RE: Cost of Activating Second Processor

["Steinmetz, Paul" <PSteinmetz@xxxxxxxxxx>]

Probably not, 	10 batch jobs running simultaneously, still only using .5 CPU
We are mostly RPG, which is all single threaded.

-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Charles Wilt
Sent: Wednesday, June 05, 2013 10:46 AM
To: Midrange Systems Technical Discussion
Subject: Re: Cost of Activating Second Processor

A single batch job _could_ benefit...

But only if it's using SQL and the system/job configured to multi-thread SQL processing...
http://pic.dhe.ibm.com/infocenter/iseries/v7r1m0/topic/rzajq/smp0001.htm

To get the most benefit, you may need to purchase the DB2 Multisystem (SMP) LPP.

Charles


On Wed, Jun 5, 2013 at 10:21 AM, Vern Hamberg <vhamberg@xxxxxxxxxxxxxxx>wrote:

> Having an extra core will not help a single-threaded application. So a 
> single batch job may not see much benefit, if any, unless it tosses 
> off other batch jobs, perhaps.
>
> So it could be beneficial to run several batch jobs at the same time, 
> so long as they CAN run simultaneously. That's the catch of threads - 
> you have to check if submitted/spawned processes are done.
>
> Java will generally benefit, since it is multi-threaded. A scenario 
> with lots of jobs will likely benefit - interactive comes to mind.
>
> At least, that's my understanding.
>
> HTH
> Vern
>
> ----- Original Message -----
> Sam,
>
> We run batch jobs every month for our clients, so what was done was we 
> took note of how long some of these jobs ran.  Compared it with the 
> same jobs after the addition of the 2nd core and saw no change.  I 
> know, not the best comparison.  That's why I was suggesting running 
> some kind of system metric beforehand.  First to have something say, 
> "Yeah, you could see a benefit from another core."  Second, as you're 
> saying, you'll have something to compare to.
>
> I don't recall the exact impetus for getting the extra core.  It was 
> ordered along with extra disk space (which was really needed).
>
> -Kurt
>
> -----Original Message-----
> From: midrange-l-bounces@xxxxxxxxxxxx [mailto:
> midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Sam_L
> Sent: Tuesday, June 04, 2013 6:36 PM
> To: Midrange Systems Technical Discussion
> Subject: Re: Cost of Activating Second Processor
>
> Kurt,
>
> Do you have any metrics from before activating the second processor 
> that you will compare against?
>
> If we get to test a second processor I'd want to have something better 
> than "it feels faster" or "doesn't seem to have made any difference".  
> I suppose maybe the elapsed time for our month-end process might be a 
> good benchmark, but if you have an approach that you can share...
>
> It's not that we have a sluggish machine--compiles are still 
> satisfyingly fast, but some of our .NET apps that connect to the 
> iSeries seem to be taking longer to get going, and the boss sees that.  
> Of course, the network guys have virtualized all the Windows servers 
> and that might be something to do with it.  But it is seems very 
> difficult to do any measurement on a mixed environment.
>
> Sam
>
> On 6/4/2013 4:44 PM, Anderson, Kurt wrote:
> > Hi Sam,
> >
> > I don't have any answers for you, but we recently activated a second
> processor and so far we haven't noticed any impact.  I wasn't a part 
> of the decision to get the additional processor, but there must be a 
> way to get some metrics on your system to see if a second processor would provide an
> ROI.  Make sure you're up-to-date on PTFs.   We're working on the PTF thing
> (just about there) and the next step is that there are some reports 
> that can be ran that cost a couple hundred bucks (so I hear).
> > -Kurt
> --
> This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing 
> list To post a message email: MIDRANGE-L@xxxxxxxxxxxx To subscribe, 
> unsubscribe, or change list options,
> visit: http://lists.midrange.com/mailman/listinfo/midrange-l
> or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take 
> a moment to review the archives at http://archive.midrange.com/midrange-l.
>
> --
> This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing 
> list To post a message email: MIDRANGE-L@xxxxxxxxxxxx To subscribe, 
> unsubscribe, or change list options,
> visit: http://lists.midrange.com/mailman/listinfo/midrange-l
> or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take 
> a moment to review the archives at 
> http://archive.midrange.com/midrange-l.
>
> --
> This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing 
> list To post a message email: MIDRANGE-L@xxxxxxxxxxxx To subscribe, 
> unsubscribe, or change list options,
> visit: http://lists.midrange.com/mailman/listinfo/midrange-l
> or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take 
> a moment to review the archives at 
> http://archive.midrange.com/midrange-l.
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list To post a message email: MIDRANGE-L@xxxxxxxxxxxx To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take a moment to review the archives at http://archive.midrange.com/midrange-l.