Thread for reporting suspected call cache misses

Post author
Steve Huang

This is an (experimental?) effort to consolidate and gather data about suspected misses regarding call-caching. 

Sometimes, we submit workflows with call-caching on, expecting that some of the tasks will be cache-hit. But for some reason, it turns out to be an unexpected miss. If we can accumulate the metadata here, it'll provide more data for the engineering team to debug the issue.

And to do this, I have this very simple template in mind to help provide the information:

 

```

rerun information:
this is where a task is expected to be a cache-hit, but turns out to be a cache-miss
  - submission ID:
  - workflow ID:
  - subworkflow ID:
  - sub-subworkflow ID:
  - task-shard name/index:
 
previous run information:
this is where the workflow, or more precisely the relevant task, is run the first time succesfully
  - submission ID:
  - workflow ID:
  - subworkflow ID:
  - sub-subworkflow ID:
  - task-shard name/index:

```

 

Comments

0 comments

Please sign in to leave a comment.