Open ph opened 8 years ago
I wonder about a couple of things here, mostly related to web api.
I guess some of this penalty will come out of there, what do you think?
On the other side, can we also add benchmarks for the webapi, could be with another issue? . I see the json serialization/deserialization (aka decoration) adding some performance penalty.
In the tests from #4572 are without any calls made to the metrics api, even if the web api receive the snapshot in the tests. The performance impact should be negligible.
json serialization/deserialization (aka decoration
You are talking about the web api? This was not tested in the test.
Lets apply this motto here, make it work, make it pretty and make it fast, we are at step 2.
but the webserver is still up and running, so some impact I expect from it. what do you think?
On Wed, Feb 10, 2016 at 3:21 PM Pier-Hugues Pellerin < notifications@github.com> wrote:
In the tests from #4572 https://github.com/elastic/logstash/issues/4572 are without any calls made to the metrics api, even if the web api receive the snapshot in the tests. The performance impact should be negligible.
json serialization/deserialization (aka decoration
You are talking about the web api? This was not tested in the test.
Lets apply this motto here, make it work, make it pretty and make it fast, we are at step 2.
— Reply to this email directly or view it on GitHub https://github.com/elastic/logstash/issues/4664#issuecomment-182394026.
The current metrics system add a 15% slowdown in the events processing of the pipeline. We should improve that situation. I believe we have some room to grow to improve the situation. I think is reasonable to improve the performance by 5%.
Lets try the following:
LongAdder
to see if we can some performance improvements #4508Followup of #4572