Re: [HACKERS] issue with track_commit_timestamp and server restart

2016-10-24 Thread Tom Lane
Alvaro Herrera writes: > Now let's discuss a release note entry for 9.5 and 9.6, shall we? The text in your commit message seems sufficient from here. On it now (although if somebody doesn't fix anon git PDQ, we're not releasing today anyway).

Re: [HACKERS] issue with track_commit_timestamp and server restart

2016-10-24 Thread Alvaro Herrera
Now let's discuss a release note entry for 9.5 and 9.6, shall we? -- Álvaro Herrerahttps://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your

Re: [HACKERS] issue with track_commit_timestamp and server restart

2016-10-24 Thread Alvaro Herrera
Craig Ringer wrote: > I initially could'n't see this when tested on 8f1fb7d with a > src/test/recovery/t test script. But it turns out it's OK on immediate > shutdown and crash recovery, but not on clean shutdown and restart. Oh, oops. > The attached patch adds a TAP test to src/test/recovery

Re: [HACKERS] issue with track_commit_timestamp and server restart

2016-10-24 Thread Julien Rouhaud
On 24/10/2016 06:58, Craig Ringer wrote: > On 22 October 2016 at 19:51, Julien Rouhaud wrote: >> I just noticed that if track_commit_timestamp is enabled, the >> oldestCommitTsXid and newestCommitTsXid don't persist after a server >> restart, so you can't ask for the

Re: [HACKERS] issue with track_commit_timestamp and server restart

2016-10-23 Thread Craig Ringer
On 24 October 2016 at 12:58, Craig Ringer wrote: > The attached patch adds a TAP test to src/test/recovery to show this. Added to CF. https://commitfest.postgresql.org/11/834/ ... but IMO this should go in the next bugfix release. I've also applied nearly the same fix

Re: [HACKERS] issue with track_commit_timestamp and server restart

2016-10-23 Thread Michael Paquier
(thread hijacking) On Mon, Oct 24, 2016 at 1:58 PM, Craig Ringer wrote: > To save time running the recovery suite, just > >rm src/test/recovery/00[0-8]*.pl > > (It'd be nice to have a prove_check target to run just one test file). Agreed! Or multiple chosen files. I

Re: [HACKERS] issue with track_commit_timestamp and server restart

2016-10-23 Thread Craig Ringer
On 22 October 2016 at 19:51, Julien Rouhaud wrote: > I just noticed that if track_commit_timestamp is enabled, the > oldestCommitTsXid and newestCommitTsXid don't persist after a server > restart, so you can't ask for the commit ts of a transaction that > committed