby conflict processing. Simple
patch to enhance conflict processing, following previous discussions.
Controlled by parameter minimize_standby_conflicts = on | off, with
default off allows measurement of performance impact to see whether
it should be set on all the time.
Attributes can now have options, just as relations and tablespaces do, and
the reloptions code is used to parse, validate, and store them. For
simplicity and because these options are not performance critical, we store
them in a separate cache rather than the main relcache.
Thanks to Alex Hunsaker for the review.
that would've been WAL-logged if archiving was enabled. If we encounter
such records in archive recovery anyway, we know that some data is
missing from the log. A WARNING is emitted in that case.
Original patch by Fujii Masao, with changes by me.
This includes two new kinds of postmaster processes, walsenders and
walreceiver. Walreceiver is responsible for connecting to the primary server
and streaming WAL to disk, while walsender runs in the primary server and
streams WAL from disk to the client.
Documentation still needs work, but the basics are there. We will probably
pull the replication section to a new chapter later on, as well as the
sections describing file-based replication. But let's do that as a separate
patch, so that it's easier to see what has been added/changed. This patch
also adds a new section to the chapter about FE/BE protocol, documenting the
protocol used by walsender/walreceivxer.
Bump catalog version because of two new functions,
pg_last_xlog_receive_location() and pg_last_xlog_replay_location(), for
monitoring the progress of replication.
Fujii Masao, with additional hacking by me
Previously, fastgetattr() and heap_getattr() tested their fourth argument
against a null pointer, but any attempt to use them with a literal-NULL
fourth argument evaluated to *(void *)0, resulting in a compiler error.
Remove these NULL tests to avoid leading future readers of this code to
believe that this has a chance of working. Also clean up related legacy
code in nocachegetattr(), heap_getsysattr(), and nocache_index_getattr().
The new coding standard is that any code which calls a getattr-type
function or macro which takes an isnull argument MUST pass a valid
boolean pointer. Per discussion with Bruce Momjian, Tom Lane, Alvaro
Herrera.
This patch only supports seq_page_cost and random_page_cost as parameters,
but it provides the infrastructure to scalably support many more.
In particular, we may want to add support for effective_io_concurrency,
but I'm leaving that as future work for now.
Thanks to Tom Lane for design help and Alvaro Herrera for the review.
to be just a minor extension of the previous patch that made "x IS NULL"
indexable, because we can treat the IS NOT NULL condition as if it were
"x < NULL" or "x > NULL" (depending on the index's NULLS FIRST/LAST option),
just like IS NULL is treated like "x = NULL". Aside from any possible
usefulness in its own right, this is an important improvement for
index-optimized MAX/MIN aggregates: it is now reliably possible to get
a column's min or max value cheaply, even when there are a lot of nulls
cluttering the interesting end of the index.
This is more in keeping with modern practice, and is a first step towards
porting to Win64 (which has sizeof(pointer) > sizeof(long)).
Tsutomu Yamada, Magnus Hagander, Tom Lane
Enabled by recovery_connections = on (default) and forcing archive recovery using a recovery.conf. Recovery processing now emulates the original transactions as they are replayed, providing full locking and MVCC behaviour for read only queries. Recovery must enter consistent state before connections are allowed, so there is a delay, typically short, before connections succeed. Replay of recovering transactions can conflict and in some cases deadlock with queries during recovery; these result in query cancellation after max_standby_delay seconds have expired. Infrastructure changes have minor effects on normal running, though introduce four new types of WAL record.
New test mode "make standbycheck" allows regression tests of static command behaviour on a standby server while in recovery. Typical and extreme dynamic behaviours have been checked via code inspection and manual testing. Few port specific behaviours have been utilised, though primary testing has been on Linux only so far.
This commit is the basic patch. Additional changes will follow in this release to enhance some aspects of behaviour, notably improved handling of conflicts, deadlock detection and query cancellation. Changes to VACUUM FULL are also required.
Simon Riggs, with significant and lengthy review by Heikki Linnakangas, including streamlined redesign of snapshot creation and two-phase commit.
Important contributions from Florian Pflug, Mark Kirkwood, Merlin Moncure, Greg Stark, Gianni Ciolli, Gabriele Bartolini, Hannu Krosing, Robert Haas, Tatsuo Ishii, Hiroyuki Yamada plus support and feedback from many other community members.
be part of multixacts, so allocate a slot for each prepared transaction in
the "oldest member" array in multixact.c. On PREPARE TRANSACTION, transfer
the oldest member value from the current backends slot to the prepared xact
slot. Also save and recover the value from the 2pc state file.
The symptom of the bug was that after a transaction prepared, a shared lock
still held by the prepared transaction was sometimes ignored by other
transactions.
Fix back to 8.1, where both 2PC and multixact were introduced.
that it can scribble on scan->xs_ctup.t_self while following HOT chains,
so we can't rely on that to stay valid between hashgettuple() calls.
Introduce a private variable in HashScanOpaque, instead.
hash indexes keep entries sorted by hash value. First, the original plans for
concurrency assumed that insertions would happen only at the end of a page,
which is no longer true; this could cause scans to transiently fail to find
index entries in the presence of concurrent insertions. We can compensate
by teaching scans to re-find their position after re-acquiring read locks.
Second, neither the bucket split nor the bucket compaction logic had been
fixed to preserve hashvalue ordering, so application of either of those
processes could lead to permanent corruption of an index, in the sense
that searches might fail to find entries that are present.
This patch fixes the split and compaction logic to preserve hashvalue
ordering, but it cannot do anything about pre-existing corruption. We will
need to recommend reindexing all hash indexes in the 8.4.2 release notes.
To buy back the performance loss hereby induced in split and compaction,
fix them to use PageIndexMultiDelete instead of retail PageIndexDelete
operations. We might later want to do something with qsort'ing the
page contents rather than doing a binary search for each insertion,
but that seemed more invasive than I cared to risk in a back-patch.
Per bug #5157 from Jeff Janes and subsequent investigation.
tuple size limit. Improve the error message for index-tuple-too-large
so that it includes the actual size, the limit, and the index name.
Sync with the btree occurrences of the same error.
Back-patch to 8.4 because it appears that the out-of-sync problem
is occurring in the field.
Teodor and Tom
own database's datfrozenxid, if the current value is old enough to be
forcing autovacuums or warning messages. This ensures that a bogus
value is replaced as soon as possible. Per a comment from Heikki.
Recent commits have removed the various uses it was supporting. It was a
performance bottleneck, according to bug report #4919 by Lauris Ulmanis; seems
it slowed down user creation after a billion users.
