On 2016-10-27 15:01, Jim Mulder wrote:
For master trace, I prefer something like 500K or 999K. Even the maximum of 999K is a pretty small amount of storage at today's prices, and a larger MTRACE provides the convenience of sometimes not needing to figure out where the syslog is archived. And for standalone dumps, a better chance that you won't need to dig around looking for buffers of stuff that has wrapped out of MTRACE but hasn't yet been hardened in syslog or operlog.
I agree enthusiastically with the above. With dumps we receive, it is very common that the relevant messages are outside the range of what is in master trace.
The minimum and default (since z/OS 1.10) of 1M per CPU for system trace is often adequate, and that is what is used in many of the dumps I see. (As of z/OS 2.1, IPCS SYSTRACE STATUS will display this information). Larger system traces come with some trade-offs.
The key phrase there is "often adequate". We are able to find what we want from system trace the vast majority of times, but not 100% of the time. Having said that, I would be very surprised to encounter a situation where 15M is not overkill.
As a motherhood side note, always remember to increase system trace size if you turn on branch tracing to help debug a problem. If you need the other trace entries for context, the volume of branch tracing will reduce the amount of other trace data available in a trace table of a given size.
-- Regards, Gord Tomlin Action Software International (a division of Mazda Computer Corporation) Tel: (905) 470-7113, Fax: (905) 470-6507 ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
