I've been saying we needed to do this for more than five years, and here it
finally is. This patch removes the ever-growing tangle of spaghetti logic
that grouping_planner() used to use to try to identify the best plan for
post-scan/join query steps. Now, there is (nearly) independent
consideration of each execution step, and entirely separate construction of
Paths to represent each of the possible ways to do that step. We choose
the best Path or set of Paths using the same add_path() logic that's been
used inside query_planner() for years.
In addition, this patch removes the old restriction that subquery_planner()
could return only a single Plan. It now returns a RelOptInfo containing a
set of Paths, just as query_planner() does, and the parent query level can
use each of those Paths as the basis of a SubqueryScanPath at its level.
This allows finding some optimizations that we missed before, wherein a
subquery was capable of returning presorted data and thereby avoiding a
sort in the parent level, making the overall cost cheaper even though
delivering sorted output was not the cheapest plan for the subquery in
isolation. (A couple of regression test outputs change in consequence of
that. However, there is very little change in visible planner behavior
overall, because the point of this patch is not to get immediate planning
benefits but to create the infrastructure for future improvements.)
There is a great deal left to do here. This patch unblocks a lot of
planner work that was basically impractical in the old code structure,
such as allowing FDWs to implement remote aggregation, or rewriting
plan_set_operations() to allow consideration of multiple implementation
orders for set operations. (The latter will likely require a full
rewrite of plan_set_operations(); what I've done here is only to fix it
to return Paths not Plans.) I have also left unfinished some localized
refactoring in createplan.c and planner.c, because it was not necessary
to get this patch to a working state.
Thanks to Robert Haas, David Rowley, and Amit Kapila for review.
Add four new SQL accessible functions: pg_control_system(),
pg_control_checkpoint(), pg_control_recovery(), and pg_control_init()
which expose a subset of the control file data.
Along the way move the code to read and validate the control file to
src/common, where it can be shared by the new backend functions
and the original pg_controldata frontend program.
Patch by me, significant input, testing, and review by Michael Paquier.
Previously recovery_min_apply_delay was applied even before recovery
had reached consistency. This could cause us to wait a long time
unexpectedly for read-only connections to be allowed. It's problematic
because the standby was useless during that wait time.
This patch changes recovery_min_apply_delay so that it's applied once
the database has reached the consistent state. That is, even if the delay
is set, the standby tries to replay WAL records as fast as possible until
it has reached consistency.
Author: Michael Paquier
Reviewed-By: Julien Rouhaud
Reported-By: Greg Clough
Backpatch: 9.4, where recovery_min_apply_delay was added
Bug: #13770
Discussion: http://www.postgresql.org/message-id/20151111155006.2644.84564@wrigleys.postgresql.org
Now it's possible to load recent version of Hunspell for several languages.
To handle these dictionaries Hunspell patch adds support for:
* FLAG long - sets the double extended ASCII character flag type
* FLAG num - sets the decimal number flag type (from 1 to 65535)
* AF parameter - alias for flag's set
Also it moves test dictionaries into separate directory.
Author: Artur Zakirov with editorization by me
The new bit indicates whether every tuple on the page is already frozen.
It is cleared only when the all-visible bit is cleared, and it can be
set only when we vacuum a page and find that every tuple on that page is
both visible to every transaction and in no need of any future
vacuuming.
A future commit will use this new bit to optimize away full-table scans
that would otherwise be triggered by XID wraparound considerations. A
page which is merely all-visible must still be scanned in that case, but
a page which is all-frozen need not be. This commit does not attempt
that optimization, although that optimization is the goal here. It
seems better to get the basic infrastructure in place first.
Per discussion, it's very desirable for pg_upgrade to automatically
migrate existing VM forks from the old format to the new format. That,
too, will be handled in a follow-on patch.
Masahiko Sawada, reviewed by Kyotaro Horiguchi, Fujii Masao, Amit
Kapila, Simon Riggs, Andres Freund, and others, and substantially
revised by me.
Fabien Coelho, reviewed mostly by Michael Paquier and me, but also by
Heikki Linnakangas, BeomYong Lee, Kyotaro Horiguchi, Oleksander
Shulgin, and Álvaro Herrera.
This is basically a bug fix; the old code assumes that a ForeignScan
is always parallel-safe, but for postgres_fdw, for example, this is
definitely false. It should be true for file_fdw, though, since a
worker can read a file from the filesystem just as well as any other
backend process.
Original patch by Thomas Munro. Documentation, and changes to the
comments, by me.
This patch introduces "pg_blocking_pids(int) returns int[]", which returns
the PIDs of any sessions that are blocking the session with the given PID.
