API changes to support the progress bar. - The eventual idea is for the save_stats route to be called by the workers during replication. Workers will save stats for a job id and attempt number. - Make modifications to the /jobs/list and the /jobs/get_debug_info routes to also return estimated bytes/records. We need both estimated metadata, as well as running states to calculate progress bar and throughput. - add the save_stats route. This is the route that will be called by workers. I've done my best to reuse existing openapi bodies to reduce duplication. - add the estimatedRecords and estimatedBytes fields to the AttemptStats body. This is part of the AttemptRead and the AttemptStreamStats objects. This eventually filters up to the jobs/list and jobs/get_debug_info objects. This also adds these to all the endpoints that were previously returning stats information. I think the duplicated data is a small issue and don't think it's worth splitting out a new api objects, though I will gladly do so if folks feel strongly. minor changes to the AttemptApiController to support the new route. - I've stubbed out the handlers for now since the backend is not yet implemented.
520 KiB
520 KiB