Re: Objects per user

Rob
I think I've Been There, Done That!

What is probably occuring is that you are exceeding the maimum space for an 
internal table for the profile. The problem is that the user profile that owns 
an object, also contains ALL of the specific authorities for the objects it 
owns. So, the more objects there are, the more table entries are used. Then, 
for each specific authority granted to the owned object, another table entry is 
used. (I think this also holds true for objects in QTEMP!)

A simple solution would be to remove authorities or delete objects. NOT! The 
table entries that were associated with the objeect/authorities in no longer 
used, but is not recovered! And there are no commands or programs to free up 
these used entries. Reclaim Storage and IPLs won't do the trick.

When this occurred (V3R7 or V4R1), there was only ONE way to regain the table 
entries. CRASH THE MACHINE! Yup, during the integrity checking performed after 
an abnormal shutdown, the table is rebuilt, freeing the unused entry.

Although your number of entries isn't too large, if your system generates a lot 
of temporary files under the group profile, you are burning table entries.

I think I have the APAR number at the office if you need it. PLEASE - raise 
caine with your IBM rep. The code exists to rebuild the table, IBM just needs 
to figure a way for a less extreme way to execute the stuff. You are the third 
account I know of to be struck by this, Anyone else?
Bob Larkin

Rob Berendt wrote:

> We are getting a MCH2804 message on one of our AS/400's.  We checked the user 
>profile, (a group profile), and it is set to *NOMAX on disk.  However IBM is 
>leading me to believe that the problem is not in the total size of the 
>objects, but in the number of the objects.  Also factoring, (although I am not 
>exactly sure on how the factor works), in the number of user profiles using 
>that group profile.  There is a theoretical limit of around 1,000,000 objects 
>that can be owned by anyone.  This group profile owns between 200,000 and 
>300,000 objects.  There are 112 users associated with the group profile, only 
>24 of them have signed on this month.
>
> The funny thing is, is that I think this factor is limited by the number of 
>active users and not total users.  We had someone sign off and others stopped 
>getting this message.  This message appeared on many commands:  CPYFRMQRYF, 
>SAVRSTOBJ, etc.
>
> I've done no testing to see if the number of sessions per user matters.
>
> IBM has reported this problem occurring in multiple levels of the operating 
>system.
>
> Has anyone else seen this problem?  Most likely to occur in a shop where one 
>group profile owns many many objects, or, in a shop where a group profile owns 
>many objects and there are many users in that shop on the system at the same 
>time using the same group profile.
> +---
> | This is the Midrange System Mailing List!
> | To submit a new message, send your mail to MIDRANGE-L@midrange.com.
> | To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
> | To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
> | Questions should be directed to the list owner/operator: david@midrange.com
> +---

begin:          vcard
fn:             Bob Larkin
n:              Larkin;Bob
org:            <A HREF="HTTP://web.wt.net/~blarkin/">Larkin Computer Consulting</A>
adr:            <A HREF="http://web.wt.net/~blarkin/">Bob and Diana's Page</A>;;;Houston;TX;<A HREF="http://web.wt.net/~blarkin/">;United States
email;internet: blarkin@wt.net
title:          Systems Consultant
x-mozilla-cpt:  ;4104
x-mozilla-html: FALSE
version:        2.1
end:            vcard