I'm forking this thread from
https://postgr.es/m/flat/20190202083822.gc32...@gust.leadboat.com, which
reported a race condition involving the "apparent wraparound" safety check in
SimpleLruTruncate():
On Wed, Feb 13, 2019 at 11:26:23PM -0800, Noah Misch wrote:
> 1. The result of the test is valid only until we release the SLRU ControlLock,
> which we do before SlruScanDirCbDeleteCutoff() uses the cutoff to evaluate
> segments for deletion. Once we release that lock, latest_page_number can
> advance. This creates a TOCTOU race condition, allowing excess deletion:
>
> [local] test=# table trunc_clog_concurrency ;
> ERROR: could not access status of transaction 2149484247
> DETAIL: Could not open file "pg_xact/0801": No such file or directory.
> Fixes are available:
> b. Arrange so only one backend runs vac_truncate_clog() at a time. Other than
> AsyncCtl, every SLRU truncation appears in vac_truncate_clog(), in a
> checkpoint, or in the startup process. Hence, also arrange for only one
> backend to call SimpleLruTruncate(AsyncCtl) at a time.
More specifically, restrict vac_update_datfrozenxid() to one backend per
database, and restrict vac_truncate_clog() and asyncQueueAdvanceTail() to one
backend per cluster. This, attached, was rather straightforward.
I wonder about performance in a database with millions of small relations,
particularly considering my intent to back-patch this. In such databases,
vac_update_datfrozenxid() can be a major part of the VACUUM's cost. Two
things work in our favor. First, vac_update_datfrozenxid() runs once per
VACUUM command, not once per relation. Second, Autovacuum has this logic:
* ... we skip
* this if (1) we found no work to do and (2) we skipped at least one
* table due to concurrent autovacuum activity. In that case, the other
* worker has already done it, or will do so when it finishes.
*/
if (did_vacuum || !found_concurrent_worker)
vac_update_datfrozenxid();
That makes me relatively unworried. I did consider some alternatives:
- Run vac_update_datfrozenxid()'s pg_class scan before taking a lock. If we
find the need for pg_database updates, take the lock and scan pg_class again
to get final numbers. This doubles the work in cases that end up taking the
lock, so I'm not betting it being a net win.
- Use LockWaiterCount() to skip vac_update_datfrozenxid() if some other
backend is already waiting. This is similar to Autovacuum's
found_concurrent_worker test. It is tempting. I'm not proposing it,
because it changes the states possible when manual VACUUM completes. Today,
you can predict post-VACUUM datfrozenxid from post-VACUUM relfrozenxid
values. If manual VACUUM could skip vac_update_datfrozenxid() this way,
datfrozenxid could lag until some concurrent VACUUM finishes.
Thanks,
nm
commit 65928b9 (HEAD)
Author: Noah Misch <n...@leadboat.com>
AuthorDate: Sun Feb 17 23:26:45 2019 -0800
Commit: Noah Misch <n...@leadboat.com>
CommitDate: Sun Feb 17 23:26:45 2019 -0800
Prevent concurrent SimpleLruTruncate() for any given SLRU.
The SimpleLruTruncate() header comment states the new coding rule. This
closes race conditions that presented rare opportunities for data loss.
Symptoms and recommendations to users mirror the previous commit[1].
Back-patch to 9.4 (all supported versions).
Reviewed by TBD.
