home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!utcsri!torn!nott!dgbt!netfs!ub!zaphod.mps.ohio-state.edu!darwin.sura.net!paladin.american.edu!auvm!VCUVM1.BITNET!SSTSCES
- From: SSTSCES@VCUVM1.BITNET (Charles Scott)
- Newsgroups: bit.listserv.ibm-main
- Subject: MVS Performance
- Message-ID: <IBM-MAIN%92111213174325@RICEVM1.RICE.EDU>
- Date: 12 Nov 92 19:17:43 GMT
- Sender: IBM Mainframe Discussion list <IBM-MAIN@RICEVM1.BITNET>
- Lines: 23
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
-
-
- *** Reply to note of 11/12/92 13:57
- Rob, You're right. Big question. I'll give you my opinion which is based
- on working in several size shops from 3 to 100 system people. Take what
- you like and feel free to ignore the rest. It's just one man's opinion.
-
- The size is important because a large shop just takes more time to keep
- up with. There are shops that don't bother to plan or can't and just wait
- for the crisis to react. If you can do that, keep your customers happy, and
- buy more when you need it, a minimal performance/capacity management effort
- is OK. I tend to support the separation of system maintenance and daily hands
- on work from performance work. It is just too tempting for most people to be
- objective when looking to justify something 'new'.
-
- One thing most shops fail to do is define when to tune. It sounds too simple
- but you can spend a lot of time tinkering and getting very little done. Some
- base tuning efforts on customer complaint, staff opinion, or something else.
- This can end up with an inconsistent effort. If you can, set guidelines for
- tuning. Service levels like: online response time, print queue times, batch
- execution times, etc. Internal guidelines are also useful like : disk
- service times, paging rates, swap time, etc.
-
- Hope this helps. I'll be glad to discuss if you want to mail me directly.
-