The new facility makes it easier to optimize bulk loading, as the
logic for buffering, WAL-logging, and syncing the relation only needs
to be implemented once. It's also less error-prone: We have had a
number of bugs in how a relation is fsync'd - or not - at the end of a
bulk loading operation. By centralizing that logic to one place, we
only need to write it correctly once.
The new facility is faster for small relations: Instead of of calling
smgrimmedsync(), we register the fsync to happen at next checkpoint,
which avoids the fsync latency. That can make a big difference if you
are e.g. restoring a schema-only dump with lots of relations.
It is also slightly more efficient with large relations, as the WAL
logging is performed multiple pages at a time. That avoids some WAL
header overhead. The sorted GiST index build did that already, this
moves the buffering to the new facility.
The changes to pageinspect GiST test needs an explanation: Before this
patch, the sorted GiST index build set the LSN on every page to the
special GistBuildLSN value, not the LSN of the WAL record, even though
they were WAL-logged. There was no particular need for it, it just
happened naturally when we wrote out the pages before WAL-logging
them. Now we WAL-log the pages first, like in B-tree build, so the
pages are stamped with the record's real LSN. When the build is not
WAL-logged, we still use GistBuildLSN. To make the test output
predictable, use an unlogged index.
Reviewed-by: Andres Freund
Discussion: https://www.postgresql.org/message-id/30e8f366-58b3-b239-c521-422122dd5150%40iki.fi
This function reads directly a page from a relation, relying on
index_open() to open the index to read from. Unfortunately, this would
crash when using partitioned indexes, as these can be opened with
index_open() but they have no physical pages.
Alexander has fixed the module, while I have written the test.
Author: Alexander Lakhin, Michael Paquier
Discussion: https://postgr.es/m/18246-f4d9ff7cb3af77e6@postgresql.org
Backpatch-through: 12
When building BRIN indexes, the brinbuildCallback only advances to the
next page range when seeing a tuple that doesn't belong to the current
one. This means that the index may end up missing ranges at the end of
the table, if those pages do not contain any indexable tuples.
We tend not to have completely empty pages at the end of a relation, but
this also applies to partial indexes, where the tuples may simply not
match the index predicate. This results in inefficient scans using the
affected BRIN index - without the summaries, the page ranges have to be
read and processed, which consumes I/O and possibly also CPU time.
The existing code already added empty ranges for earlier parts of the
table, this commit makes sure we add them for the ranges at the end of
the table too.
Patch by Matthias van de Meent, with review/improvements by me.
Author: Matthias van de Meent
Reviewed-by: Tomas Vondra
Discussion: https://postgr.es/m/CAEze2WiMsPZg%3DxkvSF_jt4%3D69k6K7gz5B8V2wY3gCGZ%2B1BzCbQ%40mail.gmail.com
Non-leaf pages of GiST indexes contain key attributes, leaf pages
contain both key and non-key attributes, and gist_page_items() ignored
the handling of non-key attributes. This caused a few problems when
using gist_page_items() on a GiST index with INCLUDE:
- On a non-leaf page, the function would crash.
- On a leaf page, the function would work, but miss to display all the
values for included attributes.
This commit fixes gist_page_items() to handle such cases in a more
appropriate way, and now displays the values of key and non-key
attributes for each item separately in a style consistent with what
ruleutils.c would generate for the attribute list, depending on the page
type dealt with. In a way similar to how a record is displayed, values
would be double-quoted for key or non-key attributes if required.
ruleutils.c did not provide a routine able to control if non-key
attributes should be displayed, so an extended() routine for index
definitions is added to work around the leaf and non-leaf page
differences.
While on it, this commit fixes a third problem related to the amount of
data reported for key attributes. The code originally relied on
BuildIndexValueDescription() (used for error reports on constraints)
that would not print all the data stored in the index but the index
opclass's input type, so this limited the amount of information
available. This switch makes gist_page_items() much cheaper as there is
no need to run ACL checks for each item printed, which is not an issue
anyway as superuser rights are required to execute the functions of
pageinspect. Opclasses whose data cannot be displayed can rely on
gist_page_items_bytea().
The documentation of this function was slightly incorrect for the
output results generated on HEAD and v15, so adjust it on these
branches.
Author: Alexander Lakhin, Michael Paquier
Discussion: https://postgr.es/m/17884-cb8c326522977acb@postgresql.org
Backpatch-through: 14
Commit 3581cbdcd6 added a flag to identify empty BRIN ranges. This adds
the new flag to brin_page_items() output.
