Github user raajay commented on the issue: https://github.com/apache/spark/pull/18683 maxSizeInFlight can be large (~100-200 MB) when (a) available memory at reducer is high, or (b) when reducer spends most of its time waiting for fetchRequests. In such cases, using a hard coded value of '5' parallel fetches will result in individual fetchRequests to be bursts of size 20-40 MB. The configuration parameter, allows one have smaller sized fetchRequests while keeping the total maxSizeInFlight a constant. This configuration is helpful when the reducer spends most of its time waiting for fetchRequests to return. In such cases, we would like to increase the maxBytesInFlight
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org