>>> >>> We're about to IPL our new Z machine and the discussion >>> turned to MIM and the fact that the LPAR about to be IPL'd >>> won't be protected. We have shared DASD and shared User >>> Catalogs. SHROPTIONS are defined as (3,4). >>> >>> Are we exposing ourselves to any risk of catalog corruption?
<snip> >> >>I agree that, in general, corruption of datasets can occur. But I must >>disagree that a CATALOG could be corrupted. Perhaps a DFP person will >>clarify. I know that I have run without any cross-system serialization >>product for a long time with no CATALOG corruption. >> > >That is probably because you let your RESERVEs remain as RESERVEs or >you are just lucky. If you have a GRSplex or a MIIplex and convert >SYSIGGV2 (** which goes against the grs ring or MII without the control >file in a CF recommendation) from a RESERVE to a global ENQ and then >share the catalogs with a system that is not part of the plex, you are >in danger. The system not in the plex will reserve the catalog when >needed but nothing will stop that system from updating the catlog >while the systems in the GRSplex/MIIplex may be updating it. > >** Note that with GRS star or MII with a CF control file, many shops are >converting all reserves to global ENQs (this is required for GDPS BTW), >so IPLing a system sharing the catalogs into that environment that >is not part of the integrity manager plex could lead to corruption. > I only addressed catalogs in my append as I was responding to a previous post. What I wrote about could also apply to SYSVTOC - which should be treated the same as SYSIGGV2 (either convert both or let both stay as RESERVEs). So if SYSVTOC was being converted in your GRS/MIM environment and you were sharing DASD in your test environment that wasn't part of the GRS/MIM plex, you could corrupt the VTOC. If SYSVTOC and SYSIGGV2 are left as reserves, then you could IPL your test environment a feel a little better. But note that you would have no data set level protection. You could be editing a PDS member or flat file (or deleting it) from the test environment while the other systems were using it. So, can you do this? Yes, many do (as you can see from some of the responses). You need to be very careful, know your environment and hope you don't shoot yourself in the foot. The OP said "shared DASD" but didn't say what might be shared when IPLing the test LPAR. I have been at more than a few small shops that operate this way. However, there are good reasons you'll find warnings against doing this!! (I won't quote Shmuel here). Search the archives... this subject has been covered numerous times before. Mark -- Mark Zelden Sr. Software and Systems Architect - z/OS Team Lead Zurich North America / Farmers Insurance Group - GITO mailto:[EMAIL PROTECTED] z/OS and OS390 expert at http://searchDataCenter.com/ateExperts/ Mark's MVS Utilities: http://home.flash.net/~mzelden/mvsutil.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