XID) in checkpoint records. This eliminates the need to recompute the value
from scratch during database startup, which is one of the two remaining
reasons for the flatfile code to exist. It should also simplify life for
hot-standby operation.
To avoid bloating the checkpoint records unreasonably, I switched from
tracking the oldest database by name to tracking it by OID. This turns
out to save cycles in general (everywhere but the warning-generating
paths, which we hardly care about) and also helps us deal with the case
that the oldest database got dropped instead of being vacuumed. The prior
coding might go for a long time without updating the wrap limit in that case,
which is bad because it might result in a lot of useless autovacuum activity.
"all tuples visible" flag in heap page headers. The flag update *must*
be applied before calling XLogInsert, but heap_update and the tuple
moving routines in VACUUM FULL were ignoring this rule. A crash and
replay could therefore leave the flag incorrectly set, causing rows
to appear visible in seqscans when they should not be. This might explain
recent reports of data corruption from Jeff Ross and others.
In passing, do a bit of editorialization on comments in visibilitymap.c.
by supporting conversions in places that used to demand exact rowtype match.
Since this issue is certain to come up elsewhere (in fact, already has,
in ExecEvalConvertRowtype), factor out the support code into new core
functions for tuple conversion. I chose to put these in a new source
file since heaptuple.c is already overly long.
Heavily revised version of a patch by Pavel Stehule.
values being complained of.
In passing, also remove the arbitrary length limitation in the similar
error detail message for foreign key violations.
Itagaki Takahiro
The current implementation fires an AFTER ROW trigger for each tuple that
looks like it might be non-unique according to the index contents at the
time of insertion. This works well as long as there aren't many conflicts,
but won't scale to massive unique-key reassignments. Improving that case
is a TODO item.
Dean Rasheed
not forced out-of-line unless that is necessary to make the row fit on a
page. Previously, they were forced out-of-line if needed to get the row
down to the default target size (1/4th page).
Kevin Grittner
archive recovery. Invent a separate state variable and inquiry function
for XLogInsertAllowed() to clarify some tests and make the management of
writing the end-of-recovery checkpoint less klugy. Fix several places
that were incorrectly testing InRecovery when they should be looking at
RecoveryInProgress or XLogInsertAllowed (because they will now be executed
in the bgwriter not startup process). Clarify handling of bad LSNs passed
to XLogFlush during recovery. Use a spinlock for setting/testing
SharedRecoveryInProgress. Improve quite a lot of comments.
Heikki and Tom
during it:
When bgwriter is active, the startup process can't perform mdsync() correctly
because it won't see the fsync requests accumulated in bgwriter's private
pendingOpsTable. Therefore make bgwriter responsible for the end-of-recovery
checkpoint as well, when it's active.
When bgwriter is active (= archive recovery), the startup process must not
accumulate fsync requests to its own pendingOpsTable, since bgwriter won't
see them there when it performs restartpoints. Make startup process drop its
pendingOpsTable when bgwriter is launched to avoid that.
Update minimum recovery point one last time when leaving archive recovery.
It won't be updated by the end-of-recovery checkpoint because XLogFlush()
sees us as out of recovery already.
This fixes bug #4879 reported by Fujii Masao.
behavior in cases where we don't know the heap tuple count accurately; in
particular partial vacuum, but this also makes the API a bit more useful
for ANALYZE. This patch adds "estimated_count" flags to both structs so
that an approximate count can be flagged as such, and adjusts the logic
so that approximate counts are not used for updating pg_class.reltuples.
This fixes my previous complaint that VACUUM was putting ridiculous values
into pg_class.reltuples for indexes. The actual impact of that bug is
limited, because the planner only pays attention to reltuples for an index
if the index is partial; which probably explains why beta testers hadn't
noticed a degradation in plan quality from it. But it needs to be fixed.
The whole thing is a bit messy and should be redesigned in future, because
reltuples now has the potential to drift quite far away from reality when
a long period elapses with no non-partial vacuums. But this is as good as
it's going to get for 8.4.
is supposed to remove duplicate heap TIDs, we have to be sure to reduce the
tuple size and posting-item count accordingly in addItemPointersToTuple().
Failing to do so resulted in the effective injection of garbage TIDs into the
index contents, ie, whatever happened to be in the memory palloc'd for the
new tuple. I'm not sure that this fully explains the index corruption
reported by Tatsuo Ishii, but the test case I'm using no longer fails.
should use GinItemPointerGetBlockNumber/GinItemPointerGetOffsetNumber,
not ItemPointerGetBlockNumber/ItemPointerGetOffsetNumber, because the latter
will Assert() on ip_posid == 0, ie a "Min" pointer. (Thus, ItemPointerIsMin
has never worked at all, but it seems unused at present.) I'm not certain
that the case can occur in normal functioning, but it's blowing up on me
while investigating Tatsuo-san's data corruption problem. In any case it
seems like a problem waiting to bite someone.
Back-patch just in case this really is a problem for somebody in the field.
errors when tables are concurrently dropped. To do this we must take lock
on each relation before we check its privileges. The old code was trying
to do that the other way around, which is a bit pointless when there are lots
of other commands that lock relations before checking privileges. I did keep
it checking each relation's privilege before locking the next relation, which
is a detail that ALTER TABLE isn't too picky about.
To implement this without almost duplicating the reloption table, treat
relopt_kind as a bitmask instead of an integer value. This decreases the
range of allowed values, but it's not clear that there's need for that much
values anyway.
This patch also makes heap_reloptions explicitly a no-op for relation kinds
other than heap and TOAST tables.
Patch by ITAGAKI Takahiro with minor edits from me. (In particular I removed
the bit about adding relation kind to an error message, which I intend to
commit separately.)
method to pass extra data to the consistent() and comparePartial() methods.
This is the core infrastructure needed to support the soon-to-appear
contrib/btree_gin module. The APIs are still upward compatible with the
definitions used in 8.3 and before, although *not* with the previous 8.4devel
function definitions.
catversion bump for changes in pg_proc entries (although these are just
cosmetic, since GIN doesn't actually look at the function signature before
calling it...)
Teodor Sigaev and Oleg Bartunov
them from degrading badly when the input is sorted or nearly so. In this
scenario the tree is unbalanced to the point of becoming a mere linked list,
so insertions become O(N^2). The easiest and most safely back-patchable
solution is to stop growing the tree sooner, ie limit the growth of N. We
might later consider a rebalancing tree algorithm, but it's not clear that
the benefit would be worth the cost and complexity. Per report from Sergey
Burladyan and an earlier complaint from Heikki.
Back-patch to 8.2; older versions didn't have GIN indexes.
multiple index entries in a holding area before adding them to the main index
structure. This helps because bulk insert is (usually) significantly faster
than retail insert for GIN.
