On Mon, 2006-06-19 at 02:18 -0500, Tom Rusnak wrote: > Greetings Mark Thomen and all, > > I'm in the process of tuning our catalogs. In looking at the results of a > F CATALOG,REPORT,PERFORMANCE command I see a huge difference in the amount > of time to process a BCS ENQ between systems in a GRS-plex and those that > are not in a GRS-plex.
Catalog performance in a (non Parallel Sysplex) GRS-plex is a bitch. Integrity seems the (developers) primary goal, and then presumably performance gets a look in. Pretty hard to argue with, but an issue none-the-less. If you have catalogs on shared volumes and you convert the IGGV2 you *will* generate GRS storms. Period. Isolating catalogs to stands alone volumes and EXCL'ing IGGV2 will alleviate this enormously, and improve performance. Doing all this, and using VLF will (probably) not get your numbers back to levels approaching your non-ring numbers. If you can guarantee the devices will *NEVER* be brought online to more than one system, maybe change them to non-shared volumes (in HCD). I'm guessing the improvement will be tremendous. I've never been game, because the exposure is too severe should the unthinkable happen. If you have a shared MCAT, you will always have a significant residual block you can't do anything with. Yes I know it's read only and 100% cached - supposedly. Go run the GRS monitor for a while and see what's happening. Shane ... ---------------------------------------------------------------------- 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
