Unless your session timeout clobbered them all at once, that should not have
happened. Your network monitoring should tell you if there was a momentary
drop out of the network, and it should show in the logs as well.
I'd call IBM and start a case.
--
Jim Oberholtzer
Agile Technology Architects
-----Original Message-----
From: MIDRANGE-L <midrange-l-bounces@xxxxxxxxxxxxxxxxxx> On Behalf Of Kevin
Monceaux
Sent: Wednesday, April 29, 2020 1:08 PM
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxxxxxxxx>
Subject: Re: Batch and Interactive User Jobs Hang
Jim,
On Wed, Apr 29, 2020 at 12:51:04PM -0500, midrangel@xxxxxxxxxxxxxxxxx wrote:
Once the checkpoint is reached you should not have issues.
That's what I was thinking. The backup seemed to be hung like the other
batch and interactive user jobs. We've had this system in production for
about three years with the save-while-active backups running since the
beginning. We hadn't seen an issue like this until recently, once last week
and once this week.
With the flood of CPF5140 right before things started moving I'm wondering
if a network glitch caused multiple interactive sessions to loose their
connections in mid-update, then the flood of CPF5140 messages happened when
the telnet keep alive timeout was reached and things started moving when the
hung sessions were cleaned up. The LPAR's telnet keep alive timeout setting
is currently set to *CALC.
midrange.com
