Re: Running disruptor async performance tests

2014-03-30 Thread Rebecca Ahlvarsson
Thanks again, Remko. After you detailed explanation everything makes sense now. I am using an i7 machine with Ubuntu. -Becky On Sat, Mar 29, 2014 at 8:49 PM, Remko Popma wrote: > When you run PerfTest with these options below you get the same as the > first test run by PerfTestDriver "Log4j2:

Re: Running disruptor async performance tests

2014-03-29 Thread Remko Popma
When you run PerfTest with these options below you get the same as the first test run by PerfTestDriver "Log4j2: Loggers all async (single thread) ": -Dlog4j.configurationFile=perf3PlainNoLoc.xml -DLog4jContextSelector=org.apache.logging.log4j.core.async. AsyncLoggerContextSelector PerfTestDriver

Re: Running disruptor async performance tests

2014-03-29 Thread Rebecca Ahlvarsson
Thank you very much, Remko. So I ran both tests on my machine and got the numbers below. I am interested in the latency numbers and throughput when we have one (i.e. threadCount 1 ?) asynchronous logger thread using disruptor. Can you point me out to these numbers on my results below? Not sure what

Re: Running disruptor async performance tests

2014-03-29 Thread Remko Popma
Hi Becky, The performance test code is a bit messy. Apologies for that. To answer your questions: I used the longer message for throughput tests, and the shorter message for latency tests. PerfTest parameters: [0] = runner class (like org.apache.logging.log4j.core. async.perftest.RunLog4j2) [1] =

Re: Running disruptor async performance tests

2014-03-28 Thread Matt Sicker
Going to guess Remko has something useful to say on this, but since he's in Japan, just wait a bit for the timezones to make sense. On 28 March 2014 22:03, Rebecca Ahlvarsson wrote: > I am trying to run the async performance tests described on the link below > on my machine. > > > http://loggin