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
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
The operation that removes the remaining dead tuples from the page must
be WAL-logged before the setting of the VM bit. Otherwise, if you replay
the WAL to between those two records, you end up with the VM bit set, but
the dead tuples are still there.
Backpatch to 9.3, where this bug was introduced.
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
Vacuum recognizes that it can update relfrozenxid by checking whether it has
processed all pages of a relation. Unfortunately it performed that check
after truncating the dead pages at the end of the relation, and used the new
number of pages to decide whether all pages have been scanned. If the new
number of pages happened to be smaller or equal to the number of pages
scanned, it incorrectly decided that all pages were scanned.
This can lead to relfrozenxid being updated, even though some pages were
skipped that still contain old XIDs. That can lead to data loss due to xid
wraparounds with some rows suddenly missing. This likely has escaped notice
so far because it takes a large number (~2^31) of xids being used to see the
effect, while a full-table vacuum before that would fix the issue.
The incorrect logic was introduced by commit
b4b6923e03. Backpatch this fix down to 8.4,
like that commit.
Andres Freund, with some modifications by me.
Previously, if VACUUM skipped vacuuming a page because it's pinned, it
didn't count that page as scanned. However, that meant that relfrozenxid
was not bumped up either, which prevented anti-wraparound vacuum from
doing its job.
Report by Миша Тюрин, analysis and patch by Sergey Burladyn and Jeff Janes.
Backpatch to 9.2, where the skip-locked-pages behavior was introduced.
Use a critical section when setting the all-visible flag on an empty page,
and WAL-logging it. log_newpage_buffer() contains an assertion that it
must be called inside a critical section, and it's the right thing to do
when modifying a buffer anyway.
Also, the page should be marked dirty before calling log_newpage_buffer(),
per the comment in log_newpage_buffer() and src/backend/access/transam/README.
Patch by Andres Freund, in response to my report. Backpatch to 9.2, like
the patch that introduced these bugs (a6370fd9).
There is a rare race condition, when a transaction that inserted a tuple
aborts while vacuum is processing the page containing the inserted tuple.
Vacuum prunes the page first, which normally removes any dead tuples, but
if the inserting transaction aborts right after that, the loop after
pruning will see a dead tuple and remove it instead. That's OK, but if the
page is on a table with no indexes, and the page becomes completely empty
after removing the dead tuple (or tuples) on it, it will be immediately
marked as all-visible. That's OK, but the sanity check in vacuum would
throw a warning because it thinks that the page contains dead tuples and
was nevertheless marked as all-visible, even though it just vacuumed away
the dead tuples and so it doesn't actually contain any.
Spotted this while reading the code. It's difficult to hit the race
condition otherwise, but can be done by putting a breakpoint after the
heap_page_prune() call.
Backpatch all the way to 8.4, where this code first appeared.
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
Previously, these functions took a HeapTupleHeader, but upcoming
patches for logical replication will introduce new a new snapshot
type under which the tuple's TID will be used to lookup (CMIN, CMAX)
for visibility determination purposes. This makes that information
available. Code churn is minimal since HeapTupleSatisfiesVisibility
took the HeapTuple anyway, and deferenced it before calling the
satisfies function.
Independently of logical replication, this allows t_tableOid and
t_self to be cross-checked via assertions in tqual.c. This seems
like a useful way to make sure that all callers are setting these
values properly, which has been previously put forward as
desirable.
Andres Freund, reviewed by Álvaro Herrera
Previously this state was represented by whether the view's disk file had
zero or nonzero size, which is problematic for numerous reasons, since it's
breaking a fundamental assumption about heap storage. This was done to
allow unlogged matviews to revert to unpopulated status after a crash
despite our lack of any ability to update catalog entries post-crash.
However, this poses enough risk of future problems that it seems better to
not support unlogged matviews until we can find another way. Accordingly,
revert that choice as well as a number of existing kluges forced by it
in favor of creating a pg_class.relispopulated flag column.
Patch b19e4250b4 attempted to
preserve existing behavior regarding statistics generation in the
case that a truncation attempt was canceled due to lock conflicts.
It failed to do this accurately in two regards: (1) autovacuum had
previously generated statistics if the truncate attempt failed to
initially get the lock rather than having started the attempt, and
(2) the VACUUM ANALYZE command had always generated statistics.
