[galaxy-dev] jobs don't get killed when histories deleted
Hi all, It seems that this problem has been reported before http://dev.list.galaxyproject.org/jobs-don-t-get-killed-when-histories-deleted-tp4658452.html and should be solved https://trello.com/card/jobs-stop-job-function-in-jobs-admin-interface-is-broken/506338ce32ae458f6d15e4b3/627. However I am using the Apr-2014 version (at least the version with the fancy Load Data button at top of tools panel which is a quite new feature). The jobs from deleted histories still hang around in the manage jobs admin page with some cannot-finish setting_metadata jobs. The interesting thing is that the last update time is always less than 24 hours. How to delete these unwanted jobs? Best regards! Jun ___ Please keep all replies on the list by using reply all in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/ To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
Re: [galaxy-dev] jobs don't get killed when histories deleted?
Hi all, I've created Trello cards for both of these issues: https://trello.com/card/jobs-stop-jobs-on-history-deletion/506338ce32ae458f6d15e4b3/626 https://trello.com/card/jobs-stop-job-function-in-jobs-admin-interface-is-broken/506338ce32ae458f6d15e4b3/627 --nate On Feb 14, 2013, at 2:58 AM, Hans-Rudolf Hotz wrote: Hi Thon I can't help you killing the job on the clusterbut regarding the admin section (and the database): On very rare occasions, we do have jobs (which are not running anymore) but are labelled as running on the Manage jobs page in the admin section. Using the Stop Jobs option does not work to remove them. So I manually change the state to error in the 'job' table, eg: update job set state='error' where id=12345; As always, be very careful when you directly access the MySQL or PostgreSQL database! Regards, Hans-Rudolf On 02/13/2013 09:14 PM, Anthonius deBoer wrote: Hi, I am continue to struggle with the proccess of killing jobs that run on the SGE cluster when deleting histories. I would assume that jobs get removed from the queue when a history they are part of is deleted but in most cases, the jobs stay on the cluster and are not killed I also never have been able to get the kill jobs in the admin section to work (using the 8-Feb version)...It just sites there, returns and the jobs are still there on the cluster and still there in the database Is there anything in the database I can do to really kick out those jobs? Thanks Thon ___ Please keep all replies on the list by using reply all in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/ ___ Please keep all replies on the list by using reply all in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/ ___ Please keep all replies on the list by using reply all in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/
[galaxy-dev] jobs don't get killed when histories deleted?
Hi,I am continue to struggle with the proccess of killing jobs that run on the SGE cluster when deleting histories.I would assume that jobs get removed from the queue when a history they are part of is deleted but in most cases, the jobs stay on the cluster and are not killedI also never have been able to get the kill jobs in the admin section to work (using the 8-Feb version)...It just sites there, returns and the jobs are still there on the cluster and still there in the databaseIs there anything in the database I can do to really kick out those jobs?ThanksThon ___ Please keep all replies on the list by using reply all in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/
Re: [galaxy-dev] jobs don't get killed when histories deleted?
Hi Thon I can't help you killing the job on the clusterbut regarding the admin section (and the database): On very rare occasions, we do have jobs (which are not running anymore) but are labelled as running on the Manage jobs page in the admin section. Using the Stop Jobs option does not work to remove them. So I manually change the state to error in the 'job' table, eg: update job set state='error' where id=12345; As always, be very careful when you directly access the MySQL or PostgreSQL database! Regards, Hans-Rudolf On 02/13/2013 09:14 PM, Anthonius deBoer wrote: Hi, I am continue to struggle with the proccess of killing jobs that run on the SGE cluster when deleting histories. I would assume that jobs get removed from the queue when a history they are part of is deleted but in most cases, the jobs stay on the cluster and are not killed I also never have been able to get the kill jobs in the admin section to work (using the 8-Feb version)...It just sites there, returns and the jobs are still there on the cluster and still there in the database Is there anything in the database I can do to really kick out those jobs? Thanks Thon ___ Please keep all replies on the list by using reply all in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/ ___ Please keep all replies on the list by using reply all in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/