This is kept as a separate commit as it should not be backpatched.
Reviewed-by: Justin Pryzby, Matthias van de Meent, Alvaro Herrera
Discussion: https://postgr.es/m/402430e4-7d9d-6cf1-09ef-464d80afff3b@enterprisedb.com
Attempting to use this function with a raw page not coming from a GiST
index would cause a crash, as it was missing the same sanity checks as
gist_page_items_bytea(). This slightly refactors the code so as all the
basic validation checks for GiST pages are done in a single routine,
in the same fashion as the pageinspect functions for hash and BRIN.
This fixes an issue similar to 076f4d9. A test is added to stress for
this case. While on it, I have added a similar test for
brin_page_items() with a combination make of a valid GiST index and a
raw btree page. This one was already protected, but it was not tested.
Reported-by: Egor Chindyaskin
Author: Dmitry Koval
Discussion: https://postgr.es/m/17815-fc4a2d3b74705703@postgresql.org
Backpatch-through: 14
This is like the existing bt_page_stats() function, but it can
report on a range of pages rather than just one at a time.
I don't have a huge amount of faith in the portability of the
new test cases, but they do pass in a 32-bit FreeBSD VM here.
Further adjustment may be needed depending on buildfarm results.
Hamid Akhtar, reviewed by Naeem Akhter, Bertrand Drouvot,
Bharath Rupireddy, and myself
Discussion: https://postgr.es/m/CANugjht-=oGMRmNJKMqnBC69y7vr+wHDmm0ZK6-1pJsxoBKBbA@mail.gmail.com
pageinspect has occasionally failed on slow buildfarm members,
with symptoms indicating that the expected effects of VACUUM
FREEZE didn't happen. This is presumably because a background
transaction such as auto-analyze was holding back global xmin.
We can work around that by using a temp table in the test.
Since commit a7212be8b, that will use an up-to-date cutoff xmin
regardless of other processes. And pageinspect itself shouldn't
really care whether the table is temp.
Back-patch to v14. There would be no point in older branches
without back-patching a7212be8b, which seems like more trouble
than the problem is worth.
Discussion: https://postgr.es/m/2892135.1668976646@sss.pgh.pa.us
Getting from get_raw_page() an all-zero page is considered as a valid
case by the buffer manager and it can happen for example when finding a
corrupted page with zero_damaged_pages enabled (using zero_damaged_pages
to look at corrupted pages happens), or after a crash when a relation
file is extended before any WAL for its new data is generated (before a
vacuum or autovacuum job comes in to do some cleanup).
However, all the functions of pageinspect, as of the index AMs (except
hash that has its own idea of new pages), heap, the FSM or the page
header have never worked with all-zero pages, causing various crashes
when going through the page internals.
This commit changes all the pageinspect functions to be compliant with
all-zero pages, where the choice is made to return NULL or no rows for
SRFs when finding a new page. get_raw_page() still works the same way,
returning a batch of zeros in the bytea of the page retrieved. A hard
error could be used but NULL, while more invasive, is useful when
scanning relation files in full to get a batch of results for a single
relation in one query. Tests are added for all the code paths
impacted.
Reported-by: Daria Lepikhova
Author: Michael Paquier
Discussion: https://postgr.es/m/561e187b-3549-c8d5-03f5-525c14e65bd0@postgrespro.ru
Backpatch-through: 10
A couple of code paths use the special area on the page passed by the
function caller, expecting to find some data in it. However, feeding
an incorrect page can lead to out-of-bound reads when trying to access
the page special area (like a heap page that has no special area,
leading PageGetSpecialPointer() to grab a pointer outside the allocated
page).
The functions used for hash and btree indexes have some protection
already against that, while some other functions using a relation OID
as argument would make sure that the access method involved is correct,
but functions taking in input a raw page without knowing the relation
the page is attached to would run into problems.
This commit improves the set of checks used in the code paths of BRIN,
btree (including one check if a leaf page is found with a non-zero
level), GIN and GiST to verify that the page given in input has a
special area size that fits with each access method, which is done
though PageGetSpecialSize(), becore calling PageGetSpecialPointer().
The scope of the checks done is limited to work with pages that one
would pass after getting a block with get_raw_page(), as it is possible
to craft byteas that could bypass existing code paths. Having too many
checks would also impact the usability of pageinspect, as the existing
code is very useful to look at the content details in a corrupted page,
so the focus is really to avoid out-of-bound reads as this is never a
good thing even with functions whose execution is limited to
superusers.