Both of these changes were unintended, and are reverted by this
patch. On review, there seems to be consensus that the previous
failure to generate statistics when the truncate was terminated
was more an unfortunate consequence of how that effort was
previously terminated than a feature we want to keep; so this
patch generates statistics even when an autovacuum truncation
attempt terminates early. Another unintended change which is kept
on the basis that it is an improvement is that when a VACUUM
command is truncating, it will the new heuristic for avoiding
blocking other processes, rather than keeping an
AccessExclusiveLock on the table for however long the truncation
takes.
Per multiple reports, with some renaming per patch by Jeff Janes.
Backpatch to 9.0, where problem was created.
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.
Remove use of PageSetTLI() from all page manipulation functions
and adjust README to indicate change in the way we make changes
to pages. Repurpose those bytes into the pd_checksum field and
explain how that works in comments about page header.
Refactoring ahead of actual feature patch which would make use
of the checksum field, arriving later.
Jeff Davis, with comments and doc changes by Simon Riggs
Direction suggested by Robert Haas; many others providing
review comments.
There's a high chance that a page becomes all-visible when the second phase
of vacuum removes all the dead tuples on it, so it makes sense to check for
that. Otherwise the visibility map won't get updated until the next vacuum.
Pavan Deolasee, reviewed by Jeff Janes.
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
In situations where there are over 8MB of empty pages at the end of
a table, the truncation work for trailing empty pages takes longer
than deadlock_timeout, and there is frequent access to the table by
processes other than autovacuum, there was a problem with the
autovacuum worker process being canceled by the deadlock checking
code. The truncation work done by autovacuum up that point was
lost, and the attempt tried again by a later autovacuum worker. The
attempts could continue indefinitely without making progress,
consuming resources and blocking other processes for up to
deadlock_timeout each time.
This patch has the autovacuum worker checking whether it is
blocking any other thread at 20ms intervals. If such a condition
develops, the autovacuum worker will persist the work it has done
so far, release its lock on the table, and sleep in 50ms intervals
for up to 5 seconds, hoping to be able to re-acquire the lock and
try again. If it is unable to get the lock in that time, it moves
on and a worker will try to continue later from the point this one
left off.
While this patch doesn't change the rules about when and what to
truncate, it does cause the truncation to occur sooner, with less
blocking, and with the consumption of fewer resources when there is
contention for the table's lock.
The only user-visible change other than improved performance is
that the table size during truncation may change incrementally
instead of just once.
This problem exists in all supported versions but is infrequently
reported, although some reports of performance problems when
autovacuum runs might be caused by this. Initial commit is just the
master branch, but this should probably be backpatched once the
build farm and general developer usage confirm that there are no
surprising effects.
Jan Wieck
During VACUUM if we pause to perform a cycle
of index cleanup we drop the vmbuffer pin,
so we should do the same thing when heap
scan completes. This avoids holding vmbuffer
pin across the main index cleanup in VACUUM,
which could be minutes or hours longer than
necessary for correctness.
Bug report and suggested fix from Pavan Deolasee
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.
In lazy_scan_heap, we could issue bogus warnings about incorrect
information in the visibility map, because we checked the visibility
map bit before locking the heap page, creating a race condition. Fix
by rechecking the visibility map bit before we complain. Rejigger
some related logic so that we rely on the possibly-outdated
all_visible_according_to_vm value as little as possible.
In heap_multi_insert, it's not safe to clear the visibility map bit
before beginning the critical section. The visibility map is not
crash-safe unless we treat clearing the bit as a critical operation.
Specifically, if the transaction were to error out after we set the
bit and before entering the critical section, we could end up writing
the heap page to disk (with the bit cleared) and crashing before the
visibility map page made it to disk. That would be bad. heap_insert
has this correct, but somehow the order of operations got rearranged
when heap_multi_insert was added.
Also, add some more comments to visibilitymap_test, lazy_scan_heap,
and IndexOnlyNext, expounding on concurrency issues.
Per extensive code review by Andres Freund, and further review by Tom
Lane, who also made the original report about the bogus warnings.
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.
