Re: pg_upgrade < 9.3 -> >=9.3 misses a step around multixacts

On Fri, May 30, 2014 at 02:16:31PM +0200, Andres Freund wrote:
> Hi,
>
> When upgrading a < 9.3 cluster pg_upgrade doesn't bother to keep the old
> multixacts around because they won't be read after the upgrade (and
> aren't compatible). It just resets the new cluster's nextMulti to the
> old + 1.
> Unfortunately that means that there'll be a offsets/0000 file created by
> initdb around. Sounds harmless enough, but that'll actually cause
> problems if the old cluster had a nextMulti that's bigger than that
> page.
>
> When vac_truncate_clog() calls TruncateMultiXact() that'll scan
> pg_multixact/offsets to find the earliest existing segment. That'll be
> 0000. If the to-be-truncated data is older than the last existing
> segment it returns. Then it'll try to determine the last required data
> in members/ by accessing the oldest data in offsets/.
>
> Unfortunately, due to the existing 0000/ segment, that means it'll
> sometimes try to access a nonexistant offsets/ file. Preventing vacuum
> et al from succeeding.
>
>
> It seems to me the fix for this is to a) rmtree("pg_multixact/members",
> false) in copy_clog_xlog_xid() in the oldcluster < 9.3 case b) add a
> warning to the release notes that everyone that used pg_upgrade and has
> a 0000 file lying around should report to the mailinglist.
>
> b) is a bit unsatisfactory, but I don't want to suggest removing the
> file. In too many situations it might actually still be needed.

This is a bug in 9.3 pg_upgrade as well?  Why has no one reported it
before?

--
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + Everyone has their own god. +