The safest approach could be to rework the functions so as these fetch a
block using a relation OID and a block number, but there are also cases
where using a raw page is useful.
Tests are added to cover all the code paths that needed such checks, and
an error message for hash indexes is reworded to fit better with what
this commit adds.
Reported-By: Alexander Lakhin
Author: Julien Rouhaud, Michael Paquier
Discussion: https://postgr.es/m/16527-ef7606186f0610a1@postgresql.org
Discussion: https://postgr.es/m/561e187b-3549-c8d5-03f5-525c14e65bd0@postgrespro.ru
Backpatch-through: 10
This commit fixes a set of issues related to the use of the SQL
functions in this module when the caller is able to pass down raw page
data as input argument:
- The page size check was fuzzy in a couple of places, sometimes
looking after only a sub-range, but what we are looking for is an exact
match on BLCKSZ. After considering a few options here, I have settled
down to do a generalization of get_page_from_raw(). Most of the SQL
functions already used that, and this is not strictly required if not
accessing an 8-byte-wide value from a raw page, but this feels safer in
the long run for alignment-picky environment, particularly if a code
path begins to access such values. This also reduces the number of
strings that need to be translated.
- The BRIN function brin_page_items() uses a Relation but it did not
check the access method of the opened index, potentially leading to
crashes. All the other functions in need of a Relation already did
that.
- Some code paths could fail on elog(), but we should to use ereport()
for failures that can be triggered by the user.
Tests are added to stress all the cases that are fixed as of this
commit, with some junk raw pages (\set VERBOSITY ensures that this works
across all page sizes) and unexpected index types when functions open
relations.
Author: Michael Paquier, Justin Prysby
Discussion: https://postgr.es/m/20220218030020.GA1137@telsasoft.com
Backpatch-through: 10
The GiST sorted build currently chooses split points according to the only page
space utilization. That may lead to higher non-leaf keys overlap and, in turn,
slower search query answers.
This commit makes the sorted build use the opclass's picksplit method. Once
four pages at the level are accumulated, the picksplit method is applied until
each split partition fits the page. Some of our split algorithms could show
significant performance degradation while processing 4-times more data at once.
But those opclasses haven't received the sorted build support and shouldn't
receive it before their split algorithms are improved.
Discussion: https://postgr.es/m/CAHqSB9jqtS94e9%3D0vxqQX5dxQA89N95UKyz-%3DA7Y%2B_YJt%2BVW5A%40mail.gmail.com
Author: Aliaksandr Kalenik, Sergei Shoulbakov, Andrey Borodin
Reviewed-by: Björn Harrtell, Darafei Praliaskouski, Andres Freund
Reviewed-by: Alexander Korotkov
ld_upper, ld_lower, pd_special and the page size have been using
smallint as return type, which could cause those fields to return
negative values in certain cases for builds configures with a page size
of 32kB.
Bump pageinspect to 1.10. page_header() is able to handle the correct
return type of those fields at runtime when using an older version of
the extension, with some tests are added to cover that.
Author: Quan Zongliang
Reviewed-by: Michael Paquier, Bharath Rupireddy
Discussion: https://postgr.es/m/8b8ec36e-61fe-14f9-005d-07bc85aa4eed@yeah.net
Most error messages about a relkind that was not supported or
appropriate for the command was of the pattern
"relation \"%s\" is not a table, foreign table, or materialized view"
This style can become verbose and tedious to maintain. Moreover, it's
not very helpful: If I'm trying to create a comment on a TOAST table,
which is not supported, then the information that I could have created
a comment on a materialized view is pointless.
Instead, write the primary error message shorter and saying more
directly that what was attempted is not possible. Then, in the detail
message, explain that the operation is not supported for the relkind
the object was. To simplify that, add a new function
errdetail_relkind_not_supported() that does this.
In passing, make use of RELKIND_HAS_STORAGE() where appropriate,
instead of listing out the relkinds individually.
Reviewed-by: Michael Paquier <michael@paquier.xyz>
Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://www.postgresql.org/message-id/flat/dc35a398-37d0-75ce-07ea-1dd71d98f8ec@2ndquadrant.com
Otherwise we risk "leaking" deleted pages by making them non-recyclable
indefinitely. Commit 6655a729 did the same thing for deleted pages in
GiST indexes. That work was used as a starting point here.
Stop storing an XID indicating the oldest bpto.xact across all deleted
though unrecycled pages in nbtree metapages. There is no longer any
reason to care about that condition/the oldest XID. It only ever made
sense when wraparound was something _bt_vacuum_needs_cleanup() had to
consider.