Discussion: https://postgr.es/m/TBD
[1] slru-truncate-modulo-v2.patch of
https://postgr.es/m/20190217040913.ga1305...@rfd.leadboat.com
diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 5c1408b..b7ddc73 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -849,7 +849,7 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss 11:34
0:00 postgres: ser
<tbody>
<row>
- <entry morerows="63"><literal>LWLock</literal></entry>
+ <entry morerows="65"><literal>LWLock</literal></entry>
<entry><literal>ShmemIndexLock</literal></entry>
<entry>Waiting to find or allocate space in shared memory.</entry>
</row>
@@ -1044,6 +1044,16 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss
11:34 0:00 postgres: ser
the oldest transaction id available to it.</entry>
</row>
<row>
+ <entry><literal>WrapLimitsVacuumLock</literal></entry>
+ <entry>Waiting to update limits on transaction id and multixact
+ consumption.</entry>
+ </row>
+ <row>
+ <entry><literal>AsyncQueueTailLock</literal></entry>
+ <entry>Waiting to update limit on notification message
+ storage.</entry>
+ </row>
+ <row>
<entry><literal>clog</literal></entry>
<entry>Waiting for I/O on a clog (transaction status) buffer.</entry>
</row>
@@ -1128,7 +1138,7 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss
11:34 0:00 postgres: ser
counters during Parallel Hash plan execution.</entry>
</row>
<row>
- <entry morerows="9"><literal>Lock</literal></entry>
+ <entry morerows="10"><literal>Lock</literal></entry>
<entry><literal>relation</literal></entry>
<entry>Waiting to acquire a lock on a relation.</entry>
</row>
@@ -1137,6 +1147,12 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss
11:34 0:00 postgres: ser
<entry>Waiting to extend a relation.</entry>
</row>
<row>
+ <entry><literal>frozen IDs</literal></entry>
+ <entry>Waiting to
+ update
<structname>pg_database</structname>.<structfield>datfrozenxid</structfield>
+ and
<structname>pg_database</structname>.<structfield>datminmxid</structfield>.</entry>
+ </row>
+ <row>
<entry><literal>page</literal></entry>
<entry>Waiting to acquire a lock on page of a relation.</entry>
</row>
diff --git a/src/backend/access/transam/slru.c
b/src/backend/access/transam/slru.c
index 3623352..8c9d5d8 100644
--- a/src/backend/access/transam/slru.c
+++ b/src/backend/access/transam/slru.c
@@ -1163,6 +1163,14 @@ SimpleLruFlush(SlruCtl ctl, bool allow_redirtied)
/*
* Remove all segments before the one holding the passed page number
+ *
+ * All SLRUs prevent concurrent calls to this function, either with an LWLock
+ * or by calling it only as part of a checkpoint. Mutual exclusion must begin
+ * before computing cutoffPage. Mutual exclusion must end after any limit
+ * update that would permit other backends to write fresh data into the
+ * segment immediately preceding the one containing cutoffPage. Otherwise,
+ * when the SLRU is quite full, SimpleLruTruncate() might delete that segment
+ * after it has accrued freshly-written data.
*/
void
SimpleLruTruncate(SlruCtl ctl, int cutoffPage)
diff --git a/src/backend/access/transam/subtrans.c
b/src/backend/access/transam/subtrans.c
index cbc6129..a480dd5 100644
--- a/src/backend/access/transam/subtrans.c
+++ b/src/backend/access/transam/subtrans.c
@@ -347,8 +347,8 @@ ExtendSUBTRANS(TransactionId newestXact)
/*
* Remove all SUBTRANS segments before the one holding the passed transaction
ID
*
- * This is normally called during checkpoint, with oldestXact being the
- * oldest TransactionXmin of any running transaction.
+ * oldestXact is the oldest TransactionXmin of any running transaction. This
+ * is called only during checkpoint.
*/
void
TruncateSUBTRANS(TransactionId oldestXact)
diff --git a/src/backend/commands/async.c b/src/backend/commands/async.c
index 5a7ee0d..66e888c 100644
--- a/src/backend/commands/async.c
+++ b/src/backend/commands/async.c
@@ -223,19 +223,22 @@ typedef struct QueueBackendStatus
/*
* Shared memory state for LISTEN/NOTIFY (excluding its SLRU stuff)
*
- * The AsyncQueueControl structure is protected by the AsyncQueueLock.
+ * The AsyncQueueControl structure is protected by the AsyncQueueLock and
+ * AsyncQueueTailLock.
*
- * When holding the lock in SHARED mode, backends may only inspect their own
- * entries as well as the head and tail pointers. Consequently we can allow a
- * backend to update its own record while holding only SHARED lock (since no
- * other backend will inspect it).
+ * When holding AsyncQueueLock in SHARED mode, backends may only inspect their
+ * own entries as well as the head and tail pointers. Consequently we can
+ * allow a backend to update its own record while holding only SHARED lock
+ * (since no other backend will inspect it).
*
- * When holding the lock in EXCLUSIVE mode, backends can inspect the entries
- * of other backends and also change the head and tail pointers.
+ * When holding AsyncQueueLock in EXCLUSIVE mode, backends can inspect the
+ * entries of other backends and also change the head pointer. When holding
+ * both AsyncQueueLock and AsyncQueueTailLock in EXCLUSIVE mode, backends can
+ * change the tail pointer.
*
* AsyncCtlLock is used as the control lock for the pg_notify SLRU buffers.
- * In order to avoid deadlocks, whenever we need both locks, we always first
- * get AsyncQueueLock and then AsyncCtlLock.
+ * In order to avoid deadlocks, whenever we need multiple locks, we first get
+ * AsyncQueueTailLock, then AsyncQueueLock, and lastly AsyncCtlLock.
