From: Andreas Ljunggren (andreas_at_ljunggren.net)
Date: Tue 14 Jan 2003 - 18:05:35 GMT
----- Original Message -----
From: "Herbert Poetzl" <herbert_at_13thfloor.at>
To: <vserver_at_solucorp.qc.ca>
Sent: Tuesday, January 14, 2003 4:18 PM
Subject: Re: [vserver] Grandma got run over by newvserver
> On Tue, Jan 14, 2003 at 05:49:46PM +0300, Lyashkov Alexey wrote:
> > Hello Herbert,
> >
> > Tuesday, January 14, 2003, 5:22:26 PM, you wrote:
> >
> > HP> On Tue, Jan 14, 2003 at 01:53:41AM -0500, Adam H. Pendleton wrote:
> > >> I was hoping to get the answer without having to provide that
> > >> information, since I'm expecting that the vserver/kernel version is
> > >> going to immediately be pointed to as the problem, but nevertheless,
I
> > >> am using the latest patches by Lyahkov Aleks, but they are currently
> > >> unreleased by him, so they are still considered test patches. These
> > >> patches are applied against 2.4.18-19.7.x, the latest RedHat 7.3
> > >> kernel. I am in contact with Lyahkov to see if his patches may be to
> > >> blame, but does anyone have any other ideas?
> >
> > HP> vserver-0.20 scripts?
> > yes. i use it verion.
>
> maybe but not necessarily this could be related to
>
> - vserver exec missing ulimits
> - vrpm missing chcontext
>
> I would suggest to look for code which should be
> executed 'within' a vserver context, but isn't and
> this way affecting the physical server ...
>
> in vserver-0.18 any sshd upgrade could disconnect
> your machine from the network, because of the
> badly designed pre/post install scripts and the
> missing chcontext ...
>
> best,
> Herbert
>
I'we also tried the Aleks patched version, and I'm pretty sure it didn't
work as expected since I
had some similar problems as stated above.
Thou I tried chbind and verified that it did NOT work, thus the patch must
have failed somehow, even
thou process context where working.
I didn't look much at it, but rather switched straight over to a "vanilla +
vserver" kernel instead.
Regards
Andreas