Re: Is anyone aware of data loss causing MultiXact bugs in 9.3.2?

On Wed, Feb 19, 2014 at 12:40 AM, Andres Freund <andres@2ndquadrant.com> wrote:
> Was there an index only scan or just a index scan? Any chance of a
> corrupted index?

Just an index scan. I think it's unlikely to be a corrupt index,
because the customer said that he dropped and restored the index, and
it's difficult to imagine how VACUUM FREEZE could then have impacted a
plan with only a sequential scan.

> Do you still have the page from before you did the VACUUM?

It seems likely that I could get it, given that the problem persisted
over several days (apparently there was an earlier, manual attempt to
repair the damage by hand). I'm reasonably confident that I could get
back an affected database by performing a point-in-time recovery. That
would also give me the benefit of an environment that I could break as
needed.


-- 
Peter Geoghegan