/cromwell-gwf-log

Primary LanguagePythonMIT LicenseMIT

cromwell-gwf-log

Prerequisites

  • AWS CLI
  • Python requests library

How to Use

For a given task of your WDL/Cromwell workflow, this gets the job ID from AWS Batch, then prints out the log via AWS CloudWatch.

python get_log.py \
    -k ~/keys/cromwell-secrets.json \
    -w 66431c43-af28-4df3-a678-7990db3c8a73 \
    -t CellRangerGex.Count \
    -r us-east-1

In a normal case, a task name consists of three names:

Workflow Name + Subworkflow Name + Task Name

e.g. Sharp.Preprocess.CiteSeqCount

python get_log.py \
    -k ~/keys/cromwell-secrets.json \
    -w dc41ed29-0e92-415f-ba6e-d574c80d9960 \
    -t Sharp.Preprocess.CiteSeqCount \
    -r us-east-1

In the case where a job was restarted and already completed tasks were picked up from the cache, do not specify a workflow name in the task name. For example:

Subworkflow Name + Task Name

e.g. Preprocess.CiteSeqCount

python get_log.py \
    -k ~/keys/secrets-aws.json \
    -w dc41ed29-0e92-415f-ba6e-d574c80d9960 \
    -t Preprocess.CiteSeqCount \
    -r us-east-1