So another question, which I've already received from the field, is how
can you detect this kind of corruption in the first place, if it's not
causing a user-visible error?

Got that question from someone who failed over between master and
replica on 9.3.2 last weekend.  They're not seeing PK/FK errors, but
that doesn't mean they're clean.  How can they tell?

Josh Berkus
PostgreSQL Experts Inc.

Sent via pgsql-hackers mailing list (
To make changes to your subscription:

Reply via email to