From: Herbert Poetzl (herbert_at_13thfloor.at)
Date: Tue 27 Jan 2004 - 23:04:49 GMT
On Tue, Jan 27, 2004 at 11:56:46AM -0700, Matthew Excell wrote:
> Good morning, all! (GMT-7) :-)
> It is definitely getting closer!
> SMP AMD Opteron (32-bit kernel)
> kernel-org kernel + MPPE patch + vs0.06
> gcc-3.3 (32-bit only - not biarch)
> - hostname "leaks" to context 0 (context 0 has the hostname of last
> vserver started.)
okay, missing hostname (utsname) virtualization,
so that was (almost) expected ...
> - I don't know if the "vserver stop reboots server" problem is still
> there. (vserver stop fails for another reason - see below.)
> stable util-vserver-0.27
> - vserver-stat is basically working (still has issues with uptime)
> - vserver start and enter will bind and change contexts but something
> (ps?) is complaining about proc not being mounted.
> - ps complains about proc not being mounted when in ctx 1 - so vps
> doesn't work. Ditto for "vserver stop"
hmm, the vs0.06 version is very similar to the devel
releases, so the procfs is 'secure' by default,
which means you probably have to 'enable' some proc
entries first, before the 'current' vserver tools
will work (and probably vserver services too)
/proc/*info and /proc/sta* are good candidates ...
> rc util-vserver-0.27.91
> - no apparent difference from 0.27
> alpha util-vserver-0.27.199
> - vps complains about proc
> - vserver-stat: open(): no such file or directory (strace is trying to
> load libraries. . .)
> - vserver cannot find vserver-setup
interesting, enrico, any ideas?
> I'll have test results on x86_64 soon.
> Contact me directly about 2.6 sponsoring - what did you have in mind?
I'll do so, thanks for testing this, I really
appreciate it ...
> Matthew Excell <matt_at_possibilityforge.com>
> The Possibility Forge, Inc.
Vserver mailing list