Re: QINACTITV does not appear to be working

Dan
I'm at home so I don't have the descriptions handy (it's 7:00am here in NZ)
but this should get you started....

Batch work:

I created a subsystem called NIGHTB that was essentially the same as QBATCH
with a Jobq entry for a jobq called NIGHTB. I submit all the jobs to this
jobq and start it as part of my nightly processing. I use Job priority to
sequence nightly processing and batch jobs.

Interactive work:

I just copied the QInter description and then removed/altered the QCMDI
routing entry to  something else (QCMDX I think but the particular value is
not important). I then added this routing entry into QINTER as well, so
that the individuals with after hours access also had during hours access !
The routing entry in QINTER was in addition to the normal QCMDI entry. I
then modified the job description of the individuals to have QCMDX as the
routing data.

My workstation entry is now set to all as I let the routing entry do the
"security work" and that way I have access to any device at any time
depending on the after hours needs. Its a bitch finding out you have a new
device name or forgot a device at 2:00am....

Some other points - I pretty much run everything in *base after hours to
provide maximum memory for whatever is executing. Since my nightly job
queue is (currently) single threaded I saw no advantage to utilising a
second pool for interact for what was likely to be very limited interactive
work. While I was getting this right I regularly signed on to check whether
the nightly jobs were running with a Base and system pool and maybe a spool
one, the idea being that they run in as large a pool as possible with the
minimum contention.

Of course, I'm running on an 820 now so the loading makes tuning rather
difficult (average cpu of 10% during peak hours)

The last jobs in my nightly job stream end the NIGHTB and NIGHTI subsystems
so that they are not active (as a side note I start the NIGHTI subsystem
before I end Qinter just on the off chance that if something goes wrong....)

The only gotcha is that with workstations set to all, if NIGHTI remains
active during the day then users will sometimes get that subsystem instead
of QINTER. I haven't bothered to fix this (I can't remember how anyway)
because it is only an occasional problem and the substem should not be
active anyway.

If you need any more detail let me know.

Hope this helps

Regards
Evan Harris


>Oh, and on the pools, I have ours as:
>
>Pool        Storage     Activity
>  ID        Size (K)      Level
>   1           *BASE
>   2       *INTERACT
>
>
>"Bale, Dan" wrote:
>
> > This is a multi-part message in MIME format.
> > --
> > [ Picked text/plain from multipart/alternative ]
> > Ya know, I was just getting ready to ask that question.  End QINTER, but
> > still allow me access to sign on.
> >
> > Evan, could you detail both the subsystem and jobqueue you use for
> > backup, as well as second subsystem with the "workstation entries" (?)
> > and "routing entry" (?).  This appears to be exactly what I need/want to
> > do in our situation.
> >
> > Dan Bale
> > IT - AS/400
> > Handleman Company
> > 248-362-4400  Ext. 4952
> > D.Bale@Handleman.com
> >
> > > -----Original Message-----
> > > From: Evan Harris [SMTP:spanner@ihug.co.nz]
> > > Sent: Thursday, August 30, 2001 2:18 PM
> > > To:   security400@midrange.com
> > > Subject:      Re: [Security400] QINACTITV does not appear to be
> > > working
> > >
> > > We do the same here. I run the backup jobs in batch through a specific
> > > dedicated subsystem and jobqueue so there is no possibility of a job
> > > clash
> > > (well, OK a much reduced possibility :)
> > >
> > > To enable access to particular individuals during the backup I starts
> > > a
> > > second subsystem that has workstation entries included in it and only
> > > allow
> > > people with a particular routing entry to sign on.
> > >
> > > Hope this helps
> > >
> > > >Dan,
> > > >
> > > >We only have one system here and I notice that some devices get
> > > knocked of
> > > >(we use 90)
> > > >and yet others, which have had no activity for HOURS are still on.
> > > The
> > > >common thread
> > > >seems to be they are PC's connected via TCP/IP and not the dumb
> > > terminals (?).
> > > >
> > > >We DEFINITELY end QINTER for backups and the backup job runs in
> > > QBATCH so
> > > >no user
> > > >submitted jobs can run against it...
> > > >
> > > >Chuck
> > > >
> > > >"Bale, Dan" wrote:
> > > >
> > > > > System value QINACTITV is set to 60 minutes on all of our
> > > production
> > > > > systems.  However, the system seems to recognize this setting on
> > > only
> > > > > two of these five boxes.  One one system, it takes up to six hours
> > > for
> > > > > the system to finally boot me off.  Another system takes about
> > > three
> > > > > hours.
> > > > >
> > > > > Yesterday, our backup failed because one user's workstation had
> > > not been
> > > > > signed off, objects were locked, and the backup "failed".  This
> > > was
> > > > > quickly elevated to "resolve this now" by our auditing team.
> > _______________________________________________
> > This is the Security Administration on the AS400 / iSeries
> (Security400) mailing list
> > To post a message email: Security400@midrange.com
> > To subscribe, unsubscribe, or change list options,
> > visit: http://lists.midrange.com/cgi-bin/listinfo/security400
> > or email: Security400-request@midrange.com
> > Before posting, please take a moment to review the archives
> > at http://archive.midrange.com/security400.
>
>_______________________________________________
>This is the Security Administration on the AS400 / iSeries (Security400)
>mailing list
>To post a message email: Security400@midrange.com
>To subscribe, unsubscribe, or change list options,
>visit: http://lists.midrange.com/cgi-bin/listinfo/security400
>or email: Security400-request@midrange.com
>Before posting, please take a moment to review the archives
>at http://archive.midrange.com/security400.