The btm_oldest_btpo_xact metapage field has been repurposed and renamed.
It is now btm_last_cleanup_num_delpages, which is used to remember how
many non-recycled deleted pages remain from the last VACUUM (in practice
its value is usually the precise number of pages that were _newly
deleted_ during the specific VACUUM operation that last set the field).
The general idea behind storing btm_last_cleanup_num_delpages is to use
it to give _some_ consideration to non-recycled deleted pages inside
_bt_vacuum_needs_cleanup() -- though never too much. We only really
need to avoid leaving a truly excessive number of deleted pages in an
unrecycled state forever. We only do this to cover certain narrow cases
where no other factor makes VACUUM do a full scan, and yet the index
continues to grow (and so actually misses out on recycling existing
deleted pages).
These metapage changes result in a clear user-visible benefit: We no
longer trigger full index scans during VACUUM operations solely due to
the presence of only 1 or 2 known deleted (though unrecycled) blocks
from a very large index. All that matters now is keeping the costs and
benefits in balance over time.
Fix an issue that has been around since commit 857f9c36, which added the
"skip full scan of index" mechanism (i.e. the _bt_vacuum_needs_cleanup()
logic). The accuracy of btm_last_cleanup_num_heap_tuples accidentally
hinged upon _when_ the source value gets stored. We now always store
btm_last_cleanup_num_heap_tuples in btvacuumcleanup(). This fixes the
issue because IndexVacuumInfo.num_heap_tuples (the source field) is
expected to accurately indicate the state of the table _after_ the
VACUUM completes inside btvacuumcleanup().
A backpatchable fix cannot easily be extracted from this commit. A
targeted fix for the issue will follow in a later commit, though that
won't happen today.
I (pgeoghegan) have chosen to remove any mention of deleted pages in the
documentation of the vacuum_cleanup_index_scale_factor GUC/param, since
the presence of deleted (though unrecycled) pages is no longer of much
concern to users. The vacuum_cleanup_index_scale_factor description in
the docs now seems rather unclear in any case, and it should probably be
rewritten in the near future. Perhaps some passing mention of page
deletion will be added back at the same time.
Bump XLOG_PAGE_MAGIC due to nbtree WAL records using full XIDs now.
Author: Peter Geoghegan <pg@bowt.ie>
Reviewed-By: Masahiko Sawada <sawada.mshk@gmail.com>
Discussion: https://postgr.es/m/CAH2-WznpdHvujGUwYZ8sihX=d5u-tRYhi-F4wnV2uN2zHpMUXw@mail.gmail.com
By default VACUUM will skip pages that it can't immediately get
exclusive access to, which means that even activities as harmless
and unpredictable as checkpoint buffer writes might prevent a page
from being processed. Ordinarily this is no big deal, but we have
a small number of test cases that examine the results of VACUUM's
processing and therefore will fail if the page of interest is skipped.
This seems to be the explanation for some rare buildfarm failures.
To fix, add the DISABLE_PAGE_SKIPPING option to the VACUUM commands
in tests where this could be an issue.
In passing, remove a duplicated query in pageinspect/sql/page.sql.
Back-patch as necessary (some of these cases are as old as v10).
Discussion: https://postgr.es/m/413923.1611006484@sss.pgh.pa.us
Block numbers are 32-bit unsigned integers. Therefore, the smallest
SQL integer type that they can fit in is bigint. However, in the
pageinspect module, most input and output parameters dealing with
block numbers were declared as int. The behavior with block numbers
larger than a signed 32-bit integer was therefore dubious. Change
these arguments to type bigint and add some more explicit error
checking on the block range.
(Other contrib modules appear to do this correctly already.)
Since we are changing argument types of existing functions, in order
to not misbehave if the binary is updated before the extension is
updated, we need to create new C symbols for the entry points, similar
to how it's done in other extensions as well.
Reported-by: Ashutosh Bapat <ashutosh.bapat.oss@gmail.com>
Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org>
Reviewed-by: Michael Paquier <michael@paquier.xyz>
Discussion: https://www.postgresql.org/message-id/flat/d8f6bdd536df403b9b33816e9f7e0b9d@G08CNEXMBPEKD05.g08.fujitsu.local
The newly-added gist pageinspect test prints the LSNs of GiST pages,
expecting them all to be 1 (GistBuildLSN). But with wal_level=minimal,
they got updated by the whole-relation WAL-logging at commit. Fix by
wrapping the problematic tests in the same transaction with the CREATE
INDEX.
Per buildfarm failure on thorntail.
