Re: DFHSM multi-host environment

2006-05-19 Thread O'Brien, David W. (NIH/CIT) [C]
) If your level of software supports it, ONLYIF will allow you to have 1 ARCCMD member for all LPARs. -Original Message- From: Bob Henry [mailto:[EMAIL PROTECTED] Sent: Thursday, May 18, 2006 4:47 PM To: IBM-MAIN@BAMA.UA.EDU Subject: Re: DFHSM multi-host environment The following is

Re: DFHSM multi-host environment

2006-05-18 Thread John Ticic
-- snip --- Each DFHSM has a unique hostid and both have HOST=MAIN, CDSR=YES and CDSSHR=YES defined in the startup parms. As the book suggests, I am running both images with the same ARCCMD00 member which has both automigration and autobackup defined for most of the shared volumes. Aside from the

Re: DFHSM multi-host environment

2006-05-18 Thread Edward Jaffe
Unashamed? Never knew you to feel ashamed about anything! Course there's always the stuff I know nothing about ... O:-) Shane wrote: On Fri, 2006-05-19 at 04:54 +0800, Ron and Jenny Hawkins wrote: If he has Device Reserves then he has serialisation, right? *only* if ALL the updates

Re: DFHSM multi-host environment

2006-05-18 Thread Shane
On Thu, 2006-05-18 at 16:47 -0400, Bob Henry wrote: > I appreciate your concern and if it were possible, I would work toward a > GRS environment. I have two ADCD systems connected so there are some > duplications of names, ...plexes, etc., but even more restrictive is the > FLEX-ES environment

Re: DFHSM multi-host environment

2006-05-18 Thread Shane
On Fri, 2006-05-19 at 04:54 +0800, Ron and Jenny Hawkins wrote: > If he has Device Reserves then he has serialisation, right? > *only* if ALL the updates are protected by RESERVE. Locating all the required (ENQ) documentation for all the various options for multiple instances of HSM is a serious

Re: DFHSM multi-host environment

2006-05-18 Thread Ron and Jenny Hawkins
Brian, If he has Device Reserves then he has serialisation, right? Ron > > Please, set up a valid serialization environment - first. That's my > suggestion, anyway. > > >both automigration and autobackup defined for most of the shared volumes. > >Aside from the occasional RESERVE contention

Re: DFHSM multi-host environment

2006-05-18 Thread Bob Henry
I appreciate your concern and if it were possible, I would work toward a GRS environment. I have two ADCD systems connected so there are some duplications of names, ...plexes, etc., but even more restrictive is the FLEX-ES environment they're running in where there's no coupling facility suppor

Re: DFHSM multi-host environment

2006-05-18 Thread Bob Henry
The following is from the "DFSMShsm Implementation and Customization Guide", chapter 5, "Primary Address Space Startup Procedure". Note the last sentence. Maybe I misunderstood something or took too literal an interpretation. The CMD=00 keyword refers to the ARCCMD00 member of PARMLIBs discu

Re: DFHSM multi-host environment

2006-05-18 Thread Brian Peterson
You are begging for trouble if you continue down a path of sharing data without serialization, which is what I assume if I understood your statement "I do NOT have GRS installed" correctly. Please, set up a valid serialization environment - first. That's my suggestion, anyway. Brian On Thu,

Re: DFHSM multi-host environment

2006-05-18 Thread O'Brien, David W. (NIH/CIT) [C]
>As the book suggests, I am running both images with the same ARCCMD00 member which has both automigration and autobackup >defined for most of the shared volumes. I must have missed that suggestion. Where is it documented? At this shop we run 3 LPAR's with shared DASD and a shared set of CDSs.

DFHSM multi-host environment

2006-05-18 Thread Bob Henry
I'm trying to set up a multi-host environment which shares multiple 3390 volumes. I do NOT have GRS installed. I have a DFHSM host in each image which share a single set of CDS's. One image (OS/390 2.10) is set up with DFHSM as the PRIMARY=YES host and the other image (z/OS 1.6) is PRIMARY=NO.