franz1981 edited a comment on issue #2633: ARTEMIS-2317 Avoid long TTSP caused by Page::read using mmap read URL: https://github.com/apache/activemq-artemis/pull/2633#issuecomment-487471785 @qihongxu that's super interesting :O TBH I was expecting the patched version to be on par with very small messages instead, because the most of the cost (just guessing eh) was on decoding side, while with big buffers I was expecting that the zero-copy approach of mmap was unbeatable: it would be interesting to check where is the cost of the non-patched version with such big messages too. I suppose that small messages are the common case so, do you have some test result with smaller messages (and including the GC optimizations on https://github.com/apache/activemq-artemis/pull/2646)? Re > We just let it remain default to 5. Why not making it totally unbounded and let the SSD (iotop -o) tell you if I/O is maximized? If GC will occur, I'm expecting that Artemis PageCache will be cleaned up of several cached pages and limiting it just to 5 concurrent requests means could be quite limiting with very fast ssd and many cores servers... https://github.com/apache/activemq-artemis/pull/2645 could be a game changes with many concurrent page read requests :)
---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
