|
Hello,
I've tracked the following behavior twice now in my
/var/log/messages file:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Jan 31 18:55:54 lvs2 pulse[13171]: partner dead:
activating lvs
Jan 31 18:55:54 lvs2 pulse[13171]: partner active:
deactivating
lvs
Feb 2 23:04:08 lvs2 pulse[23987]: partner dead:
activating lvs
Feb 2 23:04:08 lvs2 pulse[23987]: partner active:
deactivating
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I understand these log entries to mean for a split
second my backup LVS pulse daemon loses connection
with the primary, and momentarily says it starts up
lvs services, then immediately after reconnects with
the primary, and says it deactivates the lvs services.
Now what seems to occur is that LVS does not actually
or completely get disabled on the backup lvs box, so
routing gets confused (due to LVS IP address active on
both boxes I assume).
Is there LVS-code-based reason this would occur?
Granted both boxes were up for 70+ and 50+ days each,
and I've rebooted them now, since we had a good 50
days with incident-free behavior, and I've not seen
this occur before.
Running Red Hat Linux 7.3 with all errata, and kernel
kernel-2.4.18-18.7.x. Also running piranha-0.7.0-3,
ipvsadm-1.21-4, scsi_reserve-0.7-6.
Thanks for any input on what the cause is, and how to
fix it.
Thanks
Peter
__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com
|
