When I got up this morning, I noticed that Zimbra 7.2.0 was eating 100% of the available CPU on a KVM server. Some quick searches point to a Java leap-second bug AND a fix.
\# /etc/init.d/ntp stop
\# date `date +"%m%d%H%M%C%y.%S"`
\# echo “/etc/init.d/ntp start” | at now + 12 hours
It was amazing to see the CPU use drop immediately when the date command was run. Didn’t even need to restart any of the Zimbra processes. Amazing.
Life with Zimbra is good again. I’m just glad I was looking at the performance. If I hadn’t just done a Zimbra upgrade, I would never have bothered.