On 4/21/14 1:11 PM, Jim Oberholtzer wrote:
Do yourself a favor and immediately run reclaim storage and then IPL.
You can wait for this weekend but I'm guessing you'll see some damaged
objects along the way.
Not quite immediately, but I did a RCLSTG. I brought the subsystems down
right around 5 yesterday evening, and it was ready to re-IPL by 6. A
bunch of damaged objects turned up, thankfully none of them appeared to
be permanent user objects.
We'd had a power failure one night, late last week, that might have left
a ticking time bomb in the system.
The most suspicious thing, to my eye at least, though, was a CPF0A9B
from a QYPSPFRCOL job:
Message . . . . : Probe failed for category QHTTP.
Cause . . . . . : The probe that collects data for category QHTTP data
failed. The collector will stop data collection for this category.
Recovery . . . : Refer to the job log for more details about the problem.
Start data collection for this category to try data collection again.
It happened at least 15-20 minutes after the system became unresponsive,
right about the time I first screamed for help yesterday.