On Wed, 13 Sep 2006 15:04:35 -0500, Glenn Miller <[EMAIL PROTECTED]> wrote:
>Hi Dave, > >We use the EMC Timefinder/SNAP and EMC/SRDF. I believe we have experienced >the 'issue' you have >described based on a couple of assumptions. I am assuming that when you >say: "SRDF comes along Yes, I believe that is correct. I don't manage the SRDF part and I'm not real familiar with the details of the process. It is as you've outlined the target volumes are in HELD status and a CONFIG RELEASE(YES) must be issued to allow the SRDF process to execute a consistent split. >We have had dialogs with EMC regarding this issue because the RELEASE >requirement was not documented Another one of my complaints. We did have some useful conversations with one of the EMC techs though, so we might get some action out of it. I'll let you know if I hear anything good. > >Our solution to the problem was to issue the Timefinder/SNAP CONFIG RELEASE >command via our D/R replication >software we wrote in-house to support the SRDF & Timefinder/MIRROR ( the >BCV's on the D/R DASD unit ). > So you track which volumes you've SNAP'd to ??? You only execute this as one large process ??? And release everything at the end ??? We're using the dataset SNAP daily for non-DR purposes. The target volumes change. They're in a pool that might be the target of multiple SNAPs that occur in parallel. Our needs are to release volumes that were the target of a SNAP as long as there are no other SNAP processes running that also target that volume. And probably iteratively preform that type of release as SNAP processes end. We need the HOLDs to assure consistent and usable DR data. Another disappointment concerns the latest software release from EMC. When we first experienced this problem, EMC pulled a MicroSoft on me and said I had to upgrade the software (I hadn't upgraded in a while in anticipation of a possible platform change decision). I upgraded everything. Had a nightmare dealing with their website downloading software. Finally got everything downloaded, installed and implemented. The last time we had a problem, I ran the RELEASE command and it failed. We're now at TimeFinder 5.4.0 and TimeFinder SNAP 5.6.0. At least the previous version let me manually release the volumes after ensuring no SNAPs were running. Now we have to have the EMC hardware team dial in to the array and do it for us. ---------------------------------------------------------------------- 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
