wangbo edited a comment on pull request #8768: URL: https://github.com/apache/incubator-doris/pull/8768#issuecomment-1084697088
I think the key point is how much the Timer cost in Prod Env. In my experience,millions times call of Timer could see significant performance drop. Sql runs one by one, Timer may not so cost. In Prod Env, a Be may have many sqls runs at the same time, so the Timer cost may big. I would make a concurrent query test and then Perf for a period of time to see the cpu cost of timer. If there is no noticeable increase in cpu time, I think new timer is OK. And this may not be a trade-off, find online performance bottlenecks is an occasional operation. But Timer is called all the time. -- 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]