This patch also removes GIN support for amgettuple-style index scans. The
API defined for amgettuple is difficult to support with fastupdate, and
the previously committed partial-match feature didn't really work with
it either. We might eventually figure a way to put back amgettuple
support, but it won't happen for 8.4.
catversion bumped because of change in GIN's pg_am entry, and because
the format of GIN indexes changed on-disk (there's a metapage now,
and possibly a pending list).
Teodor Sigaev
its usual buffer cleaning duties during archive recovery, and it's responsible
for performing restartpoints.
This requires some changes in postmaster. When the startup process has done
all the initialization and is ready to start WAL redo, it signals the
postmaster to launch the background writer. The postmaster is signaled again
when the point in recovery is reached where we know that the database is in
consistent state. Postmaster isn't interested in that at the moment, but
that's the point where we could let other backends in to perform read-only
queries. The postmaster is signaled third time when the recovery has ended,
so that postmaster knows that it's safe to start accepting connections.
The startup process now traps SIGTERM, and performs a "clean" shutdown. If
you do a fast shutdown during recovery, a shutdown restartpoint is performed,
like a shutdown checkpoint, and postmaster kills the processes cleanly. You
still have to continue the recovery at next startup, though.
Currently, the background writer is only launched during archive recovery.
We could launch it during crash recovery as well, but it seems better to keep
that codepath as simple as possible, for the sake of robustness. And it
couldn't do any restartpoints during crash recovery anyway, so it wouldn't be
that useful.
log_restartpoints is gone. Use log_checkpoints instead. This is yet to be
documented.
This whole operation is a pre-requisite for Hot Standby, but has some value of
its own whether the hot standby patch makes 8.4 or not.
Simon Riggs, with lots of modifications by me.
qualifier, and add support for this in pg_dump.
This allows TOAST tables to have user-defined fillfactor, and will also
enable us to move the autovacuum parameters to reloptions without taking
away the possibility of setting values for TOAST tables.
refactor the relcache code that used to do that. This allows other callers
(particularly autovacuum) to do the same without necessarily having to open
and lock a table.
be used instead of the normal exclusive lock, and make WAL redo functions
responsible for calling RestoreBkpBlocks(). They know better what kind of a
lock they need.
At the moment, this just moves things around with no functional change, but
makes the hot standby patch that's under review cleaner.
fillRelOptions routine that stores the parsed values in the struct using a
table-based approach. Per Tom suggestion. Also remove the "continue"
in HANDLE_*_RELOPTION macros, which were useless and in spirit they were
assuming too much of how the macros were going to be used. (Note that these
macros are now unused, but the intention is to introduce some usage in a
future autovacuum patch, which is why they weren't completely removed.)
Also, do not call the string validation routine when not validating. It seems
less error-prone this way, per commentary on the amoptions SGML docs.
bitmap. This is extracted from Greg Stark's posix_fadvise patch; it seems
worth committing separately, since it's potentially useful independently of
posix_fadvise.
a more complete framework for writing custom option processing routines
by user-defined access methods.
Catalog version bumped due to the general API changes, which are going to
affect user-defined "amoptions" routines.
heap page, where a set bit indicates that all tuples on the page are
visible to all transactions, and the page therefore doesn't need
vacuuming. It is stored in a new relation fork.
Lazy vacuum uses the visibility map to skip pages that don't need
vacuuming. Vacuum is also responsible for setting the bits in the map.
In the future, this can hopefully be used to implement index-only-scans,
but we can't currently guarantee that the visibility map is always 100%
up-to-date.
In addition to the visibility map, there's a new PD_ALL_VISIBLE flag on
each heap page, also indicating that all tuples on the page are visible to
all transactions. It's important that this flag is kept up-to-date. It
is also used to skip visibility tests in sequential scans, which gives a
small performance gain on seqscans.
that a Portal is a useful and sufficient additional argument for
CreateDestReceiver --- it just isn't, in most cases. Instead formalize
the approach of passing any needed parameters to the receiver separately.
One unexpected benefit of this change is that we can declare typedef Portal
in a less surprising location.
This patch is just code rearrangement and doesn't change any functionality.
I'll tackle the HOLD-cursor-vs-toast problem in a follow-on patch.
truncations in FSM code, call FreeSpaceMapTruncateRel from smgr_redo. To
make that cleaner from modularity point of view, move the WAL-logging one
level up to RelationTruncate, and move RelationTruncate and all the
related WAL-logging to new src/backend/catalog/storage.c file. Introduce
new RelationCreateStorage and RelationDropStorage functions that are used
instead of calling smgrcreate/smgrscheduleunlink directly. Move the
pending rel deletion stuff from smgrcreate/smgrscheduleunlink to the new
functions. This leaves smgr.c as a thin wrapper around md.c; all the
transactional stuff is now in storage.c.
This will make it easier to add new forks with similar truncation logic,
like the visibility map.
heap_form_tuple. Since this removes the last remaining caller of
heap_addheader, remove it.
Extracted from the column privileges patch from Stephen Frost, with further
code cleanups by me.
(but not locked, as that would risk deadlocks). Also, make it work in a small
ring of buffers to avoid having bulk inserts trash the whole buffer arena.
Robert Haas, after an idea of Simon Riggs'.
and heap_deformtuple in favor of the newer functions heap_form_tuple et al
(which do the same things but use bool control flags instead of arbitrary
char values). Eliminate the former duplicate coding of these functions,
reducing the deprecated functions to mere wrappers around the newer ones.
We can't get rid of them entirely because add-on modules probably still
contain many instances of the old coding style.
Kris Jurka
functions into one ReadBufferExtended function, that takes the strategy
and mode as argument. There's three modes, RBM_NORMAL which is the default
used by plain ReadBuffer(), RBM_ZERO, which replaces ZeroOrReadBuffer, and
a new mode RBM_ZERO_ON_ERROR, which allows callers to read corrupt pages
without throwing an error. The FSM needs the new mode to recover from
corrupt pages, which could happend if we crash after extending an FSM file,
and the new page is "torn".
Add fork number to some error messages in bufmgr.c, that still lacked it.
written to temp files by tuplesort.c and tuplestore.c. This saves 2 bytes per
row for 32-bit machines, and 6 bytes per row for 64-bit machines, which seems
worth the slight additional uglification of the tuple read/write routines.
correctly set. As result, killtuple() marks as dead
wrong tuple on page. Bug was introduced by me while fixing
possible duplicates during GiST index scan.
