The logic introduced in commit b69bf30b9b
and repaired in commits 669c7d20e6 and
7be47c56af helps to ensure that we don't
overwrite old multixact member information while it is still needed,
but a user who creates many large multixacts can still exhaust the
member space (and thus start getting errors) while autovacuum stands
idly by.
To fix this, progressively ramp down the effective value (but not the
actual contents) of autovacuum_multixact_freeze_max_age as member space
utilization increases. This makes autovacuum more aggressive and also
reduces the threshold for a manual VACUUM to perform a full-table scan.
This patch leaves unsolved the problem of ensuring that emergency
autovacuums are triggered even when autovacuum=off. We'll need to fix
that via a separate patch.
Thomas Munro and Robert Haas
After a timeline switch, we would leave behind recycled WAL segments that
are in the future, but on the old timeline. After promotion, and after they
become old enough to be recycled again, we would notice that they don't have
a .ready or .done file, create a .ready file for them, and archive them.
That's bogus, because the files contain garbage, recycled from an older
timeline (or prealloced as zeros). We shouldn't archive such files.
This could happen when we're following a timeline switch during replay, or
when we switch to new timeline at end-of-recovery.
To fix, whenever we switch to a new timeline, scan the data directory for
WAL segments on the old timeline, but with a higher segment number, and
remove them. Those don't belong to our timeline history, and are most
likely bogus recycled or preallocated files. They could also be valid files
that we streamed from the primary ahead of time, but in any case, they're
not needed to recover to the new timeline.
CREATE DATABASE and ALTER DATABASE .. SET TABLESPACE copy the source
database directory on the filesystem level. To ensure the on disk
state is consistent they block out users of the affected database and
force a checkpoint to flush out all data to disk. Unfortunately, up to
now, that checkpoint didn't flush out dirty buffers from unlogged
relations.
That bug means there could be leftover dirty buffers in either the
template database, or the database in its old location. Leading to
problems when accessing relations in an inconsistent state; and to
possible problems during shutdown in the SET TABLESPACE case because
buffers belonging files that don't exist anymore are flushed.
This was reported in bug #10675 by Maxim Boguk.
Fix by Pavan Deolasee, modified somewhat by me. Reviewed by MauMau and
Fujii Masao.
Backpatch to 9.1 where unlogged tables were introduced.
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).
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.
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.
This was not changed in HEAD, but will be done later as part of a
pgindent run. Future pgindent runs will also do this.
Report by Tom Lane
Backpatch through all supported branches, but not HEAD
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.
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.
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.
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
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
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.
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
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
