rdelaage
/
postgresql
mirror of https://git.postgresql.org/git/postgresql.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Update obsolete multixact.c comments. 

Commit 4f627f89 switched SLRU truncation for multixacts back to being a
task performed during VACUUM, but missed some comments that continued to
reference truncation happening as part of checkpointing.  Update those
comments now.

Also update comments that became obsolete when commit c3ffa731 changed
the way that vacuum_multixact_freeze_min_age is applied by VACUUM as it
computes its MultiXactCutoff cutoff (which is used by VACUUM to decide
what to freeze).  Explain the same issues by referencing how OldestMxact
is the latest valid value that relminmxid can ever be advanced to at the
end of a VACUUM (following the work in commit 0b018fab).
This commit is contained in:
Peter Geoghegan 2023-01-03 16:54:35 -08:00
parent 5f53b42cfd
commit 5212d447fa
1 changed files with 19 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
src/backend/access/transam/multixact.c
View File

@ -265,18 +265,20 @@ typedef struct MultiXactStateData
* *
* The oldest valid value among all of the OldestMemberMXactId[] and * The oldest valid value among all of the OldestMemberMXactId[] and
* OldestVisibleMXactId[] entries is considered by vacuum as the earliest * OldestVisibleMXactId[] entries is considered by vacuum as the earliest
* possible value still having any live member transaction. Subtracting * possible value still having any live member transaction -- OldestMxact.
* vacuum_multixact_freeze_min_age from that value we obtain the freezing * Any value older than that is typically removed from tuple headers, or
* point for multixacts for that table. Any value older than that is * "frozen" via being replaced with a new xmax. VACUUM can sometimes even
* removed from tuple headers (or "frozen"; see FreezeMultiXactId. Note * remove an individual MultiXact xmax whose value is >= its OldestMxact
* that multis that have member xids that are older than the cutoff point * cutoff, though typically only when no individual member XID is still
* for xids must also be frozen, even if the multis themselves are newer * running. See FreezeMultiXactId for full details.
* than the multixid cutoff point). Whenever a full table vacuum happens, *
* the freezing point so computed is used as the new pg_class.relminmxid * Whenever VACUUM advances relminmxid, then either its OldestMxact cutoff
* value. The minimum of all those values in a database is stored as * or the oldest extant Multi remaining in the table is used as the new
* pg_database.datminmxid. In turn, the minimum of all of those values is * pg_class.relminmxid value (whichever is earlier). The minimum of all
* stored in pg_control and used as truncation point for pg_multixact. At * relminmxid values in each database is stored in pg_database.datminmxid.
* checkpoint or restartpoint, unneeded segments are removed. * In turn, the minimum of all of those values is stored in pg_control.
* This is used as the truncation point for pg_multixact when unneeded
* segments get removed by vac_truncate_clog() during vacuuming.
*/ */
MultiXactId perBackendXactIds[FLEXIBLE_ARRAY_MEMBER]; MultiXactId perBackendXactIds[FLEXIBLE_ARRAY_MEMBER];
} MultiXactStateData; } MultiXactStateData;
@ -2497,8 +2499,9 @@ ExtendMultiXactMember(MultiXactOffset offset, int nmembers)
* longer have any running member transaction. * longer have any running member transaction.
* *
* It's not safe to truncate MultiXact SLRU segments on the value returned by * It's not safe to truncate MultiXact SLRU segments on the value returned by
* this function; however, it can be used by a full-table vacuum to set the * this function; however, it can be set as the new relminmxid for any table
* point at which it will be possible to truncate SLRU for that table. * that VACUUM knows has no remaining MXIDs < the same value. It is only safe
* to truncate SLRUs when no table can possibly still have a referencing MXID.
*/ */
MultiXactId MultiXactId
GetOldestMultiXactId(void) GetOldestMultiXactId(void)
@ -2799,9 +2802,8 @@ ReadMultiXactCounts(uint32 *multixacts, MultiXactOffset *members)
* this will also be sufficient to keep us from using too many offsets. * this will also be sufficient to keep us from using too many offsets.
* However, if the average multixact has many members, we might exhaust the * However, if the average multixact has many members, we might exhaust the
* members space while still using few enough members that these limits fail * members space while still using few enough members that these limits fail
* to trigger full table scans for relminmxid advancement. At that point, * to trigger relminmxid advancement by VACUUM. At that point, we'd have no
* we'd have no choice but to start failing multixact-creating operations * choice but to start failing multixact-creating operations with an error.
* with an error.
* *
* To prevent that, if more than a threshold portion of the members space is * To prevent that, if more than a threshold portion of the members space is
* used, we effectively reduce autovacuum_multixact_freeze_max_age and * used, we effectively reduce autovacuum_multixact_freeze_max_age and