dyf6372 commented on issue #6719: use SnapTreeMap instead of ConcurrentSkipListMap URL: https://github.com/apache/incubator-druid/pull/6719#issuecomment-446072294 > While 35% is spent in the .get call is true but most of that time 34% is consumed by the Druid comparator code, so not sure how the snapTreeMap is going to solve this? can you present some results or benchmarks ? @b-slim I'm not sure, I think it may decrease the number of compare or optimize the concurrency. Since the project seems to be not alive, I agree that it's may not a good idea to use this code now. However, it's really necessary for me to improve the performance in index since we have more than 1200 peon task for one cluster and data growth is fast. I'll try it in my cluster and if it's really work without any mistake I'll give the detail result or benchmark in the future.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
