failback lan caused server reboot

[Shankar_6]

HI,

I have a 2*node cluster running hpux 11.11 and SG version 11.14. When there was a switch port failure in primary card,it happily swtiched to standby card.After 3hrs,when switch port came online,it switched back to primary card. Within 5 secs,i could see it lost the connection to other node causing server to reboot..

probably i may not be able to send the logs,as server got rebooted
I would like to know if there is any known issue with this version when the failback happens it will lose the connectivity and cluster membership will go away..

points will be allotted.

[John Bigg]

The simple answer is no. There are no known problems like this. You really are going to need to look at the logs to get some clues as to what may have happened. Since the system reboot you will need to look at OLDsyslog.log but also the syslog.log file of the node which did not go down could well also hold some clues. The TOC dump may also hold clues as to what happened. You should probably contact support providing all these files.

[Viktor Balogh]

This TOC could have been initiated by Serviceguard also, when the node doesn't reach the other node and the quroum either, then in some configurations it gives itself a TOC in order to quickly reboot. Check your Serviceguard config also.

[Stephen Doud]

This issue deserves investigation by Serviceguard support personnel. I recommend that you open a software support case with HP.