Caveat:  "daily digest, delayed list responses, yada, yada, yada..."

Skip/Jesse/Wayne:  I have 3 items for you to validate viz. OVERFLOW volume 
selection.

Has the OVERFLOW StorGrp got OVERFLOW=YES?  If not, it's just another, normal 
group to SMS.  I've been using a *true* SMS overflow pool for years to 
eliminate x37 abends and they are selected last.  All of the volumes are 
ENABLEd but the StorGrp is marked as OVERFLOW=YES.  The latter essentially 
defaults the StorGrp as QuiNew [1] placing all the volumes as 2nd+ tier in the 
selection criteria.  (same result as Vary,SMS,StorGrp(xxx,ALL),QUIESCE,NEW)  
For certainty, I also defined the group's status as QuiNew in ISMF.  

Does the primary allocation put all the volumes in SVCDUMP over the High 
Threshold? [2]  That means SMS starts with choosing from the 2nd+ tier (that 
includes Overflow) until it finds sufficient primary space.

You should also check the StorClas entries for "Multi-Tiered SG"=Yes.  This 
indicates you want volumes from SVCDUMP preferred to SG2 and OVERFLOW when the 
StorGrp ACS routine assigns &STORGRP= "SVCDUMP", "SG2", "OVERFLOW".  It's been 
my default for a long time, now.

[1]  "Volumes in Overflow storage groups will be selected for primary space 
allocation only when all the volumes in non-overflow storage groups can not 
satisfy the allocation amount without exceeding the storage group high 
threshold."  (ISMF StorGrp help text)
[2]  "The primary list consists of online volumes that meet all the requested 
preference attributes, are below their threshold, and whose volume status and 
storage group status are enabled."  "PRIMARY THRESHOLD Volume has sufficient 
space in the target addressing space for the allocation amount without 
exceeding the storage group HIGH THRESHOLD value."   Conventional volume 
selection, Chap 7, Storage Admin Ref.

-------->  signature = 8 lines follows  <--------
Neil Duffee, Joe Sysprog, uOttawa, Ottawa, Ont, Canada
telephone:1 613 562 5800 x4585                  fax:1 613 562 5161
mailto:NDuffee of uOttawa.ca     http:/ /aix1.uOttawa.ca/ ~nduffee
"How *do* you plan for something like that?"  Guardian Bob, Reboot
"For every action, there is an equal and opposite criticism."
"Systems Programming: Guilty, until proven innocent"  John Norgauer 2004
"Schrodinger's backup: The condition of any backup is unknown until a restore 
is attempted."  John McKown 2015


-----Original Message-----
From: Schroeder, Wayne [mailto:[email protected]] 
Sent: April 25, 2017 10:33
Subject: Re: SMS STORGRP question

I have the same setup but I have my OVERFLOW volumes set as QUINEW so they are 
only used if the original SG can't hold all of the data. Hope this helps.

-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf 
Of Jesse 1 Robinson
Sent: Friday, April 21, 2017 7:43 PM
To: [email protected]
Subject: SMS STORGRP question

'Forever' we have directed DUMPSRV SVC dumps via ACS routine to STORGRP 
'SVCDUMP', which is defined to a small set of volumes called SVCDxx. We had a 
(major) problem a while back when DB2/CICS wanted to take a humongous dump that 
exceeded the available capacity of of the SVCDxx volumes. So we *added* group 
'OVERFLOW' to the ACS routine expecting this much larger group to accommodate 
dumps that SVCDUMP could not handle. Much to our chagrin, from that moment 
forward, *all* SVCDUMPs went directly to OVERFLOW without even trying to find 
space in group SVCDUMP. We have an RYO mechanism to manage dumps created on 
volume SVCDxx, so dumps created elsewhere are a problem we can live with on 
occasion but not regularly.

I know there are various ways to resolve to this problem, but I want to know 
why SMS is behaving this way and whether there is a purely SMS solution: try 
SVCDUMP first; if no room, then try OVERFLOW. In that order.

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to