From the manual we learned that for a DASD log, the size of the staging dataset would affect the dataspace of the system logger. However, if the staging dataset is too small, more offload will be needed to perform and more CPU power will be required.
The AKP frequency also affects the performance of the logger. We found that it is not easy for us to adjust the above value. Apart from the formula that IBM provides, do you have any advise about the setting for a system that has about 80 CICS regions running?
I'm sorry that I cannot be more specific about tuning the logger dataset sizes used by CICS. The formula you mention is a sizing thing but not especially useful in your circumstance.
I would not worry about changing the keypoint value, it's too important for your CICS system performance to change without a lot of consideration.
The real bottom line is that the staging datasets are critical to how and, more importantly when, offload staging occurs. Setting the high utilisation number is probably the most critical setting. Get it too high and you will cause waiting. Too low and it wastes resources.
I'd also be concerned if there is a large number of DASD shifts occurring. The extent size is probably too low in this case. If DFHLG0743 messages do not appear, then you are not getting the logtail processing which aims to reduce the size of the logger files (and this function is related to the activity keypoint frequency).
The LGDFINT SIT parm can also be reduced from the default to reduce the delay waiting for records to be put into the buffer before being force out.
Also, have a look at the logger dataspace size that relates to the amount of data in the structure or the staging dataset. This can have an effect on system paging.
I think the best thing you could do is to have a look at some real-life examples of logger settings on the Guide and Share Web sites. These actual customer type experiences are probably the best source of information.
CICS Technical Strategist -- CICS expert at Search390.com
Editor's note: Do you agree with this expert's response? If you have more to share, post it in one of our .VO7aaqqaAFk.0@/search390>discussion forums.
Dig deeper on Mainframe operating systems and management
Related Q&A from Robert Crawford
The mainframe is IT's original cloud, and there are still ways to float cloud operations onto big iron today. But is it pragmatic?continue reading
With 3270 bridge, you can't stack input messages into one structure. The bridge can't process them all at once. It takes a little more work.continue reading
CICS expert Robert Crawford offers advice on determining the connection between CICS transactions and MQ Queue name.continue reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.