Discussion: https://www.postgresql.org/message-id/3B4F97E5-40FB-4142-8CAA-B301CDFBF982%40iki.fi
1. The raw bytea representation of the point-type keys used in the test
depends on endianess. Remove the raw key_data column from the test.
2. The items stored on non-leftmost gist page depends on how many items
git on the other pages. This showed up as a failure on 32-bit i386
systems. To fix, only test the gist_page_items() function on the
leftmost leaf page.
Per Andrey Borodin and the buildfarm.
Discussion: https://www.postgresql.org/message-id/9FCEC1DC-86FB-4A57-88EF-DD13663B36AF%40yandex-team.ru
Add a new bt_metap() column to display the metapage's allequalimage
field. Also add three new columns to contrib/pageinspect's
bt_page_items() function:
* Add a boolean column ("dead") that displays the LP_DEAD bit value for
each non-pivot tuple.
* Add a TID column ("htid") that displays a single heap TID value for
each tuple. This is the TID that is returned by BTreeTupleGetHeapTID(),
so comparable values are shown for pivot tuples, plain non-pivot tuples,
and posting list tuples.
* Add a TID array column ("tids") that displays TIDs from each tuple's
posting list, if any. This works just like the "tids" column from
pageinspect's gin_leafpage_items() function.
No version bump for the pageinspect extension, since there hasn't been a
stable Postgres release since the last version bump (the last bump was
part of commit 58b4cb30).
Author: Peter Geoghegan
Discussion: https://postgr.es/m/CAH2-WzmSMmU2eNvY9+a4MNP+z02h6sa-uxZvN3un6jY02ZVBSw@mail.gmail.com
After more discussion, the new function added by ddbd5d8 could have been
designed in a better way. Based on an idea from Álvaro, instead of
returning one column which includes both the raw and combined flags, use
two columns, with one for the raw flags and one for the combined flags.
This also takes care of some issues with HEAP_LOCKED_UPGRADED and
HEAP_XMAX_IS_LOCKED_ONLY which are not really combined flags as they
depend on conditions defined by other raw bits, as mentioned by Amit.
While on it, fix an extra issue with combined flags. A combined flag
was returned if at least one of its bits was set, but all its bits need
to be set to include it in the result.
Author: Michael Paquier
Reviewed-by: Álvaro Herrera, Amit Kapila
Discussion: https://postgr.es/m/20190913114950.GA3824@alvherre.pgsql
Flags of t_infomask and t_infomask2 for each tuple are already included
in the information returned by heap_page_items as integers, and we
lacked a way to make that information human-readable.
Per discussion, the function includes an option which controls if
combined flags should be decomposed or not. The default is false, to
not decompose combined flags.
The module is bumped to version 1.8.
Author: Craig Ringer, Sawada Masahiko
Reviewed-by: Peter Geoghegan, Robert Haas, Álvaro Herrera, Moon Insung,
Amit Kapila, Michael Paquier, Tomas Vondra
Discussion: https://postgr.es/m/CAMsr+YEY7jeaXOb+oX+RhDyOFuTMdmHjGsBxL=igCm03J0go9Q@mail.gmail.com
This feature was using a process local map to track the first few blocks
in the relation. The map was reset each time we get the block with enough
freespace. It was discussed that it would be better to track this map on
a per-relation basis in relcache and then invalidate the same whenever
vacuum frees up some space in the page or when FSM is created. The new
design would be better both in terms of API design and performance.
List of commits reverted, in reverse chronological order:
06c8a5090e Improve code comments in b0eaa4c51b.
13e8643bfc During pg_upgrade, conditionally skip transfer of FSMs.
6f918159a9 Add more tests for FSM.
9c32e4c350 Clear the local map when not used.
29d108cdec Update the documentation for FSM behavior..
08ecdfe7e5 Make FSM test portable.
b0eaa4c51b Avoid creation of the free space map for small heap relations.
Discussion: https://postgr.es/m/20190416180452.3pm6uegx54iitbt5@alap3.anarazel.de
Make nbtree treat all index tuples as having a heap TID attribute.
Index searches can distinguish duplicates by heap TID, since heap TID is
always guaranteed to be unique. This general approach has numerous
benefits for performance, and is prerequisite to teaching VACUUM to
perform "retail index tuple deletion".