Historically people have obtained such information using a self-join on
the pg_locks view, but it's unreasonably tedious to do it that way with any
modicum of correctness, and the addition of parallel queries has pretty
much broken that approach altogether. (Given some more columns in the view
than there are today, you could imagine handling parallel-query cases with
a 4-way join; but ugh.)
The new function has the following behaviors that are painful or impossible
to get right via pg_locks:
1. Correctly understands which lock modes block which other ones.
2. In soft-block situations (two processes both waiting for conflicting lock
modes), only the one that's in front in the wait queue is reported to
block the other.
3. In parallel-query cases, reports all sessions blocking any member of
the given PID's lock group, and reports a session by naming its leader
process's PID, which will be the pg_backend_pid() value visible to
clients.
The motivation for doing this right now is mostly to fix the isolation
tests. Commit 38f8bdcac4 lobotomized
isolationtester's is-it-waiting query by removing its ability to recognize
nonconflicting lock modes, as a crude workaround for the inability to
handle soft-block situations properly. But even without the lock mode
tests, the old query was excessively slow, particularly in
CLOBBER_CACHE_ALWAYS builds; some of our buildfarm animals fail the new
deadlock-hard test because the deadlock timeout elapses before they can
probe the waiting status of all eight sessions. Replacing the pg_locks
self-join with use of pg_blocking_pids() is not only much more correct, but
a lot faster: I measure it at about 9X faster in a typical dev build with
Asserts, and 3X faster in CLOBBER_CACHE_ALWAYS builds. That should provide
enough headroom for the slower CLOBBER_CACHE_ALWAYS animals to pass the
test, without having to lengthen deadlock_timeout yet more and thus slow
down the test for everyone else.
The "Session Information Functions" table seems to be sorted mostly
alphabetically (although it's not perfect), which would be all right
if it didn't lead to some related functions being described in a
pretty nonintuitive order. Also, the prose discussions after the table
were in an order that hardly matched the table at all. Rearrange to
make things a bit easier to follow.
This will parse strings in the format produced by pg_size_pretty() and
return sizes in bytes. This allows queries to be written with clauses
like "pg_total_relation_size(oid) > pg_size_bytes('10 GB')".
Author: Pavel Stehule with various improvements by Vitaly Burovoy
Discussion: http://www.postgresql.org/message-id/CAFj8pRD-tGoDKnxdYgECzA4On01_uRqPrwF-8LdkSE-6bDHp0w@mail.gmail.com
Reviewed-by: Vitaly Burovoy, Oleksandr Shulgin, Kyotaro Horiguchi,
Michael Paquier and Robert Haas
Up to now, there's been an assumption that all Paths for a given relation
compute the same output column set (targetlist). However, there are good
reasons to remove that assumption. For example, an indexscan on an
expression index might be able to return the value of an expensive function
"for free". While we have the ability to generate such a plan today in
simple cases, we don't have a way to model that it's cheaper than a plan
that computes the function from scratch, nor a way to create such a plan
in join cases (where the function computation would normally happen at
the topmost join node). Also, we need this so that we can have Paths
representing post-scan/join steps, where the targetlist may well change
from one step to the next. Therefore, invent a "struct PathTarget"
representing the columns we expect a plan step to emit. It's convenient
to include the output tuple width and tlist evaluation cost in this struct,
and there will likely be additional fields in future.
While Path nodes that actually do have custom outputs will need their own
PathTargets, it will still be true that most Paths for a given relation
will compute the same tlist. To reduce the overhead added by this patch,
keep a "default PathTarget" in RelOptInfo, and allow Paths that compute
that column set to just point to their parent RelOptInfo's reltarget.
(In the patch as committed, actually every Path is like that, since we
do not yet have any cases of custom PathTargets.)
I took this opportunity to provide some more-honest costing of
PlaceHolderVar evaluation. Up to now, the assumption that "scan/join
reltargetlists have cost zero" was applied not only to Vars, where it's
reasonable, but also PlaceHolderVars where it isn't. Now, we add the eval
cost of a PlaceHolderVar's expression to the first plan level where it can
be computed, by including it in the PathTarget cost field and adding that
to the cost estimates for Paths. This isn't perfect yet but it's much
better than before, and there is a way forward to improve it more. This
costing change affects the join order chosen for a couple of the regression
tests, changing expected row ordering.
Dead or half-dead index leaf pages were incorrectly reported as live, as a
consequence of a code rearrangement I made (during a moment of severe brain
fade, evidently) in commit d287818eb5.
