Thanks Yoni,
that makes sense, lemme try that.(I'd say that busy waiting is a "bad" default.)
-- Lars
On Wednesday, December 26, 2018, 2:58:16 AM PST, Yoni Gottesman
<yon...@apache.org> wrote:
Hi, you are right. The default setting for omid is to run in HIGH_THROUGHPUT
mode. This means that the threads in the disruptor are busy-waiting.If you want
omid to run in a blocking mode, add to
bin/omid-server-configuration.yml:waitStrategy: LOW_CPUDoes this make sense?
On Wed, Dec 26, 2018 at 11:54 AM la...@apache.org <la...@apache.org> wrote:
I'm running Omid via its Phoenix integration.
After building 1.1.0-SNAPSHOT from source and starting Omid I find that it
consumes 400% CPU (i.e. 4 cores) just being idle.
Indeed with jstack I find four thread like this (which may or may not be
related):
"persist-0" #64 prio=5 os_prio=0 tid=0x00007f045c3a5800 nid=0x3745 runnable
[0x00007f043dedf000]
java.lang.Thread.State: RUNNABLE
at
com.lmax.disruptor.BusySpinWaitStrategy.waitFor(BusySpinWaitStrategy.java:36)
at
com.lmax.disruptor.ProcessingSequenceBarrier.waitFor(ProcessingSequenceBarrier.java:56)
at com.lmax.disruptor.WorkProcessor.run(WorkProcessor.java:148)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
Am I the only one seeing this?
Thanks.
-- Lars
