Pinned topic CTG V7.1 MIP usage
Answered question This question has been answered.
Unanswered question This question has not been answered yet.
Hi - just a general interest query/question - CTG V7.1 is zAAP eligible, which is great. However, we seem to have seen an 'unbalanced' move from GP to zAAP MIPs. On completion of our migration to V7.1, we have produced a MIP usage chart, and the results are quite interesting - rougyly 100 GP MIPS reduction, against a 250 zAAP MIPs increase - with no extra applications using the CTGs. Is this what we should expect?
Updated on 2011-12-01T09:39:37Z at 2011-12-01T09:39:37Z by RobJones
Re: CTG V7.1 MIP usage2011-06-23T13:48:55ZThis is the accepted answer. This is the accepted answer.Hi,
Can you please confirm that you are comparing V7.1 entirely on CP vs V7.1 entirely on zAAP+GP ? Or are you comparing V7.1 on zAAP+GP with an older version entirely on GP?
Re: CTG V7.1 MIP usage2011-07-28T09:24:31ZThis is the accepted answer. This is the accepted answer.Hi Rob
The chart was produced and covered a period from May 2010 to June 2011. Overt his time, we were migrating applications from CTG V5.1 over to CTG V7.1., with the rollout being completed about March time. For this period, the chart shows the MIP usage for both GP and ZAAP - but separately, if you know what I mean. Which is how we see a reduction in the GP and the increase in ZAAP - but the numbers don't match. It's obviously not a problem as such, I was just curious as to the mismatch. I would have thought that V7.1 would have been a more efficient beast, and was expecting a reduction in mip usage.
Re: CTG V7.1 MIP usage2011-08-03T07:20:40ZThis is the accepted answer. This is the accepted answer.
- shadavamy 100000397M
I have been discussing your post with colleagues and we have a few points that we would like to clarify:
1. You mentioned that the version of CICS TG previously used (5.1) was ineligible for zAAP offload. It is actually the JVM that is either eligible or not, zAAP was available from Java 1.42 onwards, and CICS TG 5.1 was compatible with Java 1.42. Were you using Java 1.41 or earlier with CICS TG 5.1?
2. What data/tooling are you using to measure the CPU consumption? Is it RMF based? Could we see the data?
3. The MIPS difference you quoted - was that an average taken over a long time, or based on a snapshot during typical workloads?
I will perform a similar analysis in the lab when I can, but it may be some time before I can, due to other commitments.
Re: CTG V7.1 MIP usage2011-08-25T13:43:31ZThis is the accepted answer. This is the accepted answer.
- RobJones 060000F25C
I will try and answer your questions, below:
1. Our V5.1 CTG used J1.3.1SR25.
2. Our guys used the SMF30 records to produce the reports.
3. I have attached a copy of the report for your perusal.
Thanks for looking at this - no great urgency for reply.
Re: CTG V7.1 MIP usage2011-12-01T09:39:37ZThis is the accepted answer. This is the accepted answer.
- shadavamy 100000397M
Thank you for the data, now we have completed work on CICS TG V8.1 I will look at this question more closely with the lab performance team.