> > REINDEX is really intended for disaster recovery, not routine space
> > management.
>
> That's nice. Except the fact Postgres is so bad at maintaining its own
> indexes makes us have to *use* it as routine space management. Having a
> 200MB index that's still getting bigger, and really should only be 200k,
> tells me that there's no better use for it. Sure, I could wait until
> the index takes up my entire disk, and then consider it disaster
> recovery, but that's just silly.
I think was Tom meant is, that REINDEX should not have to be used for
routine space management. It *should* be only there for disaster recovery.
Unfortunately this is not the case in the latest releases, so we all use it
now temporarily for space management, although this is surely going to be
fixed in one of the next releases.
Greetings,
Bjoern