This patch eliminates the marking of subtransactions as SUBCOMMITTED in pg_clog
during their commit; instead they remain in-progress until main transaction
commit. At main transaction commit, the commit protocol is atomic-by-page
instead of one transaction at a time. To avoid a race condition with some
subtransactions appearing committed before others in the case where they span
more than one pg_clog page, we conserve the logic that marks them subcommitted
before marking the parent committed.
Simon Riggs with minor help from me
is NULL but SK_SEARCHNULL is not set. Add checking IS NULL of keys
to set during key initialization. If key is NULL and SK_SEARCHNULL is not
set then nothnig can be satisfied.
With assert-enabled compilation that causes coredump.
Bug was introduced in 8.3 by support of IS NULL index scan.
of referencing a WITH item that's not yet in scope according to the SQL
spec's semantics. This seems to be an easy error to make, and the bare
"relation doesn't exist" message doesn't lead one's mind in the correct
direction to fix it.
free space information is stored in a dedicated FSM relation fork, with each
relation (except for hash indexes; they don't use FSM).
This eliminates the max_fsm_relations and max_fsm_pages GUC options; remove any
trace of them from the backend, initdb, and documentation.
Rewrite contrib/pg_freespacemap to match the new FSM implementation. Also
introduce a new variant of the get_raw_page(regclass, int4, int4) function in
contrib/pageinspect that let's you to return pages from any relation fork, and
a new fsm_page_contents() function to inspect the new FSM pages.
value. This means that hash index lookups are always lossy and have to be
rechecked when the heap is visited; however, the gain in index compactness
outweighs this when the indexed values are wide. Also, we only need to
perform datatype comparisons when the hash codes match exactly, rather than
for every entry in the hash bucket; so it could also win for datatypes that
have expensive comparison functions. A small additional win is gained by
keeping hash index pages sorted by hash code and using binary search to reduce
the number of index tuples we have to look at.
Xiao Meng
This commit also incorporates Zdenek Kotala's patch to isolate hash metapages
and hash bitmaps a bit better from the page header datastructures.
at once and ItemPointers are collected in memory.
Remove tuple's killing by killtuple() if tuple was moved to another
page - it could produce unaceptable overhead.
Backpatch up to 8.1 because the bug was introduced by GiST's concurrency support.
of multiple forks, and each fork can be created and grown separately.
The bulk of this patch is about changing the smgr API to include an extra
ForkNumber argument in every smgr function. Also, smgrscheduleunlink and
smgrdounlink no longer implicitly call smgrclose, because other forks might
still exist after unlinking one. The callers of those functions have been
modified to call smgrclose instead.
This patch in itself doesn't have any user-visible effect, but provides the
infrastructure needed for upcoming patches. The additional forks envisioned
are a rewritten FSM implementation that doesn't rely on a fixed-size shared
memory block, and a visibility map to allow skipping portions of a table in
VACUUM that have no dead tuples.
thereby forestalling any problems with alignment of the data structure placed
there. Since SizeOfPageHeaderData is maxalign'd anyway in 8.3 and HEAD, this
does not actually change anything right now, but it is foreseeable that the
header size will change again someday. I had to fix a couple of places that
were assuming that the content offset is just SizeOfPageHeaderData rather than
MAXALIGN(SizeOfPageHeaderData). Per discussion of Zdenek's page-macros patch.
SizeOfPageHeaderData instead of sizeof(PageHeaderData) in places where that
makes the code clearer, and avoid casting between Page and PageHeader where
possible. Zdenek Kotala, with some additional cleanup by Heikki Linnakangas.
I did not apply the parts of the proposed patch that would have resulted in
slightly changing the on-disk format of hash indexes; it seems to me that's
not a win as long as there's any chance of having in-place upgrade for 8.4.
corresponding struct definitions. This allows other headers to avoid including
certain highly-loaded headers such as rel.h and relscan.h, instead using just
relcache.h, heapam.h or genam.h, which are more lightweight and thus cause less
unnecessary dependencies.
forks. XLogOpenRelation() and the associated light-weight relation cache in
xlogutils.c is gone, and XLogReadBuffer() now takes a RelFileNode as argument,
instead of Relation.
For functions that still need a Relation struct during WAL replay, there's a
new function called CreateFakeRelcacheEntry() that returns a fake entry like
XLogOpenRelation() used to.
algorithm, replacing the original intention of a one-pass search, which
had been hacked up over time to be partially two-pass in hopes of handling
various corner cases better. It still wasn't quite there, especially as
regards emitting unwanted NOTICE messages. More importantly, this approach
lets us fix a number of open bugs concerning concurrent DROP scenarios,
because we can take locks during the first pass and avoid traversing to
dependent objects that were just deleted by someone else.
There is more that can be done here, but I'll go ahead and commit the
base patch before working on the options.
Formerly, the default value of wal_sync_method was determined inside xlog.c,
but now it is determined inside guc.c. guc.c was reading xlogdefs.h
without having read <fcntl.h>, leading to wrong determination of
DEFAULT_SYNC_METHOD. Obviously xlogdefs.h needs to include <fcntl.h>
for itself to ensure stable results.
unnecessary #include lines in it. Also, move some tuple routine prototypes and
macros to htup.h, which allows removal of heapam.h inclusion from some .c
files.
For this to work, a new header file access/sysattr.h needed to be created,
initially containing attribute numbers of system columns, for pg_dump usage.
While at it, make contrib ltree, intarray and hstore header files more
consistent with our header style.
<craig@postnewspapers.com.au>.
It was my mistake, I missed limitation of number of held locks, now GIN doesn't
use continiuous locks, but still hold buffers pinned to prevent interference
with vacuum's deletion algorithm.
Backpatch is needed.
corrupted. (Neither is very important if SIGTERM is used to shut down the
whole database cluster together, but there's a problem if someone tries to
SIGTERM individual backends.) To do this, introduce new infrastructure
macros PG_ENSURE_ERROR_CLEANUP/PG_END_ENSURE_ERROR_CLEANUP that take care
of transiently pushing an on_shmem_exit cleanup hook. Also use this method
for createdb cleanup --- that wasn't a shared-memory-corruption problem,
but SIGTERM abort of createdb could leave orphaned files lying around.
Backpatch as far as 8.2. The shmem corruption cases don't exist in 8.1,
and the createdb usage doesn't seem important enough to risk backpatching
further.
instead of plan time. Extend the amgettuple API so that the index AM returns
a boolean indicating whether the indexquals need to be rechecked, and make
that rechecking happen in nodeIndexscan.c (currently the only place where
it's expected to be needed; other callers of index_getnext are just erroring
out for now). For the moment, GIN and GIST have stub logic that just always
sets the recheck flag to TRUE --- I'm hoping to get Teodor to handle pushing
that control down to the opclass consistent() functions. The planner no
longer pays any attention to amopreqcheck, and that catalog column will go
away in due course.
systable_endscan_ordered that have API similar to systable_beginscan etc
(in particular, the passed-in scankeys have heap not index attnums),
but guarantee ordered output, unlike the existing functions. For the moment
these are just very thin wrappers around index_beginscan/index_getnext/etc.
