There were no takers before, so here's the update anf head-up: If you're running CA Datacomm in a MUFPLEX and you have the MUF shadow on the surviving system of any outage of the system running the MUF MULTI- USER AVAILABLE member, you're going to have a broken CA11!
We have confirmed via several hard outages (sadump, deactivate of lpar without shutdown, with and without reply to ixc402D) that the MUF shadow is incapable of properly taking over for the failed system. Sadump was the initial discrvery of the problem (incredulous comment from CA: No ones takes sadumps these days!) Deactivating the system running the MUF-Available member and then re-IPLing without further ado reproduces the problem every time. We tested by not replying to IXC402D. It was impossible to get the (former) MUF shadow address space on the surviving system to terminate. CA11 came down, but the MUF did not, telling me that there was an active job named CA11. WRONG! Until I used FORCE, bypassing all CA controls to get it down. Then I replied DOWN to IXC402D. Since now *both* MUF members were restarted after the outage and before IPL, CA11 did not show the problem. We also tested with a timely DOWN to IXC402D (which would be the same as letting SFM take the system out automatically). As long as the MUF shadow is terminated/restarted *before* the failed system is re-IPL'd, you're in the clear. But if you have this scenario and do not restart the MUF shadow, expect the following: *Some* jobs (it is still unclear to me which are chosen, other than some that are submitted frequently) will show message U11-011 with different rc/rsn, even the first time around. CA11 still allows the job to run, and it runs with a cc=0. But these jobs are marked bad in CA11, and any subsequent attempt to run them results again in that message and then abendu0999. You have to somehow 'unlock' the database in order to fix the problem. Now, CA has STILL not gotten what they're doing wrong when they're not doing their sysplex cleanup properly. My colleague tells me that he feels that CA STILL don't understand what they're doing and are STILL looking in the wrong place. There is no solution in sight. How does one escalate CA problems? Best regards, Barbara ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
