Search390.com
Expert Question of the Week 
September 24, 2001

===================================================
SPONSORED BY: Postmaster Direct
===================================================
GET THE INFO YOU WANT TODAY
What do YOU want in your e-mailbox? How about FREE info and offers
about computers, networking, wireless devices, finance and lots more?
Visit  http://search390.techtarget.com/postmasterDirect/, choose the
topics that interest you, and you'll get the e-mail you want.
===================================================

---------------------------------------------------
Welcome to Search390's Expert Question of the Week newsletter.
Remember, no question is too simple for Ask the Experts! If you have
a 390-related question, send it to [EMAIL PROTECTED] Selected
questions will be answered by our experts. 

This week's question was answered by Gerhard Adam, search390's
Systems Management Expert.

THIS WEEK'S QUESTION:

Q:  Given the historical problems associated with disk enq & reserve
lockouts on OS/390 systems and the subsequent feature developments of
software monitors such as TMON, will there be an exacerbation of
these problems when PAV's are robustly implemented on large
OS/390-z/OS environments? If so, what would you suggest
operators/sysprogs do to minimize the impact? The overall scenario is
that there is a small, finite amount of time to debug a enq/reserve
lockout, prior to management decision to IPL to clear the issue. I'm
implying that a longer enq/reserve list will take a correspondingly
longer time to debug, during a live situation. Also, what pro/con
situations do you see with the new SYNCRES(yes) GRS parm that affects
the order of the enq/reserve processing?

A:  Well, let me begin by answering the last part of your question
first. The SYNCHRES parm has been available since OS/390 Rel 7 and
its primary use is to ensure that a hardware reserve is successfully
executed before a GRS ENQ is issued. The intent is to avoid deadlocks
that might occur by having an ENQ outstanding and yet be unable to
complete RESERVE processing against the device. Since the RESERVE
doesn't take effect until the first I/O is issued, this approach is a
reasonable way of reducing the potential for deadlocks.

PAV's are intended to provide simultaneous access to a volume from a
single system thus reducing IOSQ time. The primary consideration is
that the domain over which the I/O operates (ie: the extents) don't
conflict. In the case of WRITE operations, serialization will still
occur for channel commands operating within the same domain.

Multiple Allegiance is the ESS feature that deals with the issue of
I/O operations occurring across systems. In this case, there is no
problem unless the I/O conflicts with a RESERVE. If a RESERVE occurs,
the other channel commands would be queued in the ESS.

As to GRS ENQ/RESERVE processing, I don't see how adding PAV's or
other storage devices will make any difference to ENQ resolution
scenarios. Since the conflicts that may or may not occur would occur
today. I'm not sure why you think the ENQ/RESERVE list should get any
longer. Since the ESS features are designed to allow greater
concurrency of operation, this wouldn't result in higher levels of
ENQ/RESERVE competition. In other words, PAV and MA are intended to
minimize delays that already exist; not increase workloads.

------------------------------------------------------------
Gerhard Adam is this week's featured expert.  If you have a question
for Gerhard, you can submit it here: 
http://search390.techtarget.com/ateQuestion/0,289624,sid10_cid365820_tax285033,00.html

If you have performance management questions, start posting them in
our Operating Systems forum.  Remember, Senior Technical Consultant
and Performance and Workload Management Expert Steve Samson will be
live in the OS forum this Friday, Sept 28 from 3:00 pm EDT to 4:00pm
to answer your questions.  To post a question for Steve, go to: 
http://search390.discussions.techtarget.com/WebX?50@@.ee83ff6 

==========================================================
THIS WEEK'S search390 FEATURED TOPIC:  Backup and recovery
========================================================== 
Natural and sometimes not so natural disasters can bring your systems
down. But there are steps you can take to ensure a smooth recovery.
The information we've put together here will help you prepare for and
handle down times.
http://search390.techtarget.com/featuredTopic/0,290042,sid10_gci771289,00.html
==========================================================

==========================================================
If you would like to sponsor this or any TechTarget newsletter,
please contact Tim Keough at [EMAIL PROTECTED]
==========================================================


If you no longer wish to receive this newsletter simply reply to 
this message with "REMOVE" in the subject line.  Or, visit 
http://search390.techtarget.com/register 
and adjust your subscriptions accordingly. 

If you choose to unsubscribe using our automated processing, you 
must send the "REMOVE" request from the email account to which 
this newsletter was delivered.  Please allow 24 hours for your 
"REMOVE" request to be processed.

Reply via email to