On 10/30/2015 3:23 PM, Farzad Panahi wrote:
Mark - Could you elaborate a bit more on what you mean by "resolution
of the timestamps for the log messages"? If you are referring to the
log4j latency (https://logging.apache.org/log4j/2.x/manual/async.html#Latency),
it is in microseconds.
If it has anything to do with the logger why is it happening on thread
switch only and is so consistent? In your opinion what is a good way
to monitor or measure the performance of tomcat websocket message
processing?
It has to do with the resolution of the timer function being called by
the logger to generate the time stamp, which in turn depends on the OS.
I see conflicting values when I search around, but 10ms and 50ms are
common values I've seen, and also 1/16 of a second, which is ~66ms.
Different gettime methods have different resolutions too.
Cheers
Farzad
On Fri, Oct 30, 2015 at 3:16 AM, Mark Thomas <ma...@apache.org> wrote:
On 30/10/2015 00:03, Farzad Panahi wrote:
Hi,
I am using tomcat 8.0.23 to terminate my websocket connections. I was
looking at my trace logs and noticed that when tomcat worker thread
responsible for processing websocket messages switches to a different
thread, there is about 80ms delay.
No there isn't. There is about an 80ms resolution of the timestamps for
the log messages.
Mark
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org
