MarkGaox commented on issue #2594: URL: https://github.com/apache/helix/issues/2594#issuecomment-1709081211
Suppose that the delay window configuration is set to be 16 hours, the min active replica is 2, and the replication factor is 3. When a node is brought down for maintenance, it's still considered as holding replicas that are initially assigned to it. Consequently, the actual active replica remains 2 for subsequent 16 hours. And Espresso, one of Helix's clients, is asking for a feature that can quickly bootstrap the third replica even during the delay window. This functionality is crucial to ensure that they have sufficient cluster availability guarantee to start their cluster-wise maintenance task. For your 5min delay problem, [here](https://github.com/apache/helix/blob/fbefa64909781d479ec486a93d22bba416322fb8/helix-core/src/main/java/org/apache/helix/model/ClusterConfig.java#L62) is how Helix configure the delay time. Based on my understanding, Pinot should have a similar configuration. And you can probably tune it to reduce the rebalance lag. @jadami10 Hope this answers your question. -- 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]