Someday they might need to get smarter; but for now this is just a code
refactoring exercise to reduce the number of direct callers of index_getnext,
in preparation for changing that function's API.
In passing, remove index_getnext_indexitem, which has been dead code for
quite some time, and will have even less use than that in the presence
of run-time-lossy indexes.
indexscan always occurs in one call, and the results are returned in a
TIDBitmap instead of a limited-size array of TIDs. This should improve
speed a little by reducing AM entry/exit overhead, and it is necessary
infrastructure if we are ever to support bitmap indexes.
In an only slightly related change, add support for TIDBitmaps to preserve
(somewhat lossily) the knowledge that particular TIDs reported by an index
need to have their quals rechecked when the heap is visited. This facility
is not really used yet; we'll need to extend the forced-recheck feature to
plain indexscans before it's useful, and that hasn't been coded yet.
The intent is to use it to clean up 8.3's horrid @@@ kluge for text search
with weighted queries. There might be other uses in future, but that one
alone is sufficient reason.
Heikki Linnakangas, with some adjustments by me.
currently support this because we must be able to build Vars referencing
join columns, and varattno is only 16 bits wide. Perhaps this should be
improved in future, but considering that it never came up before, I'm not
sure the problem is worth much effort. Per bug #4070 from Marcello
Ceschia.
The problem seems largely academic in 8.0 and 7.4, because they have
(different) O(N^2) performance issues with such wide joins, but
back-patch all the way anyway.
snapmgmt.c file for the former. The header files have also been reorganized
in three parts: the most basic snapshot definitions are now in a new file
snapshot.h, and the also new snapmgmt.h keeps the definitions for snapmgmt.c.
tqual.h has been reduced to the bare minimum.
This patch is just a first step towards managing live snapshots within a
transaction; there is no functionality change.
Per my proposal to pgsql-patches on 20080318191940.GB27458@alvh.no-ip.org and
subsequent discussion.
bucket number, so as to ensure locality of access to the index during the
insertion step. Without this, building an index significantly larger than
available RAM takes a very long time because of thrashing. On the other
hand, sorting is just useless overhead when the index does fit in RAM.
We choose to sort when the initial index size exceeds effective_cache_size.
This is a revised version of work by Tom Raney and Shreya Bhargava.
two buckets at the start, we create a number of buckets appropriate for the
estimated size of the table. This avoids a lot of expensive bucket-split
actions during initial index build on an already-populated table.
This is one of the two core ideas of Tom Raney and Shreya Bhargava's patch
to reduce hash index build time. I'm committing it separately to make it
easier for people to test the effects of this separately from the effects
of their other core idea (pre-sorting the index entries by bucket number).
before it goes groveling through the ProcArray. In situations where the same
recently-committed transaction ID is checked repeatedly by tqual.c, this saves
a lot of shared-memory searches. And it's cheap enough that it shouldn't
hurt noticeably when it doesn't help.
Concept and patch by Simon, some minor tweaking and comment-cleanup by Tom.
it accumulates the set of changes to be made and then applies them. It had
to accumulate the set of changes anyway to prepare a WAL record for the
pruning action, so this isn't an enormous change; the only new complexity is
to not doubly mark tuples that are visited twice in the scan. The main
advantage is that we can substantially reduce the scope of the critical
section in which the changes are applied, thus avoiding PANIC in foreseeable
cases like running out of memory in inval.c. A nice secondary advantage is
that it is now far clearer that WAL replay will actually do the same thing
that the original pruning did.
This commit doesn't do anything about the open problem that
CacheInvalidateHeapTuple doesn't have the right semantics for a CTID change
caused by collapsing out a redirect pointer. But whatever we do about that,
it'll be a good idea to not do it inside a critical section.
temporary table; we can't support that because there's no way to clean up the
source backend's internal state if the eventual COMMIT PREPARED is done by
another backend. This was checked correctly in 8.1 but I broke it in 8.2 :-(.
Patch by Heikki Linnakangas, original trouble report by John Smith.
data structures and backend internal APIs. This solves problems we've seen
recently with inconsistent layout of pg_control between machines that have
32-bit time_t and those that have already migrated to 64-bit time_t. Also,
we can get out from under the problem that Windows' Unix-API emulation is not
consistent about the width of time_t.
There are a few remaining places where local time_t variables are used to hold
the current or recent result of time(NULL). I didn't bother changing these
since they do not affect any cross-module APIs and surely all platforms will
have 64-bit time_t before overflow becomes an actual risk. time_t should
be avoided for anything visible to extension modules, however.
its second pass over the table. It has to start at block zero, else the
"merge join" logic for detecting which TIDs are already in the index
doesn't work. Hence, extend heapam.c's API so that callers can enable or
disable syncscan. (I put in an option to disable buffer access strategy,
too, just in case somebody needs it.) Per report from Hannes Dorbath.
constraint status of copied indexes (bug #3774), as well as various other
small bugs such as failure to pstrdup when needed. Allow INCLUDING INDEXES
indexes to be merged with identical declared indexes (perhaps not real useful,
but the code is there and having it not apply to LIKE indexes seems pretty
unorthogonal). Avoid useless work in generateClonedIndexStmt(). Undo some
poorly chosen API changes, and put a couple of routines in modules that seem
to be better places for them.
but no database changes have been made since the last CommandCounterIncrement.
This should result in a significant improvement in the number of "commands"
that can typically be performed within a transaction before hitting the 2^32
CommandId size limit. In particular this buys back (and more) the possible
adverse consequences of my previous patch to fix plan caching behavior.
The implementation requires tracking whether the current CommandCounter
value has been "used" to mark any tuples. CommandCounter values stored into
snapshots are presumed not to be used for this purpose. This requires some
small executor changes, since the executor used to conflate the curcid of
the snapshot it was using with the command ID to mark output tuples with.
Separating these concepts allows some small simplifications in executor APIs.
Something for the TODO list: look into having CommandCounterIncrement not do
AcceptInvalidationMessages. It seems fairly bogus to be doing it there,
but exactly where to do it instead isn't clear, and I'm disinclined to mess
with asynchronous behavior during late beta.
Else, in a 64-bit machine with maintenance_work_mem set to above 4Gb,
the counter overflows and we never recognize having reached the
maintenance_work_mem limit. I believe this explains out-of-memory
failure recently reported by Sean Davis.
