Github user leventov commented on the issue:
https://github.com/apache/logging-log4j2/pull/87
<h3>Latency comparison</h3>
```
java -jar log4j-perf/target/benchmarks.jar
".*Log4j2AppenderLatencyBenchmark.end2end" -f 1 -wi 1 -i 30 -t 2 -p
loggerType=MMap
```
On my MBP with 4 cores and SSD.
https://cdn.rawgit.com/leventov/288f6ff4ea960998e6479001976b0755/raw/aeb2f65560241deaa9394ff1ef8044bb961ea37e/plotFiles.html
(red line is before this PR, blue line is this PR)
Tail latency is much better than it used to be, but worse than I expected.
Calling `MappedByteBuffer.load()` significantly improves latency around
99.99-99.999 percentiles, but dramatically drops the extreme tail latency, from
4.5 milliseconds to ~300 milliseconds. I don't know the reason for that.
Also quite "fun" than hand-rolled version of MappedByteBuffer.load():
https://github.com/leventov/logging-log4j2/tree/wait-free-MemoryMappedFileManager-crash
crashes the JVM when I try to run the same command
```
java -jar log4j-perf/target/benchmarks.jar
".*Log4j2AppenderLatencyBenchmark.end2end" -f 1 -wi 1 -i 30 -t 2 -p
loggerType=MMap
```
Could somebody else please try it on other machines?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
