From - Tue Aug 13 15:50:18 2002 Return-Path: Received: from listserv.fnal.gov (listserv.fnal.gov [131.225.9.3]) by castor.ts.infn.it (8.11.6/8.11.6) with ESMTP id g7DDk1S24135 for ; Tue, 13 Aug 2002 15:46:02 +0200 (MET DST) Received: from listserv (listserv.fnal.gov) by listserv.fnal.gov (LSMTP for Windows NT v1.1b) with SMTP id <0.00326F8C@listserv.fnal.gov>; Tue, 13 Aug 2002 8:46:02 -0500 Received: from LISTSERV.FNAL.GOV by LISTSERV.FNAL.GOV (LISTSERV-TCP/IP release 1.8d) with spool id 40984 for CDFSAM-ADMIN@LISTSERV.FNAL.GOV; Tue, 13 Aug 2002 08:46:02 -0500 Received: from heffalump.fnal.gov by listserv.fnal.gov (LSMTP for Windows NT v1.1b) with SMTP id <0.00326F8B@listserv.fnal.gov>; Tue, 13 Aug 2002 8:46:02 -0500 Received: from CONVERSION-DAEMON.smtp.fnal.gov by smtp.fnal.gov (PMDF V6.0-24 #37519) id <0H0S00401BKQ4C@smtp.fnal.gov> for cdfsam-admin@listserv.fnal.gov (ORCPT cdfsam-admin@fnal.gov); Tue, 13 Aug 2002 08:46:02 -0500 (CDT) Received: from castor.ts.infn.it ([140.105.6.112]) by smtp.fnal.gov (PMDF V6.0-24 #37519) with ESMTP id <0H0S00BIDBKPOM@smtp.fnal.gov> for cdfsam-admin@listserv.fnal.gov (ORCPT cdfsam-admin@fnal.gov); Tue, 13 Aug 2002 08:46:01 -0500 (CDT) Received: from ts.infn.it (axts12.ts.infn.it [140.105.6.163]) by castor.ts.infn.it (8.11.6/8.11.6) with ESMTP id g7DDjuS24051; Tue, 13 Aug 2002 15:45:56 +0200 (MET DST) Date: Tue, 13 Aug 2002 15:45:57 +0200 From: Stefano Belforte Subject: Re: Sam QUOTA problem To: r.stdenis@physics.gla.ac.uk Cc: cdfsam-admin@fnal.gov Reply-to: stefano.belforte@ts.infn.it Message-id: <3D590D95.6400D103@ts.infn.it> MIME-version: 1.0 X-Mailer: Mozilla 4.76 [en] (X11; U; OSF1 V4.0 alpha) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: Sender: owner-cdfsam-admin@LISTSERV.FNAL.GOV that's very intersting. Thanks !! After carefull read of http://cdfdb.fnal.gov/sam_doc/doc/cdf/install/install.html I think what happend is that: 1. I started with a 2GB dik cache 2. I created test as idicated 3. with a lot of effort and help by you, Gabriele and Luciano, I managed to redefine the disk from 2 to 20GB once I could physically expand it but did not realize the "group thing". I just changed the test group max quota from 2 to 20GB (sam set group --group=test --max-disk=20GB), it appears it worked :-) But, why does it say that current used quota is 2 projects and ~2GB, when all my projects (interactive processes) are terminated and I have only 64MB of data files on the cache disk ? How can I reset this ? The station dump on the web should show no project now, correct ? Stefano "Rick St. Denis" wrote: > > You have put a quota of 2G on the group test. So that is what stops > things. > > You have a physical disk of 20G and sam knows that. > > ehm. what do you want? the test group is good say at 2G. Would allow you > two single test files. If you are the only user, this is probalby ok. > > Then you want to allocate 18G to another group. But I admit, there is the > fair share biz and i dont understand that algo. > > oh, in case you are confused by the node I did this on. I set the > SAM_STATION environment variable to samtrieste-1. All sam dump station > does is go to the database via a corba server to find out what your > reality is. It depends on the database knowing the reality. Of course, > since the database defines your reality, (it is not a spoon) then we have > control. > > bahhahaha.... (evil laugh) > > rick > > [stdenis@nglas09 ~]$ sam dump station --groups > *** BEGIN DUMP STATION samtrieste-1 version v4_2_0_3 running at pccdf2 11 > days 23 hours 47 minutes 53 seconds, admins: belforte > No replica selection criteria > There are 60 authorized transfer groups > Minimum delivery is 1KB; external deliveries are constrained to (empty) > Excess consumer satisfaction: 1 > AUTHORIZED GROUPS: > group test: admins: belforte , swap policy: LRU, fair share: 1, quotas > (cur/max): projects = 2/10, disk: 2037845KB/2GB, locks:0B/0KB > > *** END OF STATION DUMP *** > > ***************************** > Dr. Richard St. Denis,Dept. of Phys.& Astr., Glasgow University > Glasgow G12 8QQ; United Kingdom; UK Phone: [44] (141) 330 5887 > UK Fax : [44] (141) 330 5881 > ====================================== > FermiLab PO Box 500; MS 318 Batavia, Illinois 60510 USA > FNAL Phone: [00] 1-630-840-2943 FNAL Fax: [00] 1-630-840-2968 > Sidet: [00] 1-630-840-8630 FCC: [00] 1-630-840-3707 > > On Tue, 13 Aug 2002, Stefano Belforte wrote: > > > I think I found the origin of my odd cache behavior and > > hang job. I have a 20GB disk for sam cache, but cache manager > > thinks it is 2GB: > > > > pccdf2:/private> sam dump station --disks > > [...] > > STATION DISKS: > > disk 23 pccdf2.ts.infn.it:/home/sam/cache1, 18535765KB/20641788KB free > > ---------------------NOTE THIS IS 20GB !-----------------^ > > [...] > > *** END OF STATION DUMP *** > > > > While in the station log file I find: > > > > -- pccdf2:/private> tail -2 station__pccdf2__prd__samtrieste-1/trace > > 08/13/02 12:16:59 samtrieste-1.SM.CacheMan test 31153: Needed: > > 1050225KB. can have: 28020985KB. Cache manager uses 2024095KB out of > > 2GB. Delivering of 1050225KB would exceed the quota of 2GB > > 08/13/02 12:16:59 samtrieste-1.SM.Repler 31153: No more deliveries > > possible > > > > Why does one place say 20GB and the other 2 ? > > How can I fix this ? > > > > Thanks > > > > Stefano > >