This document in the handbook for the UK Biobank RAP states, "If you decide not to use a retry policy, occasionally, some jobs may fail due to sample-specific issues or due to external factors such as spot instance termination or other intermittent system errors. "
That implies that we can choose to use a retry policy. But there is no description of a "retry policy" elsewhere on that page nor in the `dx run` manual page.
When running jobs with `dx run`, how can we set a retry policy? As an example, it would be nice to allow spot jobs that are terminated by AWS to be retried several times.
To use it with dx run, I believe that syntax would be similar to what we showed during the Large scale data analysis webinar, e.g. for setting timeoutPolicy for Regenie
Comments
2 comments
GIven this doc page: https://documentation.dnanexus.com/developer/api/running-analyses/workflows-and-analyses#execution-policy, I would try to specify retry policy as JSON.
To use it with dx run, I believe that syntax would be similar to what we showed during the Large scale data analysis webinar, e.g. for setting timeoutPolicy for Regenie
dx run app-regenie --extra-args '{
"timeoutPolicyByExecutable": {
"app-MAIN_REGENIE_APP_ID": {
"*": { "hours": 12 }
},
"app-STEP1_APP_ID": {
"*": {"hours": 200}
},
"app-STEP2_APP_ID": {
"*": {"hours": 200 }}}}'
Please sign in to leave a comment.