Re: Z900 (2064 - 216) Problem.
TO: 1 recipientShow Details
Well this is interesting: 
Excp, Conn are the same!! Apparently there is no difference in the actual I/O.
TCB and SIO are higher. What is SIO, where does this come from? What is the 
relationship between Excp and SIO? Do you use hardware compression, can this be 
switched off after 12 hours?

Kees.


"Mehrshad Manshadi" <m_mansh...@yahoo.com> wrote in message 
news:<1345115098.18256.yahoomail...@web125603.mail.ne1.yahoo.com>...
> We have this degradation for all the jobs after 12 hours of POR and IPL.
> We are using Goal Mode.
> We are also change the volumes on other channels but the problem is same.
> 
> For test, we are using a simple DUMP JOB with following result :
> 
> Before Problem:
> 
> NP   JOBNAME  StepName ProcStep JobID    Owner    C Pos DP Real Paging    SIO 
>      BKDUMP   ADRDSSU           JOB02189 IBMUSER  A IN  C0 1421   0.00     
> 8572.2 
> --------------------------------------------------------------------------------------------------------------------------------------------
> -STEPNAME PROCSTEP    RC   EXCP   CONN    TCB    SRB  CLOCK 
> -ADRDSSU              00   302K  13318    .31    .00                      .6 
> 
> 
> ===============================================================================
> 
> After Problem : After 12 hours 
> 
> NP   JOBNAME  StepName ProcStep JobID    Owner    C Pos DP Real Paging    SIO 
>      BKDUMP   ADRDSSU           JOB02100 IBMUSER  A IN  C0 1421   0.00      
> 3200.2 
> -------------------------------------------------------------------------------------------------------------------------------------------- 
> 
> 
> -STEPNAME PROCSTEP    RC   EXCP   CONN    TCB    SRB  CLOCK  
> -DELET1               00     11      3    .00    .00     .0  
> -ADRDSSU              00   302K  13059   1.56    .01                      1.6 
> 1
> 
> Best regards
> Manshadi
> 
> 
> ________________________________
>  From: Elardus Engelbrecht <elardus.engelbre...@sita.co.za>
> To: IBM-MAIN@LISTSERV.UA.EDU 
> Sent: Thursday, August 16, 2012 2:50 PM
> Subject: Re: Z900 (2064 - 216) Problem.
>  
> Mehrshad Manshadi wrote:
> 
> >We have a parallel Sysplex with 2 members.one of them is a CPU Z900 (2064 
> >-1C5) with Z/OS 1.8 as operating system and the other is CPU Z900(2064-216) 
> >with the same operating system.
> 
> z/OS 1.8 is unsupported. 
> 
> >When we power on reset the CPU and IPL from that,It works OK. for about 12 
> >hours and after that it degrades and the jobs take much more times to 
> >complete.(about 4 times).
> 
> What jobs? How many jobs on what LPARs? Are they sharing workload and 
> middleware? How long does the degrading take places? 
> 
> Try moving you data to another volsers to see if the problem is with some 
> disks or channels (ESCON or FICON).
> 
> What type of workload do you have? What middleware do you have?
> 
> Can you post any SYSLOG message(s)? 
> 
> What is RMF, GRS, WLM and SMF showing you?
> 
> >The major things is decreasing No. of SIO for the jobs.after that even REIPL 
> >is not resolve the problem and we have to POWER ON RESET the CPU and REIPL 
> >for the next 12 hours!!.
> 
> POR is too much taken in account of design of hardware and z/OS. Open a 
> ticket with IBM rep. 
> 
> Groete / Greetings
> Elardus Engelbrecht
> 


rom: Elardus Engelbrecht <elardus.engelbre...@sita.co.za>
To: IBM-MAIN@LISTSERV.UA.EDU 
Sent: Thursday, August 16, 2012 3:23 PM
Subject: Re: Z900 (2064 - 216) Problem.
 
Mehrshad Manshadi wrote:

> No this happen for all the jobs.Like ( Reorg , load ....)

Re-org is sometimes heavy on resources, but managable. YMMV.

> We are using the simple dump just for test.

Please describe your 'simple dump'. Memory dump? Dump of volsers? Is it dump by 
application or by DFDSS?

>There is no any user on the machine during these test.
>The CICS and DB2 were down.

During or before degradation?

You should really check your EREP again and make sure you get all your records 
from all LPARS. 

Oh, yes, check your Coupling Facility (with your IBM CE) if you have any 
problems.

Kees asked you lots of other questions, could you perhaps post your replies 
here?

Groete / Greetings
Elardus Engelbrecht

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

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

Reply via email to