Kees, --- <snip> --- I suppose you select the icons on the HMC and do a deactivate for them. We concluded that our CF's were activated before the z/OS LPARS and in our case, DB2 immediately started recovery for its lost structures. Shortly after that, the z/OS LPAR was deactivated and on a restart the found situation was not as we expected, because DB2 had already started recovery and was interupted in that process. ---</snip> ---
I payed attention to this during earlier tests (systems and dasd failure) but I am not sure how we have done it this time. I'll try again. I haven't look in the operlog (operlog structure was in the lost CF and was never rebuilt), just in the syslogs. --- <snip> --- We found out that we had to deactivate the z/OS LPAR first and had to deactivate the CF LPAR within the XCF timeout value, to prevent other systems in the sysplex from starting recovery actions. This really simulated a full machine outage as expected. --- </snip> --- Thank you, Kees. Zaromil ---------------------------------------------------------------------- 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