This is a bug, so backpatch to 8.2.
it failed for splits of non-leaf pages because in such pages the first
data key on a page is suppressed, and so we can't just copy the first
key from the right page to reconstitute the left page's high key.
Problem found by Koichi Suzuki, patch by Heikki.
- create a separate archive_mode GUC, on which archive_command is dependent
- %r option in recovery.conf sends last restartpoint to recovery command
- %r used in pg_standby, updated README
- minor other code cleanup in pg_standby
- doc on Warm Standby now mentions pg_standby and %r
- log_restartpoints recovery option emits LOG message at each restartpoint
- end of recovery now displays last transaction end time, as requested
by Warren Little; also shown at each restartpoint
- restart archiver if needed to carry away WAL files at shutdown
Simon Riggs
columns, and the new version can be stored on the same heap page, we no longer
generate extra index entries for the new version. Instead, index searches
follow the HOT-chain links to ensure they find the correct tuple version.
In addition, this patch introduces the ability to "prune" dead tuples on a
per-page basis, without having to do a complete VACUUM pass to recover space.
VACUUM is still needed to clean up dead index entries, however.
Pavan Deolasee, with help from a bunch of other people.
ReadNewTransactionId from GetSnapshotData --- with a "latestCompletedXid"
variable that is updated during transaction commit or abort. Since
latestCompletedXid is written only in places that had to lock ProcArrayLock
exclusively anyway, and is read only in places that had to lock ProcArrayLock
shared anyway, it adds no new locking requirements to the system despite being
cluster-wide. Moreover, removing ReadNewTransactionId from snapshot
acquisition eliminates the need to take both XidGenLock and ProcArrayLock at
the same time. Since XidGenLock is sometimes held across I/O this can be a
significant win. Some preliminary benchmarking suggested that this patch has
no effect on average throughput but can significantly improve the worst-case
transaction times seen in pgbench. Concept by Florian Pflug, implementation
by Tom Lane.
rows will normally never obtain an XID at all. We already did things this way
for subtransactions, but this patch extends the concept to top-level
transactions. In applications where there are lots of short read-only
transactions, this should improve performance noticeably; not so much from
removal of the actual XID-assignments, as from reduction of overhead that's
driven by the rate of XID consumption. We add a concept of a "virtual
transaction ID" so that active transactions can be uniquely identified even
if they don't have a regular XID. This is a much lighter-weight concept:
uniqueness of VXIDs is only guaranteed over the short term, and no on-disk
record is made about them.
Florian Pflug, with some editorialization by Tom.
Oleg Bartunov and Teodor Sigaev, but I did a lot of editorializing,
so anything that's broken is probably my fault.
Documentation is nonexistent as yet, but let's land the patch so we can
get some portability testing done.
before reporting a transaction committed. Data consistency is still
guaranteed (unlike setting fsync = off), but a crash may lose the effects
of the last few transactions. Patch by Simon, some editorialization by Tom.
and fsync WAL at convenient intervals. For the moment it just tries to
offload this work from backends, but soon it will be responsible for
guaranteeing a maximum delay before asynchronously-committed transactions
will be flushed to disk.
This is a portion of Simon Riggs' async-commit patch, committed to CVS
separately because a background WAL writer seems like it might be a good idea
independently of the async-commit feature. I rebased walwriter.c on
bgwriter.c because it seemed like a more appropriate way of handling signals;
while the startup/shutdown logic in postmaster.c is more like autovac because
we want walwriter to quit before we start the shutdown checkpoint.
over a fairly long period of time, rather than being spat out in a burst.
This happens only for background checkpoints carried out by the bgwriter;
other cases, such as a shutdown checkpoint, are still done at full speed.
Remove the "all buffers" scan in the bgwriter, and associated stats
infrastructure, since this seems no longer very useful when the checkpoint
itself is properly throttled.
Original patch by Itagaki Takahiro, reworked by Heikki Linnakangas,
and some minor API editorialization by me.
pseudo HeapScanDesc created for a bitmap heap scan. This avoids some useless
overhead during a bitmap scan startup, in particular invoking the syncscan
code. (We might someday want to do that, but right now it's merely useless
contention for shared memory, to say nothing of possibly pushing useful
entries out of syncscan's small LRU list.) This also allows elimination of
ugly pgstat_discount_heap_scan() kluge.
delivering a well-randomized hash value. I got religion on this after
observing that performance of multi-batch hash join degrades terribly if the
higher-order bits of hash values aren't random, as indeed was true for say
hashes of small integer values. It's now expected and documented that hash
functions should use hash_any or some comparable method to ensure that all
bits of their output are about equally random.
initdb forced because this change invalidates existing hash indexes. For the
same reason, this isn't back-patchable; the hash join performance problem
will get a band-aid fix in the back branches.
buffers, rather than blowing out the whole shared-buffer arena. Aside from
avoiding cache spoliation, this fixes the problem that VACUUM formerly tended
to cause a WAL flush for every page it modified, because we had it hacked to
use only a single buffer. Those flushes will now occur only once per
ring-ful. The exact ring size, and the threshold for seqscans to switch into
the ring usage pattern, remain under debate; but the infrastructure seems
done. The key bit of infrastructure is a new optional BufferAccessStrategy
object that can be passed to ReadBuffer operations; this replaces the former
StrategyHintVacuum API.
This patch also changes the buffer usage-count methodology a bit: we now
advance usage_count when first pinning a buffer, rather than when last
unpinning it. To preserve the behavior that a buffer's lifetime starts to
decrease when it's released, the clock sweep code is modified to not decrement
usage_count of pinned buffers.
Work not done in this commit: teach GiST and GIN indexes to use the vacuum
BufferAccessStrategy for vacuum-driven fetches.
Original patch by Simon, reworked by Heikki and again by Tom.
and aborted transactions have different effects; also teach it not to assume
that prepared transactions are always committed.
Along the way, simplify the pgstats API by tying counting directly to
Relations; I cannot detect any redeeming social value in having stats
pointers in HeapScanDesc and IndexScanDesc structures. And fix a few
corner cases in which counts might be missed because the relation's
pgstat_info pointer hadn't been set.
WAL records that shows whether it is safe to remove full-page images
(ie, whether or not an on-line backup was in progress when the WAL entry
was made). Also make provision for an XLOG_NOOP record type that can be
used to fill in the extra space when decompressing the data for restore.
This is the portion of Koichi Suzuki's "full page writes" patch that
has to go into the core database. The remainder of that work is two
external compression and decompression programs, which for the time being
will undergo separate development on pgfoundry. Per discussion.