The index metapage was not counted in index_size, causing that result to
not agree with the actual index size on-disk.
Index root pages were not counted in internal_pages, which is inconsistent
compared to the case of a root that's also a leaf (one-page index), where
the root would be counted in leaf_pages. Aside from that inconsistency,
this could lead to additional transient discrepancies between the reported
page counts and index_size, since it's possible for pgstatindex's scan to
see zero or multiple pages marked as BTP_ROOT, if the root moves due to
a split during the scan. With these fixes, index_size will always be
exactly one page more than the sum of the displayed page counts.
Also, the index_size result was incorrectly documented as being measured in
pages; it's always been measured in bytes. (While fixing that, I couldn't
resist doing some small additional wordsmithing on the pgstattuple docs.)
Including the metapage causes the reported index_size to not be zero for
an empty index. To preserve the desired property that the pgstattuple
regression test results are platform-independent (ie, BLCKSZ configuration
independent), scale the index_size result in the regression tests.
The documentation issue was reported by Otsuka Kenji, and the inconsistent
root page counting by Peter Geoghegan; the other problems noted by me.
Back-patch to all supported branches, because this has been broken for
a long time.
Move and refactor the underlying code for the pg_config client
application to src/common in support of sharing it with a new
system information SRF called pg_config() which makes the same
information available via SQL. Additionally wrap the SRF with a
new system view, as called pg_config.
Patch by me with extensive input and review by Michael Paquier
and additional review by Alvaro Herrera.
Clarify the description of which transactions will block a CREATE INDEX
CONCURRENTLY command from proceeding, and mention that the index might
still not be usable after CREATE INDEX completes. (This happens if the
index build detected broken HOT chains, so that pg_index.indcheckxmin gets
set, and there are open old transactions preventing the xmin horizon from
advancing past the index's initial creation. I didn't want to explain what
broken HOT chains are, though, so I omitted an explanation of exactly when
old transactions prevent the index from being used.)
Per discussion with Chris Travers. Back-patch to all supported branches,
since the same text appears in all of them.
In runtime.sgml, the old formulas for calculating the reasonable
values of SEMMNI and SEMMNS were incorrect. They have forgotten to
count the number of semaphores which both the checkpointer process
(introduced in 9.2) and the background worker processes (introduced
in 9.3) need.
This commit fixes those formulas so that they count the number of
semaphores which the checkpointer process and the background worker
processes need.
Report and patch by Kyotaro Horiguchi. Only the patch for 9.3 was
modified by me. Back-patch to 9.2 where the checkpointer process was
added and the number of needed semaphores was increased.
Author: Kyotaro Horiguchi
Reviewed-by: Fujii Masao
Backpatch: 9.2
Discussion: http://www.postgresql.org/message-id/20160203.125119.66820697.horiguchi.kyotaro@lab.ntt.co.jp
Commit 4de82f7d7 increased the WAL flush rate, mainly to increase the
likelihood that hint bits can be set quickly. More quickly set hint bits
can reduce contention around the clog et al. But unfortunately the
increased flush rate can have a significant negative performance impact,
I have measured up to a factor of ~4. The reason for this slowdown is
that if there are independent writes to the underlying devices, for
example because shared buffers is a lot smaller than the hot data set,
or because a checkpoint is ongoing, the fdatasync() calls force cache
flushes to be emitted to the storage.
This is achieved by flushing WAL only if the last flush was longer than
wal_writer_delay ago, or if more than wal_writer_flush_after (new GUC)
unflushed blocks are pending. Based on some tests the default for
wal_writer_delay is 1MB, which seems to work well both on SSD and
rotational media.
To avoid negative performance impact due to 4de82f7d7 an earlier
commit (db76b1e) made SetHintBits() more likely to succeed; preventing
performance regressions in the pgbench tests I performed.
Discussion: 20160118163908.GW10941@awork2.anarazel.de
Many automated test suites call pg_ctl. Buildfarm members axolotl,
hornet, mandrill, shearwater, sungazer and tern have failed when server
shutdown took longer than the pg_ctl default 60s timeout. This addition
permits slow hosts to easily raise the timeout without us editing a
--timeout argument into every test suite pg_ctl call. Back-patch to 9.1
(all supported versions) for the sake of automated testing.
Reviewed by Tom Lane.
If we've got a relatively straightforward join between two tables,
this pushes that join down to the remote server instead of fetching
the rows for each table and performing the join locally. Some cases
are not handled yet, such as SEMI and ANTI joins. Also, we don't
yet attempt to create presorted join paths or parameterized join
paths even though these options do get tried for a base relation
scan. Nevertheless, this seems likely to be a very significant win
in many practical cases.
