> Yeah, this is a bug. It seems that the WAL summarizer process, when
> restarted, wants to restart from wherever it was previously
> summarizing WAL, which is correct if that WAL is still around, but if
> summarize_wal has been turned off in the meanwhile, it might not be
> correct. Here's a patch to fix that.
Thanks for checking this!
> Thanks. New version attached.
And besides that, thanks for the patch, of course!
I compiled Postgres locally with your patch. I attempted to break it several
times, both manually and through a shell script.
No success on that -- which in this case is actually success :)
The WAL summarizer seems able to always resume from a valid point,
so `pg_basebackup` isn't failing anymore.