Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Helena Rasche
    @erasche
    galaxy=> select max(job.update_time) - min(job.create_time) from job where job.history_id = 141383;
       ?column?
----------------------
 1 day 13:50:07.27901
(1 row)
    that's probably approximately close for "how long from when you started history until it finished"
    Milad Miladi
    @mmiladi
    Not well defined :D let's say the big time killing monster steps. (sorted by runtime). A
    Thanks much. What is the difference between the two times
    Helena Rasche
    @erasche
    I guess for a workflow, bottlenecks would also be more complex, and any places where jobs were 'stuck' waiting. But that's a larger project
    difference between? Just that we run lots of jobs in parallel so the first number is much higher.
    Total amount of time in each tool:
    galaxy=> select job.tool_id, (sum(job_metric_numeric.metric_value)::integer || 'seconds')::interval from job left join job_metric_numeric on job.id = job_metric_numeric.job_id where job.history_id = 141383 and metric_name = 'runtime_seconds' group by job.tool_id order by interval desc;
                                                   tool_id                                                   | interval
-------------------------------------------------------------------------------------------------------------+-----------
 toolshed.g2.bx.psu.edu/repos/bgruening/infernal/infernal_cmsearch/1.1.2.0                                   | 532:23:33
 toolshed.g2.bx.psu.edu/repos/rnateam/graphclust_nspdk/nspdk_sparse/9.2.3                                    | 85:48:50
 toolshed.g2.bx.psu.edu/repos/rnateam/structure_to_gspan/structure_to_gspan/0.4                              | 08:25:12
 toolshed.g2.bx.psu.edu/repos/bgruening/infernal/infernal_cmbuild/1.1.2.0                                    | 07:17:00
 toolshed.g2.bx.psu.edu/repos/rnateam/graphclust_nspdk/NSPDK_candidateClust/9.2.3                            | 06:59:18
 toolshed.g2.bx.psu.edu/repos/rnateam/graphclust_align_cluster/graphclust_align_cluster/0.1                  | 06:25:01
 toolshed.g2.bx.psu.edu/repos/rnateam/viennarna_rnafold/viennarna_rnafold/2.2.10.4                           | 04:25:14
 toolshed.g2.bx.psu.edu/repos/rnateam/graphclust_cmfinder/cmFinder/0.4                                       | 02:36:47
 toolshed.g2.bx.psu.edu/repos/rnateam/graphclust_mlocarna/locarna_best_subtree/0.4                           | 02:13:22
 toolshed.g2.bx.psu.edu/repos/rnateam/graphclust_postprocessing_no_align/graphclust_glob_report_no_align/0.5 | 00:26:48
 toolshed.g2.bx.psu.edu/repos/rnateam/graphclust_prepocessing_for_mlocarna/preMloc/0.4                       | 00:19:02
 toolshed.g2.bx.psu.edu/repos/rnateam/graphclust_preprocessing/preproc/0.5                                   | 00:04:02
 toolshed.g2.bx.psu.edu/repos/nml/collapse_collections/collapse_dataset/4.0                                  | 00:00:12
 toolshed.g2.bx.psu.edu/repos/iuc/ucsc_fasplit/fasplit/357                                                   | 00:00:10
(14 rows)
    Milad Miladi
    @mmiladi
    Aha I see, so the first one is the total of all "Job Runtime (Wall Clock)"
    Helena Rasche
    @erasche
    yes
    Milad Miladi
    @mmiladi
    You are a superhero. Thanks a lot. Now I know what to spend time on :)
    Helena Rasche
    @erasche
    cool, happy to help.
    I'll add these queries I've written just now to our gxadmin, so it'll be easier next time too.
    Milad Miladi
    @mmiladi
    :+1:
    Maybe the last wish :D : Is there a way to get the (max(job.update_time) - min(job.create_time)) but without the job creation, queuing times..?
    Helena Rasche
    @erasche
    I can't write it in one query but, your first job started around 2019-06-16 23:28:17.583856, and ended on 2019-06-18 13:06:09.239834
    I do not think those are more than ± 20 minutes, off of the total time.
    and a caveat, that's from job_state_history table which is apparently based on polling, and not event based, so those are "when the polling process checked"
    Milad Miladi
    @mmiladi

    I can't write it in one query but, your first job started around 2019-06-16 23:28:17.583856, and ended on 2019-06-18 13:06:09.239834

    you mean one single job?

