Joel, As we were setting up the new system, splitting the dasd between Test and Prod, we frequently received the message during IPL asking which UCB needed to left offline. I expected that until I got my IODF correct. But I have SYSLOG proof that both UCBs are online at the same time, with the same VOLSER. I have not heard back from IBM yet, will let you know if I hear anything.
Steve On Mon, 4 Oct 2010 07:56:07 -0500, Joel C. Ewing <[email protected]> wrote: >But if this did indeed happen as described, you haven't answered the >basic question about how this could have been allowed by MVS. MVS is >not supposed to allow two volumes with the same volser online at the >same time, as a volser is supposed to be a unique identifier of an >online physical volume. It makes no difference if both are shared and >accessible at IPl time. If two identical volsers are found that should >be online, I have never seen a case where MVS did not complain and ask >which of the two should be placed offline before allowing the MVS start >up to proceed. Somehow MVS would have to be convinced that both were >the same physical device, which would require something really bizarre, >like MVS thinking the address of one was a PAV alias address for the >other, and information received from the DASD subsystem ought to >preclude that from happening. > Joel C Ewing > >On 10/04/2010 07:22 AM, Richbourg, Claude wrote: >> Check your IODF definition and see if the volumes are defined as >> 'Shared' and set to come online at IPL time. >> If so, when you vary them offline before an IPL, they will come online >> at IPL time. You may need to search for DASD volumes that are only >> accessible from the LPAR you wish them to be on. BTDT and it will make >> your hair gray. >> >> Hope that Helps, >> Claude >> >> --Snip -- >> >> I am posting to the list, while I go search IBM for the cause and/or >> answer, in >> case someone has the answer already. Upgraded to z/OS V1.11 on new >> 2098- >> K02, with new IBM DS6800 dasd array. Have been testing on Test LPAR >> since >> May. Night of actual cutover to new machine, did IPL, started >> everything and >> it all went mostly well (will not discuss testing protocol here). Did >> IPL >> Saturday night and things have been hosed up since. Lots of wierd >> little >> problems, all related to datasets on disk. Got called tonight, apps >> group >> unable to resolve issue. I start looking and find one of my production >> volumes >> online. Only problem, I find the Test LPAR version of this same volume >> online >> too. They have the same volser. I have been back through the log, >> looking >> at all of the IPL messages, nothing related to duplicate volser (which I >> am >> used to). Thanks in advance for your help. >> Steve >> > > >-- >Joel C. Ewing, Fort Smith, AR [email protected] > >---------------------------------------------------------------------- >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 ---------------------------------------------------------------------- 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