Shigeru Hanada and Ashutosh Bapat, reviewed by Robert Haas, with
additional review at various points by Tom Lane, Etsuro Fujita,
KaiGai Kohei, and Jeevan Chalke.
Get rid of the false implication that PRIMARY KEY is exactly equivalent to
UNIQUE + NOT NULL. That was more-or-less true at one time in our
implementation, but the standard doesn't say that, and we've grown various
features (many of them required by spec) that treat a pkey differently from
less-formal constraints. Per recent discussion on pgsql-general.
I failed to resist the temptation to do some other wordsmithing in the
same area.
When force_parallel_mode = true, we enable the parallel mode restrictions
for all queries for which this is believed to be safe. For the subset of
those queries believed to be safe to run entirely within a worker, we spin
up a worker and run the query there instead of running it in the
original process. When force_parallel_mode = regress, make additional
changes to allow the regression tests to run cleanly even though parallel
workers have been injected under the hood.
Taken together, this facilitates both better user testing and better
regression testing of the parallelism code.
Robert Haas, with help from Amit Kapila and Rushabh Lathia.
Commit a104a017fc has this check because
I added it to the submitted patch before commit, but that was entirely
wrongheaded, as explained to me by Ashutosh Bapat, who also wrote this
patch.
An example use-case is "CHECK(num_nonnulls(a,b,c) = 1)" to assert that
exactly one of a,b,c isn't NULL. The functions are variadic, so they
can also be pressed into service to count the number of null or nonnull
elements in an array.
Marko Tiikkaja, reviewed by Pavel Stehule
GetExistingLocalJoinPath() is useful for handling EvalPlanQual rechecks
properly, and GetUserMappingById() is needed to make sure you're using
the right credentials.
Shigeru Hanada, Etsuro Fujita, Ashutosh Bapat, Robert Haas
The previous RequestAddinLWLocks() method had several disadvantages.
First, the locks would be in the main tranche; we've recently decided
that it's useful for LWLocks used for separate purposes to have
separate tranche IDs. Second, there wasn't any correlation between
what code called RequestAddinLWLocks() and what code called
LWLockAssign(); when multiple modules are in use, it could become
quite difficult to troubleshoot problems where LWLockAssign() ran out
of locks. To fix, create a concept of named LWLock tranches which
can be used either by extension or by core code.
Amit Kapila and Robert Haas
Since there currently is only one possible parenthesized option, namely
VERBOSE, it's a bit pointless to show it with "{ } [, ... ]". The curly
braces are useless and therefore confusing, as seen in a recent question
from Karsten Hilbert. Remove the extra decoration for the time being;
we can put it back when and if REINDEX grows some more options.
Have varlena.c expose an interface that allows the char(n), bytea, and
bpchar types to piggyback on a now-generalized SortSupport for text.
This pushes a little more knowledge of the bpchar/char(n) type into
varlena.c than might be preferred, but that seems like the approach
that creates least friction. Also speed things up for index builds
that use text_pattern_ops or varchar_pattern_ops.
This patch does quite a bit of renaming, but it seems likely to be
worth it, so as to avoid future confusion about the fact that this code
is now more generally used than the old names might have suggested.
Peter Geoghegan, reviewed by Álvaro Herrera and Andreas Karlsson,
with small tweaks by me.
This patch doesn't put the new infrastructure to use anywhere, and
indeed it's not clear how it could ever be used for something like
postgres_fdw which has to send an SQL query and wait for a reply,
but there might be FDWs or custom scan providers that are CPU-bound,
so let's give them a way to join club parallel.
KaiGai Kohei, reviewed by me.
You can't really do anything useful with this in the form it currently
exists; among other problems, there's no way to reread whatever
information might be produced when the path is output. Work is
underway to replace this with a more useful and more general system of
extensible nodes, but let's start by getting rid of this bit.
Extracted from a larger patch by KaiGai Kohei.
The default fetch size of 100 rows might not be right in every
environment, so allow users to configure it.
Corey Huinker, reviewed by Kyotaro Horiguchi, Andres Freund, and me.
Insert sd_notify() calls at server start and stop for integration with
systemd. This allows the use of systemd service units of type "notify",
which greatly simplifies the systemd configuration.
Reviewed-by: Pavel Stěhule <pavel.stehule@gmail.com>
Provide per-script statistical info (count of transactions executed
under that script, average latency for the whole script) after a
multi-script run, adding an intermediate level of detail to existing
global stats and per-command stats.
Author: Fabien Coelho
Reviewer: Michaël Paquier, Álvaro Herrera