Re: [Evolution] Evo die with excessive disk activity

[John Stile <john stilen com>]

On Tue, 2006-01-24 at 15:40 -0200, Eduardo Spremolla wrote:
> I see the same behavior here, evo 2.4.1 on FC4.
>
> top on a text console show that gconf got half the real memory ( I have
> 380MB ).
> But some times is enough to kill all incarnations of sampd to get the
> system responsive again.
>
>
> LALO
>
> On Sun, 2006-01-22 at 14:03 +0100, Alessio Bragadini wrote:
> > On Sat, 2006-01-21 at 18:20 +0100, Andre Klapper wrote:
> >
> > > i also see this (evo-2.4 consuming more than 80% of my memory
> > > according to 'top') sometimes after running evo for a few days
> >
> > I cannot run Evolution "for a few days"! In the past I could leave it
> > open for days, but after starting to use Virtual Folders, and with my
> > mail archive growing, this is no longer possible and I've noticed that
> > hitting on the "Unmatched" VFolder (which apparently I cannot remove)
> > really degrades performances. Now I've started using Virtual Folders
> > even more via the setting of labels, and I prefer to restart Evo shell
> > and eds a couple of times during the days.
> >
> > Unfortunately I am still on 2.2.3 (via Fedora Core 4) and I don't know
> > if later releases fix this. I am an heavy user with POP, IMAP and NNTP
> > accounts and misc virtual folders used all the time.
I see the same behavior with evolution-2.4.2.1 on gentoo.
I notice that 'vmstat 1' shows the 'block out' column constantly high.
If I kill evolution and restart, initially it acts ok. I created a handy
alias to kill evolution: alias killevo='evolution --force-shutdown ;
bonobo-slay; gconftool-2 --shutdown'.  I do not use virtual folders.