Found h2susb78 locked up as of 21:02:43 PDT Aug. 28.  Restarted, models ran about 3 minutes and communication between IOC and computer went away (one-stop card/cable issue).  The IOC one-stop card was replaced with a new revision C2 card yesterday, so suspecting an early failure, the IOC one-stop card was replaced with another one.  With this second new one-stop card, one of the main PCIe buses was not visible to the computer, and the duotone timing was off by 35uS.  

Gave up, and installed an old non-updated one-stop card back in the IOC.  System came up normally, but will have to run without problems for about 3 days before confidence begins to be restored.