Friday night, the operator re-IPLed our VM system to load a new nucleus. The IPL was a simple SHUTDOWN REIPL. The OPERATOR id came up as it should. The messages regarding SPOOL were normal. A normal warm start seemed to have been done. However, none of the normal servers started. The GCS group comprised of GCS, VTAM and RSCS, was absent. TCPIP and all of its friends were not there. VM:Secure was started by AUTOLOG1. It initialized and then started AUTOLOG2. Right after the IPL confirmation message for AUTOLOG2, the system went dark. There is not a single message in any console log for a twenty minute period. There were messages in the logs of the ids auto-started by the system and VMSECURE right up to the time that darkness fell. All were normal start-up messages for the particular servers. Oddly, commands could be entered from OPERATOR; however, those commands are not in the spooled console log. We could QUERY anything. We could autolog users. We could even SMSG servers that we xautologged and get responses. None of this showed up in spooled console logs.
Finally, I told the operator to do a full SHUTDOWN and IPL from the HMC. When that was done, the system came up normally. One thing we know is that this was not due to a shortage of spool space. We have 9 full pack 3390-03s allocated for spool, which was 39% used. There were acres of space available. Allocation of the two dump files was not a problem as we have 2 additional volumes, one of them a 3390-09 having 14000+ cylinders, dedicated to Dump. Has anyone else seen anything like this, or even have a plausible explanation for it? Regards, Richard Schuh
