Steve Lane <slane(at)soliantconsulting(dot)com> writes:
> Or will postgres 8 rebuild an index if it doesn't find it where it expects?
No, but there's always REINDEX.
> I recognize, or think I do, that PG 8 tablespaces would be required to
> accomplish this.
It'd certainly make it a lot easier. Also, it's probably not worth the
pain to try to move system catalog indexes to RAMdisk --- that'll
greatly complicate the restart process. User-table indexes could be
fixed just with REINDEX.
regards, tom lane
In response to
pgsql-admin by date
|Next:||From: Fabio C. Bon||Date: 2005-07-29 16:30:46|
|Subject: Questions about fsync|
|Previous:||From: Richard_D_Levine||Date: 2005-07-29 15:57:46|
|Subject: Re: Indexes on RAM disk = insanity?|