zjf2012 commented on issue #825: URL: https://github.com/apache/incubator-uniffle/issues/825#issuecomment-1519444411
> > Maybe we can increase the granularity of partitions->shuffle servers mappings to reduce overhead. Instead of mapping per partition, we can map by partition range or partition hash/modulo. > > Just my piece of thought. > > could you elaborate a bit more? We were considering partition range, but it's not that effective. We may reconsider in the near feature. > > Even the partitions are grouped in ranges, the partition to server mapping still would be a complete view of all partitions/ranges. Have you tried the Java TreeMap? It's a good candidate for range search instead of maintaining partitions/ranges map. Please check https://docs.oracle.com/javase/8/docs/api/java/util/TreeMap.html#higherEntry-K-. Keys are range values. Values are servers. -- 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]