*
* Each backend uses the backend[] array entry with index equal to its
* BackendId (which can range from 1 to MaxBackends). We rely on this to make
@@ -2010,6 +2013,10 @@ asyncQueueAdvanceTail(void)
int newtailpage;
int boundary;
+ /* Restrict task to one backend per cluster; see SimpleLruTruncate(). */
+ LWLockAcquire(AsyncQueueTailLock, LW_EXCLUSIVE);
+
+ /* Compute the new tail. */
LWLockAcquire(AsyncQueueLock, LW_EXCLUSIVE);
min = QUEUE_HEAD;
for (i = 1; i <= MaxBackends; i++)
@@ -2018,7 +2025,6 @@ asyncQueueAdvanceTail(void)
min = QUEUE_POS_MIN(min, QUEUE_BACKEND_POS(i));
}
oldtailpage = QUEUE_POS_PAGE(QUEUE_TAIL);
- QUEUE_TAIL = min;
LWLockRelease(AsyncQueueLock);
/*
@@ -2038,6 +2044,17 @@ asyncQueueAdvanceTail(void)
*/
SimpleLruTruncate(AsyncCtl, newtailpage);
}
+
+ /*
+ * Advertise the new tail. This changes asyncQueueIsFull()'s verdict
for
+ * the segment immediately prior to the new tail, allowing fresh data
into
+ * that segment.
+ */
+ LWLockAcquire(AsyncQueueLock, LW_EXCLUSIVE);
+ QUEUE_TAIL = min;
+ LWLockRelease(AsyncQueueLock);
+
+ LWLockRelease(AsyncQueueTailLock);
}
/*
diff --git a/src/backend/commands/vacuum.c b/src/backend/commands/vacuum.c
index e91df21..326f48b 100644
--- a/src/backend/commands/vacuum.c
+++ b/src/backend/commands/vacuum.c
@@ -1225,6 +1225,14 @@ vac_update_datfrozenxid(void)
bool dirty = false;
/*
+ * Restrict this task to one backend per database. This avoids race
+ * conditions that would move datfrozenxid or datminmxid backward or
call
+ * vac_truncate_clog() with a datfrozenxid preceding a datfrozenxid
passed
+ * to an earlier vac_truncate_clog() call.
+ */
+ LockDatabaseFrozenIds(ExclusiveLock);
+
+ /*
* Initialize the "min" calculation with GetOldestXmin, which is a
* reasonable approximation to the minimum relfrozenxid for not-yet-
* committed pg_class entries for new tables; see AddNewRelationTuple().
@@ -1388,6 +1396,9 @@ vac_truncate_clog(TransactionId frozenXID,
bool bogus = false;
bool frozenAlreadyWrapped = false;
+ /* Restrict task to one backend per cluster; see SimpleLruTruncate(). */
+ LWLockAcquire(WrapLimitsVacuumLock, LW_EXCLUSIVE);
+
/* init oldest datoids to sync with my frozenXID/minMulti values */
oldestxid_datoid = MyDatabaseId;
minmulti_datoid = MyDatabaseId;
@@ -1497,6 +1508,8 @@ vac_truncate_clog(TransactionId frozenXID,
*/
SetTransactionIdLimit(frozenXID, oldestxid_datoid);
SetMultiXactIdLimit(minMulti, minmulti_datoid, false);
+
+ LWLockRelease(WrapLimitsVacuumLock);
}
diff --git a/src/backend/storage/lmgr/lmgr.c b/src/backend/storage/lmgr/lmgr.c
index e688ba8..4873835 100644
--- a/src/backend/storage/lmgr/lmgr.c
+++ b/src/backend/storage/lmgr/lmgr.c
@@ -458,6 +458,21 @@ UnlockRelationForExtension(Relation relation, LOCKMODE
lockmode)
}
/*
+ * LockDatabaseFrozenIds
+ *
+ * This allows one backend per database to execute vac_update_datfrozenxid().
