[Xen-users] xen cluster down

["J. D." <jdonline@xxxxxxxxx>]

Hey Ladies and Gentlemen,

We had a problem with our rh5.2 xen cluster last Friday. One of the nodes in our
eight node cluster locked up and became unreachable while cloning a guest.
Shortly thereafter our other seven nodes connection to the san timed out
and a fence war ensued.

Our eight nodes all have a shared directory on the san where the domU guest
disk images are stored. The directory itself is /guests. Once the nodes lost the
ability to read that directory Bad Stuff happened.

This occurred once before during the early testing phases but after rebooting all the nodes everything went
back to normal. Now we are only sporadically able to mount the shared storage /guests
directory.  Has any one else seen similar behavior, or have any ideas on which direction
to go?

Here are the details of the config:
8 1950 dell nodes w/ 32GB ram running rh5.2 dom0
domU are mostly rh5.2 HVM w/ a few rh5.2 PV and one 2003 HVM
redhat cluster suite, conga
EMC CX310 san using emcpowerpath software to provide the /guests directory to
dom0's (GFS2)
Each dom0 is dual-homed and the config is handled via a custom network-bridge
script (network-multi-bridge) on each node

Any advice greatly appreciated,

J. D.

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users