If you don't mind, YES. I'm running debian squeeze (latest patch level).
Thanks! ________________________________________ From: Loris Bennett [[email protected]] Sent: Thursday, January 31, 2013 8:45 AM To: slurm-dev Subject: [slurm-dev] RE: Ghost JobID in Database? Guglielmi Matteo <[email protected]> writes: > I'm definitely not an expert in databases... how do > I change the state of these jobs after this command: > > mysql -u root -p mypass Unless you aren't worried about loosing any data, I don't think you really want to be modifying the database on the fly. I had to correct a similar problem, where we had jobs which were in fact no longer running but showed up with sacct because they had no end time. I wrote a Perl program to generate an SQL file to make the necessary modifications, made a dump of the production database, tested the program/SQL with the dump. I then used the program on the production database. As well as generating the SQL to make the changes, I also generated the SQL to undo those changes in case anything went wrong. If you are interested, I can let you have the Perl code and give you an idea of how to modify it for your case. Cheers Loris > ________________________________________ > From: Alejandro Lucero Palau [[email protected]] > Sent: Wednesday, January 30, 2013 6:23 PM > To: slurm-dev > Subject: [slurm-dev] RE: Ghost JobID in Database? > > There have been some race conditions in the past that led to accounting > errors like this. I do not know if your case come from that time or it > is something new. > > I wonder if there are race conditions hidden and hard to detect, but > could we have some automatic action for detecting such a problem? > > The quick solution would be to have that check with some local scripts. > The good one would be slurmdbd and slurmctld taking the effort. > > What do you think Danny? > > On 01/30/2013 05:30 PM, Lipari, Don wrote: >> Matteo, >> >> I suspect something happened that prevented the state change for your jobs >> below from being propagated to the database. You are going to have to >> modify them manually in mysql. Change the state for these jobs to 3 >> (JOB_COMPLETE). I would also update the time_end field to something greater >> than time_start and less than now, though this second step may not be >> necessary for sacctmgr to remove the account. >> >> Don >> >> >>> From: Guglielmi Matteo [mailto:[email protected]] >>> Sent: Tuesday, January 29, 2013 2:48 PM >>> To: slurm-dev >>> Subject: [slurm-dev] Ghost JobID in Database? >>> >> >>> I want to remove an unused account but the sacctmgr command returns 6 ghost >>> JobID which are not running anywhere on the system (note the 6 lower JobID >>> values). >>> >>> How can I purge these ghost jobs? >>> >>> ### >>> >>> sacctmgr remove account guest >>> >>> Error with request: Job(s) active, cancel job(s) before remove >>> JobID = 594676 C = superb A = guestlp U = bruijn P = batch >>> JobID = 594680 C = superb A = guestlp U = bruijn P = batch >>> JobID = 594686 C = superb A = guestlp U = bruijn P = batch >>> JobID = 594699 C = superb A = guestlp U = bruijn P = batch >>> JobID = 594703 C = superb A = guestlp U = bruijn P = batch >>> JobID = 594707 C = superb A = guestlp U = bruijn P = batch >>> JobID = 3563965 C = superb A = guestlp U = bruijn P = batch >>> >> >> > > > WARNING / LEGAL TEXT: This message is intended only for the use of the > individual or entity to which it is addressed and may contain > information which is privileged, confidential, proprietary, or exempt > from disclosure under applicable law. If you are not the intended > recipient or the person responsible for delivering the message to the > intended recipient, you are strictly prohibited from disclosing, > distributing, copying, or in any way using this message. If you have > received this communication in error, please notify the sender and > destroy and delete any copies you may have received. > > http://www.bsc.es/disclaimer -- Dr. Loris Bennett (Mr.) ZEDAT, Freie Universität Berlin Email [email protected]
