Heikki pointed out to me that the btree delete record processing does
not respect vacuum_defer_cleanup_age. It should.
Attached patch to implement that.
Looking to commit in next few hours barring objections/suggestions, to
both HEAD and 9_0_STABLE, in time for next minor release.
--
Simon Riggs http://www.2ndQuadrant.com/books/
PostgreSQL Development, 24x7 Support, Training and Services
diff --git a/src/backend/access/nbtree/nbtxlog.c b/src/backend/access/nbtree/nbtxlog.c
index 0822f5c..5b2d58a 100644
--- a/src/backend/access/nbtree/nbtxlog.c
+++ b/src/backend/access/nbtree/nbtxlog.c
@@ -683,6 +683,12 @@ btree_xlog_delete_get_latestRemovedXid(XLogRecord *record)
UnlockReleaseBuffer(ibuffer);
/*
+ * Apply vacuum_defer_cleanup_age, if we have a valid xid.
+ */
+ if (TransactionIdIsValid(latestRemovedXid))
+ TransactionIdRetreatMany(latestRemovedXid, vacuum_defer_cleanup_age);
+
+ /*
* Note that if all heap tuples were LP_DEAD then we will be returning
* InvalidTransactionId here. That can happen if we are re-replaying this
* record type, though that will be before the consistency point and will
diff --git a/src/backend/storage/ipc/procarray.c b/src/backend/storage/ipc/procarray.c
index 6e7a6db..c16a287 100644
--- a/src/backend/storage/ipc/procarray.c
+++ b/src/backend/storage/ipc/procarray.c
@@ -1129,8 +1129,7 @@ GetOldestXmin(bool allDbs, bool ignoreVacuum)
LWLockRelease(ProcArrayLock);
/*
- * Compute the cutoff XID, being careful not to generate a "permanent"
- * XID.
+ * Compute the cutoff XID, being careful not to generate a reserved XID.
*
* vacuum_defer_cleanup_age provides some additional "slop" for the
* benefit of hot standby queries on slave servers. This is quick and
@@ -1140,9 +1139,7 @@ GetOldestXmin(bool allDbs, bool ignoreVacuum)
* wraparound --- so guc.c should limit it to no more than the
* xidStopLimit threshold in varsup.c.
*/
- result -= vacuum_defer_cleanup_age;
- if (!TransactionIdIsNormal(result))
- result = FirstNormalTransactionId;
+ TransactionIdRetreatMany(result, vacuum_defer_cleanup_age);
return result;
}
diff --git a/src/include/access/transam.h b/src/include/access/transam.h
index a7ae752..2f7070e 100644
--- a/src/include/access/transam.h
+++ b/src/include/access/transam.h
@@ -58,6 +58,19 @@
(dest)--; \
} while ((dest) < FirstNormalTransactionId)
+#define TransactionIdRetreatMany(dest, many) \
+{ \
+ if ((dest) >= (many)) \
+ { \
+ (dest) -= (many); \
+ while ((dest) < FirstNormalTransactionId) \
+ { \
+ (dest)--; \
+ } \
+ } \
+ else \
+ (dest) = MaxTransactionId - (many) + (dest); \
+}
/* ----------
* Object ID (OID) zero is InvalidOid.
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
