Hi,
I have a CPPM cluster with 8 appliances running 6.6.8.100017. In the last 24 hours two subscribers started logging dbcn, cpass-londiste restarted, and replication errors. One of the subscibers also stopped authenticating 802.1x users for about one minute today.
Both of these two servers are pegged at 100% 'Swap Memory Usage'. 'Swap Space' is over 102% used. See attached pics. All servers remain in sync and aside from the 1 minute blip the two troubled subscribers are still handling their own. I have a small maintenance window tomrorow and would like to reboot one of these as a first step.
My question is should I drop the subsciber first before rebooting it or just reboot it? If just reboot should I stop RADIUS service first?
If anyone has any thoughts about root cause or have seen this before I'd be inerested to know your experience. The only changes made was adding syslog parameters to a different subsciber (not having issues) and literally just logging into Insight to poke around.
Thanks in advance.