olly Beginner
Joined: 15 May 2007 Posts: 1 Topics: 1
|
Posted: Fri May 18, 2007 5:22 am Post subject: CPU-Time discrepancies from a DVIPA'd route into a CICSPLEX |
|
|
Hi,
Just wondered if anyone knows why there may be inconsistencies from the cpu-time of several CICS regions in a CICSPLEX from a DVIPA'd route. At present we have 20 CICS regions in a CICSPLEX spread over 5 LPARs that have a DVIPA inbound. The DVIPA is WLM'd and we expect to find 1 LPAR being favopured over another. We would also expect to see different cpu-times for each of the 4 regions per LPAR. What we are finding though is that 2 of our LPARs post almost identical times for their 4 regions whereas the other 3 LPARs post widely differing times for each of their 4 regions. Anyone know why this is? Is this a setting of WLM at LPAR level as well as the WLM of the DVIPA? One final point - the 2 LPARs that are almost identical have the heavier weighting out of the 5 and can therefore expect a heavier workload.
Thanks in advance. |
|