Re: odd buildfarm failure - "pg_ctl: control file appears to be corrupt"
От | David Steele |
---|---|
Тема | Re: odd buildfarm failure - "pg_ctl: control file appears to be corrupt" |
Дата | |
Msg-id | 7edf2ce2-e744-f4d2-21d4-3a0df3b836b0@pgmasters.net обсуждение исходный текст |
Ответ на | Re: odd buildfarm failure - "pg_ctl: control file appears to be corrupt" (Thomas Munro <thomas.munro@gmail.com>) |
Ответы |
Re: odd buildfarm failure - "pg_ctl: control file appears to be corrupt"
|
Список | pgsql-hackers |
Hi Thomas, On 7/26/23 06:06, Thomas Munro wrote: > While chatting to Robert and Andres about all this, a new idea came > up. Or, rather, one of the first ideas that was initially rejected, > now resurrected to try out a suggestion of Andres’s on how to > de-pessimise it. Unfortunately, it also suffers from Windows-specific > problems that I originally mentioned at the top of this thread but > had since repressed. Arrrghgh. > > First, the good news: > > We could write out a whole new control file, and durable_rename() it > into place. We don’t want to do that in general, because we don’t > want to slow down UpdateMinRecoveryPoint(). The new concept is to do > that only if a backup is in progress. That requires a bit of > interlocking with backup start/stop (ie when runningBackups is > changing in shmem, we don’t want to overlap with UpdateControlFile()'s > decision on how to do it). Here is a patch to try that out. No more > weasel wording needed for the docs; basebackup and low-level file > system backup should always see an atomic control file (and > occasionally also copy a harmless pg_control.tmp file). Then we only > need the gross retry-until-stable hack for front-end programs. I like the approach in these patches better than the last patch set. My only concern would be possible performance regression on standbys (when doing backup from standby) since pg_control can be written very frequently to update min recovery point. I've made a first pass through the patches and they look generally reasonable (and back patch-able). One thing: + sendFileWithContent(sink, XLOG_CONTROL_FILE, + (char *) control_file, sizeof(*control_file), + &manifest); I wonder if we should pad pg_control out to 8k so it remains the same size as now? Postgres doesn't care, but might look odd to users, and is arguably a change in behavior that should not be back patched. > And the bad news: Provided we can reasonably address the Windows issues this seems to be the way to go. Regards, -David
В списке pgsql-hackers по дате отправления: