lordcheng10 commented on PR #3062: URL: https://github.com/apache/bookkeeper/pull/3062#issuecomment-1378153641
> 1. read request doesn't wait for the completion on that threadpool. > > > When cold data is read, it will take a long time. At this time, the read request will be blocked, resulting in the write timeout. > > Can I see a test reproing this? i suspect the problem is somewhere else. > > 2. The ordering there is required for a reason. > > The test in this PR simply does not cover anything related to actual functionality. I suggest trying to enable the feature by default and running all tests to see what breaks, it's hard to come up with something more specific off the top of my head. > > 3. Do you have apples-to-apples comparison of performance, same load, same bookies etc, the feature off/on? For ledgers in the Writable state, reading and writing share the same thread pool, and there is no way to isolate reading and writing; IMO, it is useful to isolate reads and writes to different threads of execution -- 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]
