bharathv edited a comment on pull request #3371: URL: https://github.com/apache/hbase/pull/3371#issuecomment-859183336
> Thread local was used to avoid threads having to coordinate around a resource; this plus ringbuffer was meant to get us a write that was w/o locking (till we hit dfsclient at least). Ya that makes sense. I also have a version that fixes that problem by retaining Thread locals (see [this](https://github.com/bharathv/hbase/commit/cbcbf7e9d74d93515eb820163e2d6cd7b225787a)). It works by clarifying two states DONE and RELEASED, if this contention is a concern for any reason, just FYI. > Interesting on caffeine vs guava; probably this workload I don't know what % of it is noise in the environment, it is on my local machine may have some interference (tried to minimize as much as I can). > Have you looked at flamegraphs for asyncwal? Perhaps you'll see where the 1% is going? It is going into putIfAbsent(), see below.  -- 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
