Results 407 comments of roryqi

We should modify the method from ``` public StatusCode registerShuffle( String appId, int shuffleId, List partitionRanges, RemoteStorageInfo remoteStorageInfo, String user, ShuffleDataDistributionType dataDistType, int maxConcurrencyPerPartitionToWrite) { refreshAppId(appId); ShuffleTaskInfo taskInfo = shuffleTaskInfos.get(appId);...

We should catch ` BufferLimitOfHugePartitionException` in the `ShuffleServerGrpcService` and log the information of the exception.

> > > > If it's ok, I will go ahead. Gently ping. @jerqi @zuston @kaijchen @advancedxy > > > > > > > > > I think cli is...

@s0nskar The data distribution of reduce partition is important. Otherwise the reader will suffer from random IO according to our experience in the Uniffle (another remote shuffle service implement similar...

We need to guarantee the order of data, otherwise we will lose the data.

One concern is that it may occupy too much memory if we use Map to store blockIds instead of bitmap.

@EnricoMi Maybe we could to start to release 0.9.0-rc1.

https://docs.starburst.io/latest/security/biac-privileges.html