Adapting code & tests to use NuQCJob #76
Conversation
|
Still a work in progress. |
|
Note this PR has been tested against biocore/mg-scripts#110 and that PR should be merged before reviewing and merging this PR. |
Co-authored-by: Daniel McDonald <d3mcdonald@eng.ucsd.edu>
charles-cowart
changed the title
|
biocore/mg-scripts#111 should be reviewed and merged first. GitHub reference should then be reverted back to normal. Then PR should be ready for review. The new test to validate NuQCJob script is an analog to the same test for legacy QCJob and should be sufficient for now. |
Updates based on testing on Qiita-RC. Tests updated to match changes in qp-klp and mg-scripts modules.
|
Dependency biocore/mg-scripts#112 should be reviewed and merged first before reviewing and merging this PR. |
qp_klp/tests/test_step.py
Outdated
| @@ -271,6 +272,24 @@ class BaseStepTests(TestCase): | |||
| "job_pool_size": 30, | |||
| "job_max_array_length": 1000 | |||
| }, | |||
| "nu-qc": { | |||
| "nodes": 1, | |||
| "nprocs": 16, | |||
There was a problem hiding this comment.
Based on the other open PR, should this be connected/the-same as {{cores_per_task}}? If yes, could you make sure to use the same name, just for clarity?
There was a problem hiding this comment.
Good catch! I removed nprocs for NuQC because it's not being used and added cpus_per_task since it is being used. I added a comment briefly covering the difference between the two parameters after reading some literature.
No description provided.