× The internal search function is temporarily non-functional. The current search engine is no longer viable and we are researching alternatives.
As a stop gap measure, we are using Google's custom search engine service.
If you know of an easy to use, open source, search engine ... please contact support@midrange.com.



Hi, Tom (and everyone else participating in this thread):

See my embedded remarks below.

> Tom Liotta wrote:
...(snip)...
Take care here -- most of what I wrote was in hopes of teasing comments out from others. But, as you can see, essentially nobody has anything to add.

One important concept (that no one has mentioned so far?) is that the amount of memory specified for each pool is the amount of "real main storage" assigned to that pool. This can have very little to do with the amount of virtual memory that might actually be allocated in that pool at any given time (e.g. objects or parts of objects from Single Level Store that are paged into that pool, etc.)
I can't tell if (a) everything I wrote was correct, (b) nobody knows what the 'correct' things are, (c) there are those who know but won't join a discussion, (d)...?
Another important topic to consider is the degree of multiprogramming (also called the "multiprogramming level" or "MPL" in the OS/400 literature). This is controlled by specifying the maximum number of jobs that can run in a given subsystem.
One or more subsystems may be obtaining their memory from a given pool, and thus you can determine the maximum degree of multiprogramming that can take place in a given pool by setting the "Activity Level" for the pool.

A key idea is that OS/400 and i5/OS (and CPF) are true paging systems, whereas most other systems that support virtual memory (including MVS, VM, VSE, VMS, Unix and Linux) support one address space per job and have "swapping" that swaps out an entire address space (or at least the "important" parts.) So, in OS/400, if the MPL is too high in a given storage pool, you will almost inevitably see "trashing" where the OS spends more time paging in and out parts of objects, and less time doing any actual "work" in that pool. This can be alleviated by making the pool larger or reducing the Activity Level for the pool.
I remain baffled over the scarcity of work/performance management info. It's one area that always gets included in lists of "Why the (AS/400, iSeries, System i...) is so great." If nobody out in the world really knows, what good does it do?
I think the main reason there is not a lot of material about these topics is that they are somewhat esoteric and very technical and usually only a small percentage of OS/400 or i5/OS shops have the expertise in-house to do this kind of "tuning". I suspect that many or even most OS/400 and i5/OS customers just run with one of the "vanilla" configurations as shipped by IBM with the system (either QBASE as the controlling subsystem or QCTL.)
Tom Liotta
For anyone who is really interested, here are a few excellent references:

http://www.mc-store.com/5824.html
and
http://www.mc-store.com/5023.html

and, of course, the IBM documentation for "Work Management" in the V6R1 InfoCenter here:

http://publib.boulder.ibm.com/infocenter/systems/scope/i5os/topic/rzaks/rzaksprintthis.htm

Cheers,

Mark S. Waterbury

As an Amazon Associate we earn from qualifying purchases.

This thread ...

Replies:

Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].

Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.