Re: How to determine a database is intact?

On 9/4/04 5:53 AM, "Jan Wieck" <JanWieck(at)yahoo(dot)com> wrote:

> [snip]

Well, I had a big long response composed to your snide remarks, but decided
I'm not getting in a flame war with you on something that is irrelevant to
the question I posed.

> Is the server at
> least configured with ECC Ram, or is the data not important enough to
> justify for quality hardware?

As a matter of fact it does have ECC, and no errors have been reported by
the system or diagnostics. It's a Dell 6650. No disk errors have been
reported either (hardware RAID controller).

As it pertains to the question at hand, I don't care what caused the
corruption. It's totally irrelevant to the question. Perfect hardware and
perfect software don't exist. There will be hardware failures. There will
be software failures. The question was whether or not there is a feasible
way of determining at any given point in time that a database is fully
intact. We are already now doing weekly pg_dumpall's. Doing a restore of
each of those simply isn't reasonable. If there is no such integrity
utility to scan the DB in place, then that's the way it is. That's the
answer to my question. But, quit trying blame crappy hardware.

Wes