Vertica/DR won?t startup once the database is stopped after a leap second under certain RedHat/Kernel versions

Document ID : KB000031389
Last Modified Date : 14/02/2018
Show Technical Document Details

Problem:

When a Vertica DB cluster has been brought down for maintenance reasons, or for an upgrade after Jun 30th, and needs to be restarted, it is possible that the restart may fail because the nodes do not wait long enough for the other nodes. 

The Vertica database cluster state on that particular server would show as going from READY to UNSAFE when the Vertica application is restarted. 

Solution:

The machines should be rebooted as a workaround and a Linux kernel patch should be applied to avoid the problem in the future.

The leap second issue is addressed in the Redhat Kernel versions listed below:

   RHEL 6.1 EUS: Update to kernel-2.6.32-131.30.2 (from RHBA-2012-1197) or later.
   RHEL 6.2 EUS: Update to kernel-2.6.32-220.25.1.el6 (from RHBA-2012-1198) or later.
   RHEL 6.3: Update to kernel-2.6.32-279.5.2 (from RHBA-2012-1199) or later.
   RHEL 6.4GA and later: already contain the fix for this issue