rickyma commented on code in PR #1693: URL: https://github.com/apache/incubator-uniffle/pull/1693#discussion_r1599915407
########## docs/client_guide/spark_client_guide.md: ########## @@ -149,4 +149,20 @@ Other configuration: |---|---|---| |spark.rss.access.timeout.ms|10000|The timeout to access Uniffle coordinator| |spark.rss.client.access.retry.interval.ms|20000|The interval between retries fallback to SortShuffleManager| -|spark.rss.client.access.retry.times|0|The number of retries fallback to SortShuffleManager| \ No newline at end of file +|spark.rss.client.access.retry.times|0|The number of retries fallback to SortShuffleManager| + +### Partition reassign in one shuffle attempt + +To achieve better task stability, partition's reassign mechanism that requests the new replacement shuffleServers is introduced to overcome server instability Review Comment: I think this will be better: To achieve better task stability, a partition reassignment mechanism has been introduced, which requests new replacement shuffle servers to overcome server instability caused by unhealthy conditions or high memory pressure in a single shuffle attempt. Currently, this feature is not compatible with stage retry and multiple replica mechanisms (additional testing is required). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
