abhishekagarwal87 commented on issue #12701: URL: https://github.com/apache/druid/issues/12701#issuecomment-1187136044
Thank you @pchang388 for the detailed analysis. The best way to troubleshoot these degradations is to look at the flame graphs of overlord and peon. Looking forward to that information. I skimmed through the information that is already available. There is no major degradation in the system and JVM metrics. It could be that the overlord just became slower due to an application issue. I have seen that sometimes overlord HTTP threads slow down because of metadata operations. Some of these threads busy in metadata operations hold a lock that locks other threads. This brings the whole service to a crawl. We can confirm this by looking at flame graphs. If the metadata store is indeed the cause of slowness, then scaling the metadata store would certainly help. -- 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]