Also, twiddle the handling of BTREE_SPLIT records to ensure it'll be
possible to compress them (the previous coding caused essential info
to be omitted). The other commonly-used record types seem OK already,
with the possible exception of GIN and GIST WAL records, which I don't
understand well enough to opine on.
to avoid losing useful Xid information in not-so-old tuples. This makes
CLUSTER behave the same as VACUUM as far a tuple-freezing behavior goes
(though CLUSTER does not yet advance the table's relfrozenxid).
While at it, move the actual freezing operation in rewriteheap.c to a more
appropriate place, and document it thoroughly. This part of the patch from
Tom Lane.
pages it intends to zero immediately. Just to show there is some use for that
function besides WAL recovery :-).
Along the way, fold _hash_checkpage and _hash_pageinit calls into _hash_getbuf
and friends, instead of expecting callers to do that separately.
from time_t to TimestampTz representation. This provides full gettimeofday()
resolution of the timestamps, which might be useful when attempting to
do point-in-time recovery --- previously it was not possible to specify
the stop point with sub-second resolution. But mostly this is to get
rid of TimestampTz-to-time_t conversion overhead during commit. Per my
proposal of a day or two back.
messages to the stats collector. This avoids the problem that enabling
stats_row_level for autovacuum has a significant overhead for short
read-only transactions, as noted by Arjen van der Meijden. We can avoid
an extra gettimeofday call by piggybacking on the one done for WAL-logging
xact commit or abort (although that doesn't help read-only transactions,
since they don't WAL-log anything).
In my proposal for this, I noted that we could change the WAL log entries
for commit/abort to record full TimestampTz precision, instead of only
time_t as at present. That's not done in this patch, but will be committed
separately.
failed (due to lock conflicts or out-of-space). We might have already
extended the index's filesystem EOF before failing, causing the EOF to be
beyond what the metapage says is the last used page. Hence the invariant
maintained by the code needs to be "EOF is at or beyond last used page",
not "EOF is exactly the last used page". Problem was created by my patch
of 2006-11-19 that attempted to repair bug #2737. Since that was
back-patched to 7.4, this needs to be as well. Per report and test case
from Vlastimil Krejcir.
(original code *always* created a full-page image for the left page, thus
leaving the intended savings unrealized), avoid risk of not having enough room
on the page during xlog restore, squeeze out another couple bytes in the xlog
record, clean up neglected comments.
index types can be reliably distinguished by examining the special space
on an index page. Per my earlier proposal, plus the realization that
there's no need for btree's vacuum cycle ID to cycle through every possible
16-bit value. Restricting its range a little costs nearly nothing and
eliminates the possibility of collisions.
Memo to self: remember to make bitmap indexes play along with this scheme,
assuming that patch ever gets accepted.
This commit breaks any code that assumes that the mere act of forming a tuple
(without writing it to disk) does not "toast" any fields. While all available
regression tests pass, I'm not totally sure that we've fixed every nook and
cranny, especially in contrib.
Greg Stark with some help from Tom Lane
Add the latter to the values checked in pg_control, since it can't be changed
without invalidating toast table content. This commit in itself shouldn't
change any behavior, but it lays some necessary groundwork for experimentation
with these toast-control numbers.
Note: while TOAST_TUPLE_THRESHOLD can now be changed without initdb, some
thought still needs to be given to needs_toast_table() in toasting.c before
unleashing random changes.
module and teach PREPARE and protocol-level prepared statements to use it.
In service of this, rearrange utility-statement processing so that parse
analysis does not assume table schemas can't change before execution for
utility statements (necessary because we don't attempt to re-acquire locks
for utility statements when reusing a stored plan). This requires some
refactoring of the ProcessUtility API, but it ends up cleaner anyway,
for instance we can get rid of the QueryContext global.
Still to do: fix up SPI and related code to use the plan cache; I'm tempted to
try to make SQL functions use it too. Also, there are at least some aspects
of system state that we want to ensure remain the same during a replan as in
the original processing; search_path certainly ought to behave that way for
instance, and perhaps there are others.
Get rid of VARATT_SIZE and VARATT_DATA, which were simply redundant with
VARSIZE and VARDATA, and as a consequence almost no code was using the
longer names. Rename the length fields of struct varlena and various
derived structures to catch anyplace that was accessing them directly;
and clean up various places so caught. In itself this patch doesn't
change any behavior at all, but it is necessary infrastructure if we hope
to play any games with the representation of varlena headers.
Greg Stark and Tom Lane
keeping private state in each backend that has inserted and deleted the same
tuple during its current top-level transaction. This is sufficient since
there is no need to be able to determine the cmin/cmax from any other
transaction. This gets us back down to 23-byte headers, removing a penalty
paid in 8.0 to support subtransactions. Patch by Heikki Linnakangas, with
minor revisions by moi, following a design hashed out awhile back on the
pghackers list.
> Currently, an index split writes all the data on the split page to
> WAL. That's a lot of WAL traffic. The tuples that are copied to the
> right page need to be WAL logged, but the tuples that stay on the
> original page don't.
Heikki Linnakangas
describe the maximum size of index tuples (which is typically AM-dependent
anyway); and consequently remove the bogus deduction for "special space"
that was built into it.
Adjust TOAST_TUPLE_THRESHOLD and TOAST_MAX_CHUNK_SIZE to avoid wasting two
bytes per toast chunk, and to ensure that the calculation correctly tracks any
future changes in page header size. The computation had been inaccurate in a
way that didn't cause any harm except space wastage, but future changes could
have broken it more drastically.
Fix the calculation of BTMaxItemSize, which was formerly computed as 1 byte
more than it could safely be. This didn't cause any harm in practice because
it's only compared against maxalign'd lengths, but future changes in the size
of page headers or btree special space could have exposed the problem.
initdb forced because of change in TOAST_MAX_CHUNK_SIZE, which alters the
storage of toast tables.
threshold for tuple length. On 4-byte-MAXALIGN machines, the toast code
creates tuples that have t_len exactly TOAST_TUPLE_THRESHOLD ... but this
number is not itself maxaligned, so if heap_insert maxaligns t_len before
comparing to TOAST_TUPLE_THRESHOLD, it'll uselessly recurse back to
tuptoaster.c, wasting cycles. (It turns out that this does not happen on
8-byte-MAXALIGN machines, because for them the outer MAXALIGN in the
TOAST_MAX_CHUNK_SIZE macro reduces TOAST_MAX_CHUNK_SIZE so that toast tuples
will be less than TOAST_TUPLE_THRESHOLD in size. That MAXALIGN is really
incorrect, but we can't remove it now, see below.) There isn't any particular
value in maxaligning before comparing to the thresholds, so just don't do
that, which saves a small number of cycles in itself.
