Skip to content

Nextflow-SLURM job status synchronization #82

Description

@mckeowr1

Nextflow is showing jobs as long-running/submitted when they have actually failed in SLURM, leading to workflow hanging.

4 jobs appeared to be running longer than their allocated CPU time according to nextflow.log, but were actually failed jobs in SLURM that Nextflow was not detecting.

Nextflow Log Output

Jun-18 11:19:11.660 [Task monitor] DEBUG n.processor.TaskPollingMonitor - !! executor slurm > tasks to be completed: 4 -- submitted tasks are shown below
~> TaskHandler[jobId: 5484012_72; id: 774; name: PLINK_RECODE_VCF (cb.strain.set.242_foliated.maggot_0.05); status: SUBMITTED; exit: -; error: -; workDir: /scratch4/eande106/Ryan/2025_briggsae_gwas_strainset_sims_batch_2/95/a850a2bcc4a8c0579a379a71562007 started: -; exited: -; ]
~> TaskHandler[jobId: 5484012_73; id: 775; name: PLINK_RECODE_VCF (cb.strain.set.242_crazy.feline_0.05); status: SUBMITTED; exit: -; error: -; workDir: /scratch4/eande106/Ryan/2025_briggsae_gwas_strainset_sims_batch_2/b5/327cbf412a340d1cd5d8ecc0fdf9fd started: -; exited: -; ]
~> TaskHandler[jobId: 5484012_74; id: 776; name: PLINK_RECODE_VCF (cb.strain.set.242_dissuasive.steer_0.05); status: SUBMITTED; exit: -; error: -; workDir: /scratch4/eande106/Ryan/2025_briggsae_gwas_strainset_sims_batch_2/11/0d2fed45f5f3fcd8056af3972c0da1 started: -; exited: -; ]
~> TaskHandler[jobId: 5484012_75; id: 777; name: PLINK_RECODE_VCF (cb.strain.set.242_conceivable.centipede_0.05); status: SUBMITTED; exit: -; error: -; workDir: /scratch4/eande106/Ryan/2025_briggsae_gwas_strainset_sims_batch_2/de/1cbadecae51eb2395990b4d3f8d7e7 started: -; exited: -; ]

Status of the job according to slurm commands
Check for an active job

$ squeue -j 5484012_72 
slurm_load_jobs error: Invalid job id specified

Check for a completed job

$ seff 5484012_72
Job ID: 5489867
Array Job ID: 5484012_72
Cluster: slurm
User/Group: rmckeow1/eande106
State: FAILED (exit code 15)
Nodes: 1
Cores per node: 8
CPU Utilized: 00:00:00
CPU Efficiency: 0.00% of 00:00:00 core-walltime
Job Wall-clock time: 00:00:00
Memory Utilized: 0.00 MB (estimated maximum)
Memory Efficiency: 0.00% of 40.00 GB (40.00 GB/node)

Looks like the job failed right away with an exit code of 15. Even when it failed, Nextflow task monitor never detected it as finished

I resolved the issue by terminating the nextflow run and rerunning the workflow with the -resume flag, but it was challenging to identify the problem. The job hung and repeatedly pinged the scheduler for several hours (possibly indefinitely).

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Type

    Fields

    No fields configured for Bug.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions