nextflow-io/nextflow

Ability to use a time delay in 'errorStrategy' to re-submit scheduler jobs

Open

#721 opened on May 30, 2018

View on GitHub
 (17 comments) (0 reactions) (0 assignees)Groovy (3,382 stars) (784 forks)batch import
help wanted

Description

Our HPC system is giving large amounts of errors like this:

Error executing process > 'sambamba_dedup_flagstat (1)'

Caused by:
  Failed to submit process to grid scheduler for execution

Command executed:

  qsub -terse .command.run

Command exit status:
  1

Command output:
  Unable to run job: failed receiving gdi request response for mid=1 (got syncron message receive timeout error)..
  Exiting.

This is the result of the scheduler (SGE) being unable to submit jobs or respond to queries. If I try again later, for example after 5 minutes, the job may be successfully scheduled.

The errorStrategy described here shows how to increase resource allotment on error.

However, in this case I need to wait a minute or two before attempt to submit the job again.

Is it possible to accomplish this or get this feature in Nextflow? An 'errorStrategy' of 'retry in 5 minutes'.

Contributor guide

Tech stack
groovy
Domain
devops
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-2 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
clear
Prerequisites
Familiarity with NextflowKnowledge of SGE grid scheduler
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
60
Research direction
Examine the Nextflow source code for the errorStrategy implementation, particularly in the process execution and scheduler submission files. Look for how retries are currently handled and where a delay could be introduced. Check existing comments in issue #721 for further discussion.