Fix intermittent BF failures in 035_conflicts.
This commit addresses two sources of instability in the 035_conflicts
test:
Unstable VACUUM usage:
The test previously relied on VACUUM to remove a deleted column, which can
be unreliable due to concurrent background writer or checkpoint activity
tha
pg_restore: Fix comment handling with --no-publications / --no-subscriptions.
Previously, pg_restore did not skip comments on publications or subscriptions
even when --no-publications or --no-subscriptions was specified. As a result,
it could issue COMMENT commands for objects that were never crea
Fix incorrect const qualifier
Commit 7202d72787d added in passing some const qualifiers, but the one
on the postmaster_child_launch() startup_data argument was incorrect,
because the function itself modifies the pointed-to data. This is
hidden from the compiler because of casts. The qualifiers o
Fix incorrect const qualifier
Commit 7202d72787d added in passing some const qualifiers, but the one
on the postmaster_child_launch() startup_data argument was incorrect,
because the function itself modifies the pointed-to data. This is
hidden from the compiler because of casts. The qualifiers o
pg_restore: Fix comment handling with --no-publications / --no-subscriptions.
Previously, pg_restore did not skip comments on publications or subscriptions
even when --no-publications or --no-subscriptions was specified. As a result,
it could issue COMMENT commands for objects that were never crea
pg_restore: Fix comment handling with --no-publications / --no-subscriptions.
Previously, pg_restore did not skip comments on publications or subscriptions
even when --no-publications or --no-subscriptions was specified. As a result,
it could issue COMMENT commands for objects that were never crea
pg_restore: Fix comment handling with --no-policies.
Previously, pg_restore did not skip comments on policies even when
--no-policies was specified. As a result, it could issue COMMENT commands
for policies that were never created, causing those commands to fail.
This commit fixes the issue by en
pg_restore: Fix comment handling with --no-publications / --no-subscriptions.
Previously, pg_restore did not skip comments on publications or subscriptions
even when --no-publications or --no-subscriptions was specified. As a result,
it could issue COMMENT commands for objects that were never crea
pg_restore: Fix comment handling with --no-publications / --no-subscriptions.
Previously, pg_restore did not skip comments on publications or subscriptions
even when --no-publications or --no-subscriptions was specified. As a result,
it could issue COMMENT commands for objects that were never crea
pg_restore: Fix comment handling with --no-publications / --no-subscriptions.
Previously, pg_restore did not skip comments on publications or subscriptions
even when --no-publications or --no-subscriptions was specified. As a result,
it could issue COMMENT commands for objects that were never crea
pg_restore: Fix comment handling with --no-publications / --no-subscriptions.
Previously, pg_restore did not skip comments on publications or subscriptions
even when --no-publications or --no-subscriptions was specified. As a result,
it could issue COMMENT commands for objects that were never crea
Update various forward declarations to use typedef
There are a number of forward declarations that use struct but not the
customary typedef, because that could have led to repeat typedefs,
which was not allowed. This is now allowed in C11, so we can update
these to provide the typedefs as well, s
Teach nbtree to avoid evaluating row compare keys.
Add logic to _bt_set_startikey that determines whether row compare keys
are guaranteed to be satisfied by every tuple on a page that is about to
be read by _bt_readpage. This works in essentially the same way as the
existing scalar inequality log
Expand virtual generated columns in constraint expressions
Virtual generated columns in constraint expressions need to be
expanded because the optimizer matches these expressions to qual
clauses. Failing to do so can cause us to miss opportunities for
constraint exclusion.
Author: Richard Guo
D
Improve ExplainState type handling in header files
Now that we can have repeat typedefs with C11, we don't need to use
"struct ExplainState" anymore but can instead make a typedef where
necessary. This doesn't change anything but makes it look nicer.
(There are more opportunities for similar cha
Expand virtual generated columns in constraint expressions
Virtual generated columns in constraint expressions need to be
expanded because the optimizer matches these expressions to qual
clauses. Failing to do so can cause us to miss opportunities for
constraint exclusion.
Author: Richard Guo
D
CREATE STATISTICS: improve misleading error message
The previous change (commit f225473cbae) was still not on target,
because it talked about relation kinds, which are not what is being
checked here. Provide a more accurate message.
Reviewed-by: Tom Lane
Discussion:
https://postgr.es/m/CACJufx
CREATE STATISTICS: improve misleading error message
The previous change (commit f225473cbae) was still not on target,
because it talked about relation kinds, which are not what is being
checked here. Provide a more accurate message.
Reviewed-by: Tom Lane
Discussion:
https://postgr.es/m/CACJufx
CREATE STATISTICS: improve misleading error message
The previous change (commit f225473cbae) was still not on target,
because it talked about relation kinds, which are not what is being
checked here. Provide a more accurate message.
Reviewed-by: Tom Lane
Discussion:
https://postgr.es/m/CACJufx
CREATE STATISTICS: improve misleading error message
The previous change (commit f225473cbae) was still not on target,
because it talked about relation kinds, which are not what is being
checked here. Provide a more accurate message.
Reviewed-by: Tom Lane
Discussion:
https://postgr.es/m/CACJufx
CREATE STATISTICS: improve misleading error message
The previous change (commit f225473cbae) was still not on target,
because it talked about relation kinds, which are not what is being
checked here. Provide a more accurate message.
Reviewed-by: Tom Lane
Discussion:
https://postgr.es/m/CACJufx
Remove hbaPort type
This was just a workaround to avoid including the header file that
defines the Port type. With C11, we can now just re-define the Port
type without the possibility of a conflict.
Reviewed-by: Tom Lane
Reviewed-by: Chao Li
Discussion:
https://www.postgresql.org/message-id/f
Remove workarounds against repeat typedefs
This is allowed in C11, so we don't need the workarounds anymore.
Reviewed-by: Chao Li
Discussion:
https://www.postgresql.org/message-id/flat/10d32190-f31b-40a5-b177-11db55597...@eisentraut.org
Branch
--
master
Details
---
https://git.postgre
Resume conflict-relevant data retention automatically.
This commit resumes automatic retention of conflict-relevant data for a
subscription. Previously, retention would stop if the apply process failed
to advance its xmin (oldest_nonremovable_xid) within the configured
max_retention_duration and u
24 matches
Mail list logo
