Improve Stop Job Logic
boyter opened this issue · comments
The current logic to pause/stop jobs is a little hairy and not brilliant. It has issues stopping jobs when they are running and leaves open file handles. Need to resolve this.
TODO - Store pending deletes in a persistant queue and apply it on startup.
Implemented in branch issue41 verified with 15000 queued deletions as a stress test. Seemed to work fine and is a bit of an edge case, I would not expect someone to add 15,000 repos and then delete them. Might be worth adding a total reset option into the UI somewhere that resets everything.
Going to merge in and close down.