We have been working on the hardware node to indicate the cause of the issue.
Unfortunately moving to the slave server does not fix the issue as the iowait is pretty high on it as well due to bad drive.
The primary server has a bad drive inside of it as well.
The bad drive on the primary server has been disconnected for now and the sites should be loading faster a bit now.
Meanwhile we are in touch with the DC techs to replace the bad drives on both servers.
Friday, May 11, 2012