Jun 2020

1 commit
Save slot's restart_lsn when invalidated due to size
24 Jun 2020Alvaro Herreramaster14beta1REL_19_STABLE14beta1REL_18_STABLE14beta1REL_17_STABLE14beta1REL_16_STABLE14beta1REL_15_STABLE14beta1REL_14_STABLE14beta1REL_13_STABLE13beta3Also in: Review for GetWALAvailability() (part of a posted patchset)

Apr 2020

2 commits
Fix checkpoint signalling
29 Apr 2020Alvaro Herreramaster13beta1REL_19_STABLE13beta1REL_18_STABLE13beta1REL_17_STABLE13beta1REL_16_STABLE13beta1REL_15_STABLE13beta1REL_14_STABLE13beta1REL_13_STABLE13beta1
LLM analysis, 100% suremodified before commitmoderate changesfrom patchset #130

Matches the checkpoint fix and TAP test in #130, with SetLatch(MyLatch) replaced by continue and the accompanying comment rewritten, as explained in #132. The commit credits Horiguchi and Herrera as co-authors; Horiguchi supplied the original fix and test.

Check slot->restart_lsn validity in a few more places
29 Apr 2020Alvaro Herreramaster13beta1REL_19_STABLE13beta1REL_18_STABLE13beta1REL_17_STABLE13beta1REL_16_STABLE13beta1REL_15_STABLE13beta1REL_14_STABLE13beta1REL_13_STABLE13beta1
LLM analysis, 100% suremodified before commitmoderate changesfrom patchset #1251 of 1 reviewer saw the final version

Derives from #125, which expands Horiguchi's #121; the commit moves the logical SQL check after CreateDecodingContext(), preserves the physical-slot diagnostic, and revises error wording. Herrera describes those final edits in #131.

  • Álvaro Herrerasaw the final versiontrailerHe expanded the original fix in #125, reviewed diagnostic ordering in #129, and committed the adjusted version described in #131.

Mar - Apr 2020

2 commits
Allow users to limit storage reserved by replication slots
7 Apr 2020Alvaro Herreramaster13beta1REL_19_STABLE13beta1REL_18_STABLE13beta1REL_17_STABLE13beta1REL_16_STABLE13beta1REL_15_STABLE13beta1REL_14_STABLE13beta1REL_13_STABLE13beta1
LLM analysis, 100% suremodified before commitlight changesfrom patchset #1161 of 3 reviewers saw the final version

Matches the full patch in #116, whose header credits Horiguchi, with keeping/PRESERVED renamed to reserved/RESERVED, an unused function argument removed, documentation and formatting edits, and a catalog-version bump. Herrera confirms committing this version in #119, explicitly leaving out #118.

  • Masahiko Sawadasaw an earlier versiontrailerHis code reviews, tests, and delta in #52 corrected retention and monitoring arithmetic, but checkpoint-time invalidation and slot ownership were substantially redesigned afterward.
  • Jehan-Guillaume de Rorthaissaw an earlier versiontrailerHe rebased and traffic-tested the series in #85, verified revised states in #90, and reviewed APIs in #94 before the final invalidation redesign.
  • Álvaro Herrerasaw the final versiontrailerHe reviewed the APIs, reworked checkpoint invalidation and slot acquisition, tested the combined patch in #116, and committed its final form.
Remove header noise from test_decoding test
31 Mar 2020Alvaro Herreramaster13beta1REL_19_STABLE13beta1REL_18_STABLE13beta1REL_17_STABLE13beta1REL_16_STABLE13beta1REL_15_STABLE13beta1REL_14_STABLE13beta1REL_13_STABLE13beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #991 of 1 reviewer saw the final version

Extracts exactly the expanded-output changes in contrib/test_decoding/sql/ddl.sql and expected/ddl.out from #99, leaving the WAL-limit feature for later. Herrera announces this separate landing in #102; Michael Paquier is credited for the idea rather than a subsequent review.

  • Alvaro Herrerasaw the final versioncommittercommitted the patch

Credits across 5 commits

Author
Kyotaro Horiguchi1 LLM4
Reviewer
Alvaro Herreracommitter4
Jehan-Guillaume de Rorthais1
Masahiko Sawada1
Co author
Alvaro Herrera1
Kyotaro Horiguchi1
Committer
Alvaro Herrera5

N LLM means N of the credits come from LLM analysis, not from commit trailers.