We must set the visibility map bit before releasing our exclusive lock
on the heap page; otherwise, someone might clear the heap page bit
before we set the visibility map bit, leading to a situation where the
visibility map thinks the page is all-visible but it's really not.
This problem has existed since 8.4, but it wasn't critical before we
had index-only scans, since the worst case scenario was that the page
wouldn't get vacuumed until the next scan_all vacuum.
Along the way, a couple of minor, related improvements: (1) if we
pause the heap scan to do an index vac cycle, release any visibility
map page we're holding, since really long-running pins are not good
for a variety of reasons; and (2) warn if we see a page that's marked
all-visible in the visibility map but not on the page level, since
that should never happen any more (it was allowed in previous
releases, but not in 9.2).
Previously, we used SetBufferCommitInfoNeedsSave, but that's really
intended for dirty-marks we can theoretically afford to lose, such as
hint bits. As for 9.2, the PD_ALL_VISIBLE mustn't be lost in this
way, since we could then end up with a heap page that isn't
all-visible and a visibility map page that is all visible, causing
index-only scans to return wrong answers.
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.
As noted by Tom Lane, the previous coding in this area, which I
introduced in commit bbb6e559c4, was
poorly tested and caused the vacuum's second heap to go into what would
have been an infinite loop but for the fact that it eventually caused a
memory allocation failure. This version seems to work better.
This adds some I/O stats to the logging of autovacuum (when the
operation takes long enough that log_autovacuum_min_duration causes it
to be logged), so that it is easier to tune. Notably, it adds buffer
I/O counts (hits, misses, dirtied) and read and write rate.
Authors: Greg Smith and Noah Misch
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
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.
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.
It is possible for VACUUM to scan no pages at all, if the visibility map
shows that all pages are all-visible. In this situation VACUUM has no new
information to report about the relation's tuple density, so it wasn't
changing pg_class.reltuples ... but it updated pg_class.relpages anyway.
That's wrong in general, since there is no evidence to justify changing the
density ratio reltuples/relpages, but it's particularly bad if the previous
state was relpages=reltuples=0, which means "unknown tuple density".
We just replaced "unknown" with "zero". ANALYZE would eventually recover
from this, but it could take a lot of repetitions of ANALYZE to do so if
the relation size is much larger than the maximum number of pages ANALYZE
will scan, because of the moving-average behavior introduced by commit
b4b6923e03.
The only known situation where we could have relpages=reltuples=0 and yet
the visibility map asserts everything's visible is immediately following
a pg_upgrade. It might be advisable for pg_upgrade to try to preserve the
relpages/reltuples statistics; but in any case this code is wrong on its
own terms, so fix it. Per report from Sergey Koposov.
Back-patch to 8.4, where the visibility map was introduced, same as the
previous change.
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.
When we added the ability for vacuum to skip heap pages by consulting the
visibility map, we made it just not update the reltuples/relpages
statistics if it skipped any pages. But this could leave us with extremely
out-of-date stats for a table that contains any unchanging areas,
especially for TOAST tables which never get processed by ANALYZE. In
particular this could result in autovacuum making poor decisions about when
to process the table, as in recent report from Florian Helmberger. And in
general it's a bad idea to not update the stats at all. Instead, use the
previous values of reltuples/relpages as an estimate of the tuple density
in unvisited pages. This approach results in a "moving average" estimate
of reltuples, which should converge to the correct value over multiple
VACUUM and ANALYZE cycles even when individual measurements aren't very
good.
This new method for updating reltuples is used by both VACUUM and ANALYZE,
with the result that we no longer need the grotty interconnections that
caused ANALYZE to not update the stats depending on what had happened
in the parent VACUUM command.
Also, fix the logic for skipping all-visible pages during VACUUM so that it
looks ahead rather than behind to decide what to do, as per a suggestion
from Greg Stark. This eliminates useless scanning of all-visible pages at
the start of the relation or just after a not-all-visible page. In
particular, the first few pages of the relation will not be invariably
included in the scanned pages, which seems to help in not overweighting
them in the reltuples estimate.
Back-patch to 8.4, where the visibility map was introduced.
Bruce Momjian
Robert Haas
Alvaro Herrera
Heikki Linnakangas
Simon Riggs
Tom Lane
Kevin Grittner
Peter Eisentraut