franz1981 commented on pull request #3572:
URL: https://github.com/apache/activemq-artemis/pull/3572#issuecomment-846790513


   This is the last comparison of `main` vs `this PR` with 32 producers.
   `main`:
   ```
   **************
   RUN 1        EndToEnd Throughput: 52667 ops/sec
   **************
   EndToEnd SERVICE-TIME Latencies distribution in MICROSECONDS
   mean                638.65
   min                 104.45
   50.00%              573.44
   90.00%              884.74
   99.00%             2015.23
   99.90%             7045.12
   99.99%            11141.12
   max               21233.66
   count              3200000
   ```
   `this PR`
   ```
   **************
   RUN 1        EndToEnd Throughput: 57492 ops/sec
   **************
   EndToEnd SERVICE-TIME Latencies distribution in MICROSECONDS
   mean                725.65
   min                  94.72
   50.00%              501.76
   90.00%             1286.14
   99.00%             4685.82
   99.90%            10616.83
   99.99%            17563.65
   max               25690.11
   count              3200000
   ```
   Despite what looks to be a better latency profile for `main` the reality is 
that the new version is capable of pushing a bit more throughput with a better 
CPU usage overall, that could be used elsewhere (more load or GC).


-- 
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:
[email protected]


Reply via email to