Re: Preventing index scans for non-recoverable index AMs

Pavan Deolasee wrote:
> Just curious, how do we handle the case of corrupted hash index today?

We don't.

> If we can detect that the index is corrupt because of bad page
> headers etc, then its still OK; we can throw an error. But what if the
> hash index is used after recovery and it returns wrong tuple(s) ?

You get to keep both pieces..

In short, don't use hash index, unless you're prepared to run REINDEX 
manually after every crash.

--   Heikki Linnakangas  EnterpriseDB   http://www.enterprisedb.com