v1.2.0
Added caper cleanup to clean up outputs of a single workflow.
- Usage:
caper cleanup [METADATA_JSON_FILE]: Define ametadata.jsonfile for the first positional argument.caper cleanup [KEYWORD_TO_FIND_WORKFLOW]: Define a keyword to find a workflow from a runningcaper server.
- This DELETEs ALL FILES on a workflow root directory (under key
workflowRootin metadata JSON) written in metadata. Deletion is done in arm -rffashion. - This action is NOT REVERSIBLE. Use this at your own risk.
- Add
--deleteto actually delete all outputs. Otherwise,caper cleanupwill just dry-run.
Added caper gcp_monitor to monitor resources of workflows run on Google Clouds.
- Usage:
caper gcp_monitor [METADATA_JSON_FILE]: Define ametadata.jsonfile for the first positional argument.caper gcp_monitor [KEYWORD_TO_FIND_WORKFLOW]: Define a keyword to find a workflow from a runningcaper server.
- Output will be either in a TSV format or in a JSON format.
- TSV: Flatten version (with dot notation) of JSON one.
- JSON: More detailed.
- This feature is available for a workflow run with Caper>=1.2.0.
- Outputs:
- Task name / shard index / status / ...
- Instance resource info: cpu, memory, Cromwell-mounted disk.
- Statistics of resources:
mean,std,min,maxof each resource (memory, Cromwell-mounted disk's usage, cpu_pct).