BI-1864 - Experiment Export doesn't consistently work#276
BI-1864 - Experiment Export doesn't consistently work#276timparsons merged 1 commit intorelease/0.8from
Conversation
mlm483
left a comment
There was a problem hiding this comment.
This change makes sense to me. It looks OK running against my local bidb and BreedBase, but when I change my .env to point to the rel-test bidb (in order to test the program Alexandra reported), I'm seeing an unending loop of germplasm cache misses and population attemps for the same key when I try to download an experiment. I think rel-test-sugarcanebase is buckling under it's own weight.
I'm OK with merging this change, I think other performance improvements that may impact the speed and reliability (e.g. due to timeouts) of downloads should be separate stories.
I'll approve but am interested in the follow-up to this if it doesn't appear to fix Alex's issue. |
Description
Story: https://breedinginsight.atlassian.net/browse/BI-1864
Updated the experiment export logic to pull all of the germplasm for a program once at the beginning of the export process rather than hitting the cache for each observation unit being exported. This should improve the performance because the current cache implementation still fetches all germplasm records, then searches in the BrAPI object of interest to filter by the desired criteria. For experiment import, this means the full germplasm list for a program is pulled
ntimes, wherenis the number of observation units being exported.Also added more logging to the export process
Dependencies
bi-web: release/0.8
Testing
Checklist: