From consult@TC.Cornell.EDU Tue Sep 3 06:45:52 1996 X-UIDL: 3c3a33f203049f0d9fa9b1790ed3095b Return-Path: Received: from theory.tc.cornell.edu by physics.ucla.edu (SMI-8.6/SMI-SVR4) id GAA24136; Tue, 3 Sep 1996 06:45:51 -0700 Received: from localhost (localhost [127.0.0.1]) by theory.tc.cornell.edu (8.6.9/8.6.6) with ESMTP id JAA51293 for ; Tue, 3 Sep 1996 09:45:52 -0400 Message-Id: <199609031345.JAA51293@theory.tc.cornell.edu> To: tsung@physics.ucla.edu (Frank S. Tsung) Subject: Re: jobs vanishing from the queue, and other questions... In-reply-to: Your message of "Fri, 30 Aug 1996 01:46:30 PDT." <199608300846.BAA21519@raman> Reply-to: Online Consultant Date: Tue, 03 Sep 1996 09:45:48 -0400 From: Online Consultant Content-Length: 1291 Status: RO Frank, When the downtime occurs on Wednesday, any jobs running or queued are killed. However, jobs placed on hold will continue to remain in the queue. To place your job on hold, you must use the llhold command. Please read the man page on llhold for more information. You must remember to release your job after placing your job on hold. To determine the amount of physical memory for a RS/6000, enter: lsattr -E -l sys0 -a realmem To determine the amount of physical memory for a SGI, enter: hinv John Lee CTC Consulting --------forwarded message-------- >Hello: > >I have a couple of questions.... > >(1) I have noticed that sometimes my LoadLeveler job would >disappear from the queue around Wednesday morning when >the SP2 is being worked on. It puzzles me because >(i) There is no "output" and "error" files so I >assume the job was cancelled before it had a chance >to begin. (ii) When I check the queue on Wednesdays, >there are other jobs which were submitted before Wednesday, >and they were not killed. Is there a way to prevent my job >from being canceled on Wednesday mornings when the machine >is down? > >(2) How much main memory is on the RS/6000 "rocky", "jafar", > and how about the SGI "hack" and "slash"? > >thanks a lot for your help.. > >Frank