Basically you hit the high points of the steps that occurred during the test with one exception. I did delete and redefine the CFRM dataset writing the correct policy into it. We ipled again and the d**m system was still looking for the coupling facility on the incorrect CPU. I then did the only other thing I could think of which was to delete and redefine the sysplex couple dataset and try again.
Mark Jacobs -----Original Message----- From: IBM Mainframe Discussion List on behalf of Brian Peterson Sent: Fri 4/9/2010 2:48 PM To: IBM-MAIN@bama.ua.edu Subject: Re: SYSPLEX/CFRM Couple dataset(s) relationship I believe simply rewriting the policy was not sufficient. Please tell me if this is what you did: 1) First IPL at DR site specified CFRM CDS data set "one". This IPL specified CFRMPOL(policy_a). "Policy_a" was not correct. 2) Second IPL at DR site used the exact same CFRM CDS as in attempt 1. In this IPL, you specified CFRMPOL(policy_b) or else you specified CFRMPOL(policy_a) having first rewritten "policy_a" in the CFRM CDS prior to the IPL. This fails because the CFRMPOL parameter is only acted upon if there is NO active policy in the CFRM CDS. Here's the way I think about this, to keep this straight in my own mind: Whenever it is necessary to IPL at DR with a new CFRM policy, the CFRM CDS must be newly created. By the way, in my experience, the Sysplex CDS has nothing to do with this at all - this is strictly a matter of whether the CFRM CDS data set has an active policy in it, or if instead the CFRM CDS data set is newly allocated and empty when the new DR policy is loaded into it. Here's a good explanation of the one-shot nature of the CFRMPOL parameterm from Setting Up a Sysplex: -=-=-=-=-=- 4.4.2 Performing System Level Initialization CFRM performs system level initialization based on the state of the active CFRM policy. Valid CFRM policy states are: o The CFRM couple data set contains an active policy that is empty, that is, there is no currently active CFRM policy, either a policy that was never started or policy usage was stopped. A policy is activated either by issuing a SETXCF START,POLICY command or specifying CFRMPOL(POLICY-NAME) in the COUPLExx parmlib member used when the sysplex is IPLed. Policy usage is stopped by issuing a SETXCF STOP,POLICY command. o The CFRM couple data set contains an active policy that was properly defined with the administrative data utility and then activated either by using the SETXCF START command or the CFRMPOL parameter in the COUPLExx parmlib member. -=-=-=-=-=- At your first IPL, you can use CFRMPOL to load an appropriate CFRM policy. But, this only works once - if you make a mistake, you need to delete/define new CFRM CDS data sets (to make them empty again) and reload the policy, otherwise your next IPL will use the same policy as was activated by the first IPL. Brian On Thu, 8 Apr 2010 10:31:23 -0400, Mark Jacobs wrote: >At our last DR exercise we had an incorrect CPU serial number in our >CFRM policy which resulted in a failure in our IPL. We fixed the CFRM >policy and re-wrote it replacing the incorrect policy (same policy name). > >When we re-ipled, the system was still looking for the CF lpar with the >incorrect CPU serial number even though the CFRM policy with the old >serial number wasn't in the CFRM dataset. We couldn't get the system to >IPL until I deleted and redefined the SYSPLEX couple dataset. > >Does the sysplex couple dataset retain information about the CFRM policy >in use other than the name of the last used CFRM policy? > >-- >Mark Jacobs ---------------------------------------------------------------------- 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 ---------------------------------------------------------------------- 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