WHEELER@CC.UTAH.EDU (05/25/88)
Posting-Version: unknown; site unknown Subject: What resources does an idle process tie up ... or why have Here are a few reasons that I know of for having an idle process killer. (Please note, none of the system's I'm in charge of use an idle process killer) 1) User's don't log out, someone use's thier account (student's in University environment) 2) You really want the user's list to reflect who is actually working on the system. 3) You install new software and the user who never logs out doesn't get the new DCL table starts complaining. (Hassle factor) 4) You re-install the DCL tables (after installing any layered product) and all the user's who haven't logged out are still mapped to the old installed DCL table and hence you use up some global pages. (I once didn't have enough global pages to continue the installation and hence had a machine that one couldn't log into! STOP/ID cured that problem) By the way, does anyone *really* make all the user's log off and shutdown the network to install new products? Bob Wheeler bob%howard@cc.utah.edu
EPRF@SNYCENVM.BITNET (Peter Flass) (05/27/88)
The reason we run an idle process killer is to keep people from logging
in and tying up a dial-up port indefinitely. Of course if you have lots
of ports....
- Pete FlassWHEELER%CC.UTAH.EDU%KL.SRI.COM%lbl%sfsu1.hepnet@LBL.GOV (05/28/88)
Received: from KL.SRI.COM by LBL.Gov with INTERNET ;
Thu, 26 May 88 19:51:02 PDT
Received: from CC.UTAH.EDU by KL.SRI.COM with TCP; Tue 24 May 88 23:52:10-PDT
Date: Wed, 25 May 88 00:51 MDT
From: WHEELER@CC.UTAH.EDU
Subject: What resources does an idle process tie up ... or why have
To: info-vax@KL.SRI.COM
X-VMS-To: in%"info-vax@kl.sri.com"
Posting-Version: unknown; site unknown
Subject: What resources does an idle process tie up ... or why have
Here are a few reasons that I know of for having an idle process killer.
(Please note, none of the system's I'm in charge of use an idle process
killer)
1) User's don't log out, someone use's thier account (student's in University
environment)
2) You really want the user's list to reflect who is actually working on
the system.
3) You install new software and the user who never logs out doesn't get
the new DCL table starts complaining. (Hassle factor)
4) You re-install the DCL tables (after installing any layered product)
and all the user's who haven't logged out are still mapped to the old
installed DCL table and hence you use up some global pages. (I once
didn't have enough global pages to continue the installation and hence
had a machine that one couldn't log into! STOP/ID cured that problem)
By the way, does anyone *really* make all the user's log off and shutdown
the network to install new products?
Bob Wheeler
bob%howard@cc.utah.edu
cfchiesa@bsu-cs.UUCP (Christopher Chiesa) (05/29/88)
In article <8805270320.AA07440@ucbvax.Berkeley.EDU>, WHEELER@CC.UTAH.EDU writes: > Posting-Version: unknown; site unknown > Subject: What resources does an idle process tie up ... or why have > > Here are a few reasons that I know of for having an idle process killer. > (Please note, none of the system's I'm in charge of use an idle process > killer) > > [you've seen the list, so I'm omitting it] > By the way, does anyone *really* make all the user's log off and shutdown > the network to install new products? > > Bob Wheeler > bob%howard@cc.utah.edu Here at Ball State University, the entire four-node Vaxcluster is completely shut down from 5PM Friday night til 8AM Saturday morning for backups, main- tenance, and who knows what else (I've been a student so I don't know the inner arcane details). Software installation is done either at THAT time, or in an even MORE extended shutdown (this whole holiday weekend, for example) during inter-quarter Breaks... Of course, this is a HUGE system (comments on this next quote are solicited): our system manager quotes a figure of about 25,000 users... (Not all at the same time, of course, but that's about right, in terms of the number of accounts living on the system, by my own experience.) Chris Chiesa Ball State University -- UUCP: <backbones>!{iuvax,pur-ee,uunet}!bsu-cs!cfchiesa cfchiesa@bsu-cs.UUCP
cfchiesa@bsu-cs.UUCP (Christopher Chiesa) (05/29/88)
In article <3199@bsu-cs.UUCP>, I wrote: > Here at Ball State University, the entire four-node Vaxcluster is completely > shut down from 5PM Friday night til 8AM Saturday morning for backups, main- > tenance, and who knows what else... > > Of course, this is a HUGE system (comments on this next quote are solicited): > our system manager quotes a figure of about 25,000 users... (Not all at the > same time, of course, but that's about right, in terms of the number of > accounts living on the system, by my own experience.) I'd meant to add that typical system use is between 180 and 200 interactive users, during "peak" periods that I'VE seen personally. Chris Chiesa -- UUCP: <backbones>!{iuvax,pur-ee,uunet}!bsu-cs!cfchiesa cfchiesa@bsu-cs.UUCP