Naively adding a new attribute to every pivot tuple has unacceptable
overhead (it bloats internal pages), so suffix truncation of pivot
tuples is added. This will usually truncate away the "extra" heap TID
attribute from pivot tuples during a leaf page split, and may also
truncate away additional user attributes. This can increase fan-out,
especially in a multi-column index. Truncation can only occur at the
attribute granularity, which isn't particularly effective, but works
well enough for now. A future patch may add support for truncating
"within" text attributes by generating truncated key values using new
opclass infrastructure.
Only new indexes (BTREE_VERSION 4 indexes) will have insertions that
treat heap TID as a tiebreaker attribute, or will have pivot tuples
undergo suffix truncation during a leaf page split (on-disk
compatibility with versions 2 and 3 is preserved). Upgrades to version
4 cannot be performed on-the-fly, unlike upgrades from version 2 to
version 3. contrib/amcheck continues to work with version 2 and 3
indexes, while also enforcing stricter invariants when verifying version
4 indexes. These stricter invariants are the same invariants described
by "3.1.12 Sequencing" from the Lehman and Yao paper.
A later patch will enhance the logic used by nbtree to pick a split
point. This patch is likely to negatively impact performance without
smarter choices around the precise point to split leaf pages at. Making
these two mostly-distinct sets of enhancements into distinct commits
seems like it might clarify their design, even though neither commit is
particularly useful on its own.
The maximum allowed size of new tuples is reduced by an amount equal to
the space required to store an extra MAXALIGN()'d TID in a new high key
during leaf page splits. The user-facing definition of the "1/3 of a
page" restriction is already imprecise, and so does not need to be
revised. However, there should be a compatibility note in the v12
release notes.
Author: Peter Geoghegan
Reviewed-By: Heikki Linnakangas, Alexander Korotkov
Discussion: https://postgr.es/m/CAH2-WzkVb0Kom=R+88fDFb=JSxZMFvbHVC6Mn9LJ2n=X=kS-Uw@mail.gmail.com
In b0eaa4c51b, we allow FSM to be created only after 4 pages. One of the
tests check the FSM contents and to do that it populates many tuples in
the relation. The FSM contents depend on the availability of freespace in
the page and it could vary because of the alignment of tuples.
This commit removes the dependency on FSM contents.
Author: Amit Kapila
Discussion: https://postgr.es/m/CAA4eK1KADF6K1bagr0--mGv3dMcZ%3DH_Z-Qtvdfbp5PjaC6PJJA%40mail.gmail.com
Previously, all heaps had FSMs. For very small tables, this means that the
FSM took up more space than the heap did. This is wasteful, so now we
refrain from creating the FSM for heaps with 4 pages or fewer. If the last
known target block has insufficient space, we still try to insert into some
other page before giving up and extending the relation, since doing
otherwise leads to table bloat. Testing showed that trying every page
penalized performance slightly, so we compromise and try every other page.
This way, we visit at most two pages. Any pages with wasted free space
become visible at next relation extension, so we still control table bloat.
As a bonus, directly attempting one or two pages can even be faster than
consulting the FSM would have been.
Once the FSM is created for a heap we don't remove it even if somebody
deletes all the rows from the corresponding relation. We don't think it is
a useful optimization as it is quite likely that relation will again grow
to the same size.
Author: John Naylor, Amit Kapila
Reviewed-by: Amit Kapila
Tested-by: Mithun C Y
Discussion: https://www.postgresql.org/message-id/CAJVSVGWvB13PzpbLEecFuGFc5V2fsO736BsdTakPiPAcdMM5tQ@mail.gmail.com
Previously, all heaps had FSMs. For very small tables, this means that the
FSM took up more space than the heap did. This is wasteful, so now we
refrain from creating the FSM for heaps with 4 pages or fewer. If the last
known target block has insufficient space, we still try to insert into some
other page before giving up and extending the relation, since doing
otherwise leads to table bloat. Testing showed that trying every page
penalized performance slightly, so we compromise and try every other page.
This way, we visit at most two pages. Any pages with wasted free space
become visible at next relation extension, so we still control table bloat.
As a bonus, directly attempting one or two pages can even be faster than
consulting the FSM would have been.
Once the FSM is created for a heap we don't remove it even if somebody
deletes all the rows from the corresponding relation. We don't think it is
a useful optimization as it is quite likely that relation will again grow
to the same size.
Author: John Naylor with design inputs and some code contribution by Amit Kapila
Reviewed-by: Amit Kapila
Tested-by: Mithun C Y
Discussion: https://www.postgresql.org/message-id/CAJVSVGWvB13PzpbLEecFuGFc5V2fsO736BsdTakPiPAcdMM5tQ@mail.gmail.com
Commit 8b08f7d482 failed to update these modules to at least give
non-broken error messages for partitioned indexes. Add appropriate
error support to them.
