Re: Rethinking stats communication mechanisms

Greg Stark <gsstark@mit.edu> writes:

> Douglas McNaught <doug@mcnaught.org> writes:
>
>> Yeah, but if you turn on query logging in that case you'll see the
>> bajillions of short queries, so you don't need the accurate snapshot
>> to diagnose that.
>
> Query logging on a production OLTP machine? a) that would be a huge
> performance drain on the production system b) it would produce so much logging
> that it would take a significant amount of disk and i/o resources just to
> handle and c) you would need to implement special purpose tools just to make
> sense of these huge logs.

(a) and (b): of course you would only do it on a temporary basis for
problem diagnosis.  We do that with our production apps where I work
(when absolutely necessary).

(c): Perl.  :)

Ideally, you'd find the query storm problem in load testing before you
ever got to production.  I hope to someday visit that planet--it must
be nice.

-Doug