On Jul 29, 2009, at 2:30 PM, "Andrea Dell'Amico" adellam@sevenseas.org wrote:
On Wed, 2009-07-29 at 16:16 +0200, Andrea Dell'Amico wrote:
On Wed, 2009-07-29 at 09:55 -0400, Ross Walker wrote:
I'm pretty sure the crash is DRBD related: until the secondary drbd server is detached, all is working well. There are 23 guests running, right now, some of them paravirtualized, other full virtualized. Some of them use files images, other logical volumes (all of them over a drbd device). And I don't have a resource starvation, but a kernel crash and an immediate reboot.
It seems that one: http://thread.gmane.org/gmane.linux.network.drbd/17537 but I didn't loose the link between primary and secondary
The OP with the iSCSI problem saw no resource starvation either, yet the hypervisor was rate limiting his dom0 CPU usage to the point where he was missing interrupts.
-Ross