    Helena Rasche
    @erasche
    no, start of first job in running state, to end of last job in 'ok' state
    Milad Miladi
    @mmiladi
    I see. It makes sense.
    Thanks much for the help.
    Helena Rasche
    @erasche
    sure, no problem
    faizansaleem1992
    @faizansaleem1992
    A job I submitted earlier stopped with an error "The job creating this dataset has been resubmitted" and I do not seem to grasp what has gone wrong. I am in need of this analysis and do not know what just happened to it. command was cluster.split. my id id faizansaleemgalaxy@gmail.com
    QZ1130
    @QZ1130
    This happened exactly to me! ' The Job creating this dataset has been resubmitted.' Could we get some help for this ?
    Helena Rasche
    @erasche
    This means the job has failed, and we have automatically retried to run it with more memory.
    If your jobs fail, please check the (i) information icon, and see the stderr or stdout, and if those do not tell you the issue, then please usethe bug report icon to submit errors
    Hans-Rudolf Hotz
    @hrhotz
    Just noticed: if I try to login on "rna.usegalaxy.eu" using the elixir login, I get redirected to "usegalaxy.eu" (and properly logged in)....switching to "rna.usegalaxy.eu", I am not logged in
    Helena Rasche
    @erasche
    ohhhhkay interesting
    I am actually not sure there is a solution to that.
    Bother.
    Ok, thanks for the report @hrhotz I'll talk with elixir guys
    Helena Rasche
    @erasche
    I would suggest you just use elixir to login to usegalaxy.eu, set password, and login with password on subdomain for now :(
    Alex Ostrovsky
    @astrovsky01
    On the stacked bar graph of the queue, should the hover-over list show so many jobs with a 0?
    Milad Miladi
    @mmiladi
    Q: Is there a way to tag multiple history entries at once?
    Helena Rasche
    @erasche
    If it's in a collection you can use the 'apply tag to collection' tool
    otherwise no
    Milad Miladi
    @mmiladi
    Thanks
    Dave Clements
    @tnabtaf
    Hi All, I just added a hub page about Galaxy @ JOBIM. It lists two talks, but no posters. If you are presenting a Galaxy-related poster @ JOBIM then please let me know and I will add it to the page.
    (And through sheer bad luck, JOBIM again collides with GCC, for probably the 4th time in 5 years)
    faizansaleem1992
    @faizansaleem1992
    Dear Ms. Rasche,
    You mentioned that you tried to run a failed job of mine with more memory (cluster.split). Hoe can I see the progress of newly submitted command because in my accound it still ahows process3s of cluster.split as failed?. faizansaleemgalaxy@gmail.com
    Helena Rasche
    @erasche
    Hi @faizansaleem1992, no I didn't try t run it, our Galaxy automatically re-runs it once. But if it shows up as failed, then it has failed twice. Please use the bug report (small bug) icon tosubmit a report.
    QZ1130
    @QZ1130
    Hello Helena, A simple question( maybe stupid),I have mapped with HISAT2 and then used FEATURE COUNT to check the assigned reads.... It showed 0.... Could you give me some ideas that which part I might do something wrong?
    Helena Rasche
    @erasche
    Hi @QZ1130 I'm afraid I don't have any ideas, I'm the sysadmin and do not do so much bioinformatics currently. I recommend posting those sort of questions to https://help.galaxyproject.org
    QZ1130
    @QZ1130
    :) Thanks
    Milad Miladi
    @mmiladi
    Hello, can I add a tool to multiple tool_panel_section_labels?
    https://github.com/usegalaxy-eu/usegalaxy-eu-tools/edit/master/bgruening.yaml
    Helena Rasche
    @erasche
    Nope
    sorry
    When tools appear in multiple places it's a bug on our end, and due to different versions being placed in different sections