The reasons for GridICE monitoring server slowness can be more the one.
First of all, the postgresql database is assorbing almost all the memory
of the machine; we plan to increase it.
Second, the monitoring services were sampled with an high frequence ( e.g.
queues set to be sampled each minute ); we ( the GridICE team and me
) are tuning the configuration and profiling the system.
Now the situation seems to be better; any feedback is welcome.
ciao Piera
On Thu, 20 Nov 2003, John Gordon wrote:
> Piera, do we know the reason for the usual GridICE slowness?
>
> John
>
> Piera Bettini <[log in to unmask]> wrote:
> __________
> >Hi.
> >During the last hours the GridIce monitoring system has been affected by
> >some
> >problems more seious than the usual slowness; they now should be fixed.
> >In any case, an upgrade, already forseen, is necessary; I plan to do it
> >tomorrow. I will send an e-mail when the service is available again.
> >Ciao
> >Piera
> >
> >--
> >---------------------------------------------------
> > Piera Bettini
> > IT Div. - Bat.31 S-013 - CERN - CH-1211 Geneva 23
> > mailto: [log in to unmask]
> > phone : +41-22-7671513
> >---------------------------------------------------
>
--
---------------------------------------------------
Piera Bettini
IT Div. - Bat.31 S-013 - CERN - CH-1211 Geneva 23
mailto: [log in to unmask]
phone : +41-22-7671513
---------------------------------------------------
|