Peter G. was complaining about a problem of unfriendly error messages;
while we haven't fixed that yet, subsequent discussion let to discovery
of these unhandled cases.
Author: Michaël Paquier
Reported-by: Peter Geoghegan
Discussion: https://postgr.es/m/CAH2-WzkOKptQiE51Bh4_xeEHhaBwHkZkGtKizrFMgEkfUuRRQg@mail.gmail.com
Vacuum of index consists from two stages: multiple (zero of more) ambulkdelete
calls and one amvacuumcleanup call. When workload on particular table
is append-only, then autovacuum isn't intended to touch this table. However,
user may run vacuum manually in order to fill visibility map and get benefits
of index-only scans. Then ambulkdelete wouldn't be called for indexes
of such table (because no heap tuples were deleted), only amvacuumcleanup would
be called In this case, amvacuumcleanup would perform full index scan for
two objectives: put recyclable pages into free space map and update index
statistics.
This patch allows btvacuumclanup to skip full index scan when two conditions
are satisfied: no pages are going to be put into free space map and index
statistics isn't stalled. In order to check first condition, we store
oldest btpo_xact in the meta-page. When it's precedes RecentGlobalXmin, then
there are some recyclable pages. In order to check second condition we store
number of heap tuples observed during previous full index scan by cleanup.
If fraction of newly inserted tuples is less than
vacuum_cleanup_index_scale_factor, then statistics isn't considered to be
stalled. vacuum_cleanup_index_scale_factor can be defined as both reloption and GUC (default).
This patch bumps B-tree meta-page version. Upgrade of meta-page is performed
"on the fly": during VACUUM meta-page is rewritten with new version. No special
handling in pg_upgrade is required.
Author: Masahiko Sawada, Alexander Korotkov
Review by: Peter Geoghegan, Kyotaro Horiguchi, Alexander Korotkov, Yura Sokolov
Discussion: https://www.postgresql.org/message-id/flat/CAD21AoAX+d2oD_nrd9O2YkpzHaFr=uQeGr9s1rKC3O4ENc568g@mail.gmail.com
Instead of using our standard macro for this calculation, this code
did it itself ... and got it wrong, leading to incorrect display of
the null bitmap in some cases. Noted and fixed by Maksim Milyutin.
In passing, remove a uselessly duplicative error check.
Errors were introduced in commit d6061f83a; back-patch to 9.6
where that came in.
Maksim Milyutin, reviewed by Andrey Borodin
Discussion: https://postgr.es/m/ec295792-a69f-350f-6287-25a20e8f31d5@gmail.com
Per a report from AP, it's not that hard to exhaust the supply of
bitmap pages if you create a table with a hash index and then insert a
few billion rows - and then you start getting errors when you try to
insert additional rows. In the particular case reported by AP,
there's another fix that we can make to improve recycling of overflow
pages, which is another way to avoid the error, but there may be other
cases where this problem happens and that fix won't help. So let's
buy ourselves as much headroom as we can without rearchitecting
anything.
The comments claim that the old limit was 64GB, but it was really
only 32GB, because we didn't use all the bits in the page for bitmap
bits - only the largest power of 2 that could fit after deducting
space for the page header and so forth. Thus, we have 4kB per page
for bitmap bits, not 8kB. The new limit is thus actually 8 times the
old *real* limit but only 4 times the old *purported* limit.
Since this breaks on-disk compatibility, bump HASH_VERSION. We've
already done this earlier in this release cycle, so this doesn't cause
any incremental inconvenience for people using pg_upgrade from
releases prior to v10. However, users who use pg_upgrade to reach
10beta3 or later from 10beta2 or earlier will need to REINDEX any hash
indexes again.
Amit Kapila and Robert Haas
Discussion: http://postgr.es/m/20170704105728.mwb72jebfmok2nm2@zip.com.au
Since hash indexes typically have very few overflow pages, adding a
new splitpoint essentially doubles the on-disk size of the index,
which can lead to large and abrupt increases in disk usage (and
perhaps long delays on occasion). To mitigate this problem to some
degree, divide larger splitpoints into four equal phases. This means
that, for example, instead of growing from 4GB to 8GB all at once, a
hash index will now grow from 4GB to 5GB to 6GB to 7GB to 8GB, which
is perhaps still not as smooth as we'd like but certainly an
improvement.
