Re: Custom partitioning of keys with keyBy

2021-11-04 Thread naitong Xiao
uld be interesting for high-throughput > situation where you actually want to run in parallelism higher than the > number of different ‘small’ keys > > On Thu, Nov 4, 2021 at 5:48 AM naitong Xiao wrote: > > I think I had a similar scenario several months ago, he

Re: Custom partitioning of keys with keyBy

2021-11-03 Thread naitong Xiao
I think I had a similar scenario several months ago, here is my related code: val MAX_PARALLELISM = 16 val KEY_RAND_SALT = “73b46” logSource.keyBy{ value =>  val keyGroup = KeyGroupRangeAssignment.assignToKeyGroup(value.deviceUdid, MAX_PARALLELISM)  s"$KEY_RAND_SALT$keyGroup" } The keyGroup is