Handle ConcurrentJobLimit errors more intelligently #838

geopyspark driver deployements typically have limit on max number of parallel jobs, and going over that limit might cause something like:

ERROR openeo.extra.job_management: 400 ConcurrentJobLimit: Job was not started because concurrent job limit (10) is reached.

the client could handle this more intelligently in some situations
e.g. back off a bit (instead of hard failing) when creating/starting new jobs in the job manager loop

Note however that ConcurrentJobLimit is not an official error code (yet) in openEO API, so pushing for that might be part of the work here.

related to

• Recommended limits for created/running jobs openeo-api#559