This changes the on-disk format of the metapage, so bump HASH_VERSION
from 2 to 3. This will force a REINDEX of all existing hash indexes,
but that's probably a good idea anyway. First, hash indexes from
pre-10 versions of PostgreSQL could easily be corrupted, and we don't
want to confuse corruption carried over from an older release with any
corruption caused despite the new write-ahead logging in v10. Second,
it will let us remove some backward-compatibility code added by commit
293e24e507.
Mithun Cy, reviewed by Amit Kapila, Jesper Pedersen and me. Regression
test outputs updated by me.
Discussion: http://postgr.es/m/CAD__OuhG6F1gQLCgMQNnMNgoCvOLQZz9zKYJQNYvYmmJoM42gA@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoYty0jCf-pa+m+vYUJ716+AxM7nv_syvyanyf5O-L_i2A@mail.gmail.com
The warning about hash indexes not being write-ahead logged and their
use being discouraged has been removed. "snapshot too old" is now
supported for tables with hash indexes. Most importantly, barring
bugs, hash indexes will now be crash-safe and usable on standbys.
This commit doesn't yet add WAL consistency checking for hash
indexes, as we now have for other index types; a separate patch has
been submitted to cure that lack.
Amit Kapila, reviewed and slightly modified by me. The larger patch
series of which this is a part has been reviewed and tested by Álvaro
Herrera, Ashutosh Sharma, Mark Kirkwood, Jeff Janes, and Jesper
Pedersen.
Discussion: http://postgr.es/m/CAA4eK1JOBX=YU33631Qh-XivYXtPSALh514+jR8XeD7v+K3r_Q@mail.gmail.com
The contrib extensions pageinspect, pg_visibility and pgstattuple only
work against regular relations which have storage. They don't work
against foreign tables, partitioned (parent) tables, views, et al.
Add checks to the user-callable functions to return a useful error
message to the user if they mistakenly pass an invalid relation to a
function which doesn't accept that kind of relation.
In passing, improve some of the existing checks to use ereport() instead
of elog(), add a function to consolidate common checks where
appropriate, and add some regression tests.
Author: Amit Langote, with various changes by me
Reviewed by: Michael Paquier and Corey Huinker
Discussion: https://postgr.es/m/ab91fd9d-4751-ee77-c87b-4dd704c1e59c@lab.ntt.co.jp
It did that to verify that the page was an overflow page rather than
anything else, but that means that checking the status of all the
overflow bits requires reading the entire index. So don't do that.
The new code validates that the page is not a primary bucket page
or bitmap page by looking at the metapage, so that using this on
large numbers of pages can be reasonably efficient.
Ashutosh Sharma, per a complaint from me, and with further
modifications by me.
This avoids a very significant amount of buffer manager traffic and
contention when scanning hash indexes, because it's no longer
necessary to lock and pin the metapage for every scan. We do need
some way of figuring out when the cache is too stale to use any more,
so that when we lock the primary bucket page to which the cached
metapage points us, we can tell whether a split has occurred since we
cached the metapage data. To do that, we use the hash_prevblkno field
in the primary bucket page, which would otherwise always be set to
InvalidBuffer.
This patch contains code so that it will continue working (although
less efficiently) with hash indexes built before this change, but
perhaps we should consider bumping the hash version and ripping out
the compatibility code. That decision can be made later, though.
Mithun Cy, reviewed by Jesper Pedersen, Amit Kapila, and by me.
Before committing, I made a number of cosmetic changes to the last
posted version of the patch, adjusted _hash_getcachedmetap to be more
careful about order of operation, and made some necessary updates to
the pageinspect documentation and regression tests.
Per a report from Tom Lane, the ffactor reported by hash_metapage_info
and the free_size reported by hash_page_stats vary by platform.
Ashutosh Sharma and Robert Haas
Patch by Jesper Pedersen and Ashutosh Sharma, with some error handling
improvements by me. Tests from Peter Eisentraut. Reviewed by Álvaro
Herrera, Michael Paquier, Jesper Pedersen, Jeff Janes, Peter
Eisentraut, Amit Kapila, Mithun Cy, and me.
Discussion: http://postgr.es/m/e2ac6c58-b93f-9dd9-f4e6-d6d30add7fdf@redhat.com
On closer inspection, commit 84ad68d64 broke gin_leafpage_items(),
because the aligned copy of the page got palloc'd in a short-lived
context whereas it needs to be in the SRF's multi_call_memory_ctx.
This was not exposed by the regression test, because the regression
test doesn't actually exercise the function in a meaningful way.
Fix the code bug, and extend the test in what I hope is a portable
fashion.