Description
Some of the timeouts in Spark are quite low (e.g. < 5 seconds) and this doesn't work well once larger datasets are involved and GC pauses happen that take a lot of time. We should bump all timeouts to be in the ~30s range.
Some of the timeouts in Spark are quite low (e.g. < 5 seconds) and this doesn't work well once larger datasets are involved and GC pauses happen that take a lot of time. We should bump all timeouts to be in the ~30s range.