Hi Dave,
Right - you touched on another thing I'm checking on with code -- volumes
being specified in EXTENT CONTROL - but not attached to the system.   Also -
being specified in the directory but not attached.   There may be valid
reasons for these conditions, but in general, should be avoided.  Thanks!

Scott

On Tue, May 12, 2009 at 7:09 PM, David Kreuter <dkreu...@vm-resources.com>wrote:

> Hi Scott: Since the days of wide CMS usage, and device physics, I haven't
> really configured regions with any degree of sophistication. Not to say it
> isn't useful.  Maybe there's a case for if you have real mod27s or mod54s
> and you want to code a bunch 3390 mod3 size mdisks, maybe for 2nd level vm's
> then you could do some slick configuring.
>
> On another extent control note, a little nasty is that dirmaint treats
> extent control as truth.  If you code a volser cum region and there is no
> disk with that label online or attached to the system DIRMAINT will happily
> add space to it when you use autov, autor, explicit, and autog (provided the
> errant volume is in the groups section). Not nice when you really expect a
> minidisk to be there! Also good ole datamove will do the proverbially choke
> if it is told to go and format, or clonedisk to a missing volume.  A perfect
> case of your gun, bullet, and foot.  After all you told DIRMAINT about it.
> I know some out there might do this exact thing for future plans and keep
> datamove away from it, but for clarities sake I avoid it. Datamove gets
> confused enough I daresay.
>
> I support systems with hundreds of real volumes (SCSI and edev still a
> while down the road) and keeping extent control in shape remains a manual
> process.  Life's tough all over.
>
> David Kreuter
>
>
>  -------- Original Message --------
> Subject: [IBMVM] DIRMAINT 'regions'
> From: Scott Rohling <scott.rohl...@gmail.com>
> Date: Tue, May 12, 2009 8:47 pm
> To: IBMVM@LISTSERV.UARK.EDU
>
> I have never seen 'regions' defined in EXTENT CONTROL as anything other
> than the corresponding DASD volser.   I can see how it might be used if you
> were really into carving up DASD volumes strategically -- but I believe
> those days are long gone.
>
> Anyway - I'm putting together some best practices revolving around DASD,
> DIRMAINT, EXTENT CONTROL - and one thing I planned to list is to ensure that
> the regions listed match the volsers.   I've been bitten by mismatches
> (someone duplicating a line but failing to change the 2nd word -- the
> volume) and want to eliminate them.
>
> So my question is this:   Does anyone at all actually specify regions in
> EXTENT CONTROL on a DASD volume that have a different name then the DASD
> volser?   If so - is there an advantage you can tell me about?
>
> Thanks!  Scott
>
>

Reply via email to