+ */
+void
+LockDatabaseFrozenIds(LOCKMODE lockmode)
+{
+ LOCKTAG tag;
+
+ SET_LOCKTAG_DATABASE_FROZEN_IDS(tag, MyDatabaseId);
+
+ (void) LockAcquire(&tag, lockmode, false, false);
+}
+
+/*
* LockPage
*
* Obtain a page-level lock. This is currently used by some index access
@@ -1060,6 +1075,11 @@ DescribeLockTag(StringInfo buf, const LOCKTAG *tag)
tag->locktag_field2,
tag->locktag_field1);
break;
+ case LOCKTAG_DATABASE_FROZEN_IDS:
+ appendStringInfo(buf,
+
_("pg_database.datfrozenxid of database %u"),
+ tag->locktag_field1);
+ break;
case LOCKTAG_PAGE:
appendStringInfo(buf,
_("page %u of relation
%u of database %u"),
diff --git a/src/backend/storage/lmgr/lwlocknames.txt
b/src/backend/storage/lmgr/lwlocknames.txt
index db47843..d517023 100644
--- a/src/backend/storage/lmgr/lwlocknames.txt
+++ b/src/backend/storage/lmgr/lwlocknames.txt
@@ -49,3 +49,5 @@ MultiXactTruncationLock 41
OldSnapshotTimeMapLock 42
LogicalRepWorkerLock 43
CLogTruncationLock 44
+WrapLimitsVacuumLock 45
+AsyncQueueTailLock 46
diff --git a/src/backend/utils/adt/lockfuncs.c
b/src/backend/utils/adt/lockfuncs.c
index ffd1970..cbf0cab 100644
--- a/src/backend/utils/adt/lockfuncs.c
+++ b/src/backend/utils/adt/lockfuncs.c
@@ -26,6 +26,7 @@
const char *const LockTagTypeNames[] = {
"relation",
"extend",
+ "frozen IDs",
"page",
"tuple",
"transactionid",
@@ -245,6 +246,17 @@ pg_lock_status(PG_FUNCTION_ARGS)
nulls[8] = true;
nulls[9] = true;
break;
+ case LOCKTAG_DATABASE_FROZEN_IDS:
+ values[1] =
ObjectIdGetDatum(instance->locktag.locktag_field1);
+ nulls[2] = true;
+ nulls[3] = true;
+ nulls[4] = true;
+ nulls[5] = true;
+ nulls[6] = true;
+ nulls[7] = true;
+ nulls[8] = true;
+ nulls[9] = true;
+ break;
case LOCKTAG_PAGE:
values[1] =
ObjectIdGetDatum(instance->locktag.locktag_field1);
values[2] =
ObjectIdGetDatum(instance->locktag.locktag_field2);
diff --git a/src/include/storage/lmgr.h b/src/include/storage/lmgr.h
index 3d705fa..f113f55 100644
--- a/src/include/storage/lmgr.h
+++ b/src/include/storage/lmgr.h
@@ -59,6 +59,9 @@ extern bool ConditionalLockRelationForExtension(Relation
relation,
LOCKMODE lockmode);
extern int RelationExtensionLockWaiterCount(Relation relation);
+/* Lock to recompute pg_database.datfrozenxid in the current database */
+extern void LockDatabaseFrozenIds(LOCKMODE lockmode);
+
/* Lock a page (currently only used within indexes) */
extern void LockPage(Relation relation, BlockNumber blkno, LOCKMODE lockmode);
extern bool ConditionalLockPage(Relation relation, BlockNumber blkno, LOCKMODE
lockmode);
diff --git a/src/include/storage/lock.h b/src/include/storage/lock.h
index 16b927c..f3b6ef4 100644
--- a/src/include/storage/lock.h
+++ b/src/include/storage/lock.h
@@ -141,6 +141,8 @@ typedef enum LockTagType
/* ID info for a relation is DB OID + REL OID; DB OID = 0 if shared */
LOCKTAG_RELATION_EXTEND, /* the right to extend a relation */
/* same ID info as RELATION */
+ LOCKTAG_DATABASE_FROZEN_IDS, /* pg_database.datfrozenxid */
+ /* ID info for frozen IDs is DB OID */
LOCKTAG_PAGE, /* one page of a relation */
/* ID info for a page is RELATION info + BlockNumber */
LOCKTAG_TUPLE, /* one physical tuple */
@@ -206,6 +208,14 @@ typedef struct LOCKTAG
(locktag).locktag_type = LOCKTAG_RELATION_EXTEND, \
(locktag).locktag_lockmethodid = DEFAULT_LOCKMETHOD)
+#define SET_LOCKTAG_DATABASE_FROZEN_IDS(locktag,dboid) \
+ ((locktag).locktag_field1 = (dboid), \
+ (locktag).locktag_field2 = 0, \
+ (locktag).locktag_field3 = 0, \
+ (locktag).locktag_field4 = 0, \
+ (locktag).locktag_type = LOCKTAG_DATABASE_FROZEN_IDS, \
+ (locktag).locktag_lockmethodid = DEFAULT_LOCKMETHOD)
+
#define SET_LOCKTAG_PAGE(locktag,dboid,reloid,blocknum) \
((locktag).locktag_field1 = (dboid), \
(locktag).locktag_field2 = (reloid), \