These numbers should be rejiggered to minimize wasted space on toast-relation
pages, but we can't do that in the back branches because changing
TOAST_MAX_CHUNK_SIZE would force an initdb (by changing the contents of toast
tables). We can move the toast decision thresholds a bit, though, which is
what this patch effectively does.
Thanks to Pavan Deolasee for discovering the unintended recursion.
Back-patch into 8.2, but not further, pending more testing. (HEAD is about
to get a further patch modifying the thresholds, so it won't help much
for testing this form of the patch.)
In this case extractQuery should returns -1 as nentries. This changes
prototype of extractQuery method to use int32* instead of uint32* for
nentries argument.
Based on that gincostestimate may see two corner cases: nothing will be found
or seqscan should be used.
Per proposal at http://archives.postgresql.org/pgsql-hackers/2007-01/msg01581.php
PS tsearch_core patch should be sightly modified to support changes, but I'm
waiting a verdict about reviewing of tsearch_core patch.
Hashing for aggregation purposes still needs work, so it's not time to
mark any cross-type operators as hashable for general use, but these cases
work if the operators are so marked by hand in the system catalogs.
currentMarkData from IndexScanDesc to the opaque structs for the
AMs that need this information (currently gist and hash).
Patch from Heikki Linnakangas, fixes by Neil Conway.
per-column options for btree indexes. The planner's support for this is still
pretty rudimentary; it does not yet know how to plan mergejoins with
nondefault ordering options. The documentation is pretty rudimentary, too.
I'll work on improving that stuff later.
Note incompatible change from prior behavior: ORDER BY ... USING will now be
rejected if the operator is not a less-than or greater-than member of some
btree opclass. This prevents less-than-sane behavior if an operator that
doesn't actually define a proper sort ordering is selected.
cases where we already hold the desired lock "indirectly", either via
membership in a MultiXact or because the lock was originally taken by a
different subtransaction of the current transaction. These cases must be
accounted for to avoid needless deadlocks and/or inappropriate replacement of
an exclusive lock with a shared lock. Per report from Clarence Gardner and
subsequent investigation.
in PITR scenarios. We now WAL-log the replacement of old XIDs with
FrozenTransactionId, so that such replacement is guaranteed to propagate to
PITR slave databases. Also, rather than relying on hint-bit updates to be
preserved, pg_clog is not truncated until all instances of an XID are known to
have been replaced by FrozenTransactionId. Add new GUC variables and
pg_autovacuum columns to allow management of the freezing policy, so that
users can trade off the size of pg_clog against the amount of freezing work
done. Revise the already-existing code that forces autovacuum of tables
approaching the wraparound point to make it more bulletproof; also, revise the
autovacuum logic so that anti-wraparound vacuuming is done per-table rather
than per-database. initdb forced because of changes in pg_class, pg_database,
and pg_autovacuum catalogs. Heikki Linnakangas, Simon Riggs, and Tom Lane.
deletion code to avoid the case where an upper-level btree page remains "half
dead" for a significant period of time, and to block insertions into a key
range that is in process of being re-assigned to the right sibling of the
deleted page's parent. This prevents the scenario reported by Ed L. wherein
index keys could become out-of-order in the grandparent index level.
Since this is a moderately invasive fix, I'm applying it only to HEAD.
The bug exists back to 7.4, but the back branches will get a different patch.
static variables. This avoids any risk of potential non-reentrancy,
and in particular offers a much cleaner workaround for the Intel compiler
bug that was affecting ginutil.c.
we probably should make them work reliably for all arrays. Fix code
to handle NULLs and multidimensional arrays, move it into arrayfuncs.c.
GIN is still restricted to indexing arrays with no null elements, however.
agreed these symbols are less easily confused. I made new pg_operator
entries (with new OIDs) for the old names, so as to provide backward
compatibility while making it pretty easy to remove the old names in
some future release cycle. This commit only touches the core datatypes,
contrib will be fixed separately.
on the same index page; we can avoid data copying as well as buffer refcount
manipulations in this common case. Makes for a small but noticeable
improvement in mergejoin speed.
Heikki Linnakangas
of the transaction ID counter. Nothing is done with the epoch except to
store it in checkpoint records, but this provides a foundation with which
add-on code can pretend that XIDs never wrap around. This is a severely
trimmed and rewritten version of the xxid patch submitted by Marko Kreen.
Per discussion, the epoch counter seems the only part of xxid that really
needs to be in the core server.
the rel, it's easy to get rid of the narrow race-condition window that
used to exist in VACUUM and CLUSTER. Did some minor code-beautification
work in the same area, too.
than N seconds apart. This allows a simple, if not very high performance,
means of guaranteeing that a PITR archive is no more than N seconds behind
real time. Also make pg_current_xlog_location return the WAL Write pointer,
add pg_current_xlog_insert_location to return the Insert pointer, and fix
pg_xlogfile_name_offset to return its results as a two-element record instead
of a smashed-together string, as per recent discussion.
Simon Riggs
operation every so often. This improves the usefulness of PITR log
shipping for hot standby: formerly, if the standby server crashed, it
was necessary to restart it from the last base backup and replay all
the WAL since then. Now it will only need to reread about the same
amount of WAL as the master server would. The behavior might also
come in handy during a long PITR replay sequence. Simon Riggs,
with some editorialization by Tom Lane.
to happen automatically during pg_stop_backup(). Add some functions for
interrogating the current xlog insertion point and for easily extracting
WAL filenames from the hex WAL locations displayed by pg_stop_backup
and friends. Simon Riggs with some editorialization by Tom Lane.
(table or index) before trying to open its relcache entry. This fixes
race conditions in which someone else commits a change to the relation's
catalog entries while we are in process of doing relcache load. Problems
of that ilk have been reported sporadically for years, but it was not
really practical to fix until recently --- for instance, the recent
addition of WAL-log support for in-place updates helped.
Along the way, remove pg_am.amconcurrent: all AMs are now expected to support
concurrent update.
When we are about to split an index page to do an insertion, first look
to see if any entries marked LP_DELETE exist on the page, and if so remove
them to try to make enough space for the desired insert. This should reduce
index bloat in heavily-updated tables, although of course you still need
VACUUM eventually to clean up the heap.
Junji Teramoto
Andres Freund
Heikki Linnakangas
Fujii Masao
Alvaro Herrera
Peter Eisentraut
Robert Haas
Bruce Momjian
Tom Lane
Magnus Hagander
Noah Misch
Greg Stark
Jeff Davis
Simon Riggs
Kevin Grittner
Joe Conway
Teodor Sigaev
Neil Conway