Junfan Zhang
Junfan Zhang
If you have time, could u help review this proposal? @jerqi
> @zuston Do you commit the latest source? Some previous comments seems not be fixed, eg, user shouldn't be a member in RemoteStorage > > ``` > message RemoteStorage {...
I test this PR in our internal env. - [x] coordinator read the exclude-node-file from kerberlized HDFS - [x] shuffle server write shuffle data to kerbeilized HDFS and client read...
Changelog 1. Add the check of rss.security.hadoop.kerberos.relogin.interval.sec 2. Add more test cases about kerberos HDFS cluster
@jerqi . Updated, and all CI test passed.
Gentle ping @jerqi
@colinmjj @duanmeng If u have time, could you help review this? Looking forward to your reply. Thanks.
I'm confused why not reuse the exclude-node-file? And the decommission operation maybe controlled by coordinator will be better? Do u have any ideas on it? @jerqi
I can’t got your thought fully. As i know, when need to make one shuffle server decommission, just add it to the exclude file in coordinator side. There is no...
> But if we have multi coordinator, we will change exclude-node-files multi times, and if we need to start shuffle server again, we need change them again. Exclude-nodes-file could be...