From: Herbert Pötzl (herbert_at_13thfloor.at)
Date: Thu 10 Jul 2003 - 18:09:04 BST
On Thu, Jul 10, 2003 at 05:58:09PM +0100, Nuno Silva wrote:
>
> Hi!
>
> Since everyone is guessing I'll try my best shot :)
I was not guessing, I just provided facts .. 8-)
> Try running a kernel without highmem enabled and boot linux with
> mem=900M. If that restores stable operation you should upgrade
> your JRE and JDK to the very latest ones (1.4.*) from Sun or
> Blackdown and try again with highmem enabled.
this could indeed be an issue, highmem often gives
trouble ... together with lvm it's almost certain
to fail/hang/screw up ...
best,
Herbert
> Good luck,
> Nuno Silva
>
> Kyle Yencer wrote:
> >Hi,
> >
> >I have been working using the VServer 0.22 and ctx-17 in a scenario with
> >anywhere from 5 to 20 VServers running, some of which run java inside of
> >them. In the beginning phases, the server would crash anywhere from 7
> >to 10 days, however now it will crash almost daily and at that multiple
> >times sometimes. The Host OS is Redhat 7.3 and the kernel is a vanilla
> >patched 2.4.20 kernel. I have also tried the 2.4.19-ctx15 kernel as
> >will with no better results. The machine is a dual xeon with 2 gigs of
> >ram, it's quite possible the issue is hardware related, the server is
> >running a i2o raid card and eepro100 nics. Has anyone ran into similiar
> >problems? The only thing i've been able to grab before the server
> >crashed is the last proc, which coincidently happened to be a java
> >process. Please let me know if anyone has seen similar behavior, or
> >where anyone has found stable ground if this is normal.
> >
> >Kyle Yencer