I have been investigating where some lost SMF data went. We have been intermittently losing SMF records since our first IPL after increasing the amount of memory on our z/OS 1.4 system to more than 2GiB. I *think* the issue is the one referenced in II10956:
If SMFDUMP dumps 2 MAN datasets, we lose the data from the first dataset. For example, if we dump MANA and MANB, then the summary report shows the records for MANA being dumped, followed by the summary report for MANB being dumped. However, the records that are actually on the output file are only the records for MANB, plus RT2 and RT3. Resolution: Verify that the output Dataset has a DISP of MOD,CATLG Our SMF rollup job has the output dataset of SMFDUMP with a disposition of NEW,CATLG, so I'm betting this is the problem. My question -- finally, I get to the question -- is, would the move to more than 2GiB of memory have increased the amount of SMF data enough to have exposed this issue after many years of proper behavior? Thanks, Jon ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
