home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!darwin.sura.net!paladin.american.edu!auvm!YKTVMH.BITNET!PERSHNG
- Message-ID: <IBM-MAIN%92122809583635@RICEVM1.RICE.EDU>
- Newsgroups: bit.listserv.ibm-main
- Date: Mon, 28 Dec 1992 10:42:06 EST
- Sender: IBM Mainframe Discussion list <IBM-MAIN@RICEVM1.BITNET>
- From: "John A. Pershing Jr." <PERSHNG@YKTVMH.BITNET>
- Subject: Logical Partitions/Physical Partitions
- Lines: 26
-
- In-Reply-To: Posting to IBM-MAIN dated Wed, 23 Dec 1992 22:28:00 PST
-
- The problem with these VM-versus-PR/SM discussions is that they always
- degenerate into apples-vs.-bicycles comparisons. Note that VM has many
- more capabilities, in general, than PR/SM, but these additional
- features all come at a price. E.g., minidisks (as opposed to dedicated
- packs) require CCW translation. V=V guests require much more overhead
- than V=F guests (which are handled to a great extent by the machine
- facilities). If you take your "typical" (if there is such a thing)
- PR/SM configuration and map it directly into a VM/ESA configuration
- (with a bunch of V=F guests), you end up with similar performance
- except that (1) VM costs you in license fees and a bit of extra core
- (and the alleged availability exposure), and (2) VM provides you will
- tons of extra features and stuff that you will be sorely tempted to
- start exploiting (hence, the apples-and-bicycles comparisons).
-
- Note that the original question had a lot of "maybe"s -- maybe one or
- two UTS images, maybe an AIX, maybe a "new project". I was merely
- suggesting that, maybe, VM would be superior to PR/SM in this
- situation, due to the apparent fluidity of the configuration. I'm no
- expert on PR/SM, but I was under the impression that cons-ing up a new
- parition or two, or making one of them go away was a bit involved
- (e.g., pulling new channel cables, installing new control units).
-
- John "VM Bigot" Pershing PERSHNG at YKTVMH.BITNET
- IBM Research, Yorktown Heights, NY pershng @ watson.ibm.com
-