Re: concurrent updates problem - Mailing list pgsql-general

From mwaples@waples.net
Subject Re: concurrent updates problem
Date
Msg-id 3AB6366C.C5620478@waples.net
Whole thread Raw
In response to Re: concurrent updates problem  ("Richard Huxton" <dev@archonet.com>)
List pgsql-general
[snip junk]
> [snipped code that looked simple enough to a non tcl-er]
> >
> > they both produces sometimes correct results, sometimes under what the
> > result should be and sometimes over it.
> > But 5 or 6 concurrent requests always produce the correct results, any
> > more than that things go wrong.
>
> Your code looks fine (it's only sql="..." and exec $sql so there can't
> be anything wrong with it).
>
> The only things I can think of are:
> 1. bug with pg+aolserver (unlikely I think openACS uses that combo). The
> only peculiarity I know with aolsvr is that it's multi-threading, and I
> don't know if that could cause this.
>
> 2. you aren't actually fetching the number of pages you think you are.
> It could be that calls are failing and being redone without ab telling
> you. Not sure how to tell this without counting packets on the
> wire/turning logging to max on the webserver.
>
> - Richard Huxton

Im guessing 2.
I just tried using httperf with --num-conn 50 ra 50
My counter went up by 50 every time - and httperf reported around 25
concurrent requests.

Is there a good way to test out concurrent queries,  and see if they are
working as expected ?
ab doesnt seem to be reporting right - although httperf gives the
results i expect.

pgsql-general by date:

Previous
From: "Poul L. Christiansen"
Date:
Subject: Re: Cant connect if -B 1024 was set to postmaster
Next
From: "Brent R. Matzelle"
Date:
Subject: Re: Cannot start postmaster