Discussion:
CMBs Shortage - How to solve an issue?
(too old to reply)
Gilson Cesar de Oliveira
2017-12-22 15:31:08 UTC
Permalink
Dear list:

We are facing some alerts related to shortage of Console Message Buffers due to an automation that change and route sysouts from one site to another. This automation, when there are a lot of sysouts, generates a lot of HASP686 messages.

In order to fix the issue we have two ways:

1- Increase the NUMBUF in CONDEF (today this number is 270);
2- Suppress the message HASP686 from the console through the MPFLSTxx.

I'd like to hear from the experts wich one would be the best.

Thanks in advance for any help.

Regards,

Gilson Cesar

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to ***@listserv.ua.edu with the message: INFO IBM-MAIN
Ed Jaffe
2017-12-22 16:03:48 UTC
Permalink
Post by Gilson Cesar de Oliveira
We are facing some alerts related to shortage of Console Message Buffers due to an automation that change and route sysouts from one site to another. This automation, when there are a lot of sysouts, generates a lot of HASP686 messages.
Haha! When I first saw this post, I was stunned that anyone could run
out of Channel Measurement Blocks (CMBs)!

The JES2 $HASP050 message indicates a *temporary* condition that will
eventually clear itself. Having said that, storage is cheap these days
and 270 is considered a very small number. Personally, I would raise
that to whatever value you need to avoid shortages.

Despite BUFNUM=100 being the default, IBM recommends you "[start with]
BUFNUM=1000 as a minimum starting value and increase the value using the
$T CONDEF command if the $HASP050 message indicates a shortage. Because
console messages must wait for buffers, too small a value slows the system."

https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.hasa400/has2u60047.htm
--
Phoenix Software International
Edward E. Jaffe
831 Parkview Drive North
El Segundo, CA 90245
http://www.phoenixsoftware.com/

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to ***@listserv.ua.edu with the message: INFO IBM-MAIN
Edward Gould
2017-12-22 20:22:16 UTC
Permalink
Post by Gilson Cesar de Oliveira
We are facing some alerts related to shortage of Console Message Buffers due to an automation that change and route sysouts from one site to another. This automation, when there are a lot of sysouts, generates a lot of HASP686 messages.
1- Increase the NUMBUF in CONDEF (today this number is 270);
2- Suppress the message HASP686 from the console through the MPFLSTxx.
I'd like to hear from the experts wich one would be the best.
Thanks in advance for any help.
Regards,
Gilson Cesar
A long time ago we had similar issues not with the shortages of CMB’s though. For a long time we would route a lot of jobs through NJE from one node to another as that is where the printers were (long story).
The problem that caused us was a slow down in the 3705 and we had to put pacing in the PEP gen. I don’t understand why you just don’t increase the number of CMB’s?

Ed
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to ***@listserv.ua.edu with the message: INFO IBM-MAIN
Loading...