Oracle support note for Leap Second Hang problem that may result into 100% CPU utilization in Linux environment
By Anand Akela on Jul 04, 2012
On or around July 1, 2012, Oracle has become aware of an issue on Linux distributions resulting from the introduction of the leap second; this is causing problems for some customers. Leap seconds may be introduced at the end of June or December in a calendar year, like 2012, as necessary to maintain time standards. Servers hosting Oracle products which are clients of an NTP server (Network Time Protocol) may be particularly susceptible to this issue as the NTP server is updated.
Linux distributions which may be affected include Oracle Enterprise Linux, Red Hat Enterprise Linux, Oracle VM and Oracle Unbreakable Enterprise Kernel. Asianux 2 and 3, based on RHEL 4 and 5, may also be affected. One report of correction to high agent CPU using Note 1472421.1 on SLES11 has also been reported.
Not all customers will be affected, but those, who are affected, may observe higher than normal CPU consumption on their Linux environments where JVM's are utilized. In Oracle Enterprise Manager ( EM ) , this problem can manifest itself as high CPU consumption with the EM Agent process (which runs on a JVM in EM 12c, for instance). It is possible that the OMS is also affected.
would advise customers to review the description of this problem in MOS
Note 1472651.1 and take action if they observe that their environment
Contributed by Andrew Bulloch , Director, Application Systems Management Products