Re: How to cripple a postgres server - Mailing list pgsql-general

From Stephen Robert Norris
Subject Re: How to cripple a postgres server
Date
Msg-id 1022551491.2670.11.camel@ws12
Whole thread Raw
In response to Re: How to cripple a postgres server  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: How to cripple a postgres server
List pgsql-general
On Tue, 2002-05-28 at 11:57, Tom Lane wrote:
> Stephen Robert Norris <srn@commsecure.com.au> writes:
> > 1) Open many (I used 800) database connections and leave them idle.
> > 2) run: while true; do vacuum {database} ; done
>
> > Wait. Observe db has stopped doing useful work.
>
> Is your machine capable of supporting 800 active backends in the absence
> of any vacuum processing?  I'm not sure that "we're driving our server
> into swapping hell" qualifies as a Postgres bug ...
>
> If you're seeing load peaks in excess of what would be observed with
> 800 active queries, then I would agree there's something to investigate
> here.
>
>             regards, tom lane

Yep, indeed with 800 backends doing a query every second, nothing
happens.

The machine itself has 1GB of RAM, and uses no swap in the above
situation. Instead, system time goes to 99% of CPU. The machine is a
dual-CPU athlon 1900 (1.5GHz).

It _only_ happens with idle connections!

    Stephen

Attachment

pgsql-general by date:

Previous
From: Tom Lane
Date:
Subject: Re: How to cripple a postgres server
Next
From: Martijn van Oosterhout
Date:
Subject: Re: is there any backend (server) timeout undocumented?