Thanks Brock for that tip!
I ran the ps -L -p command and this is trimmed version of it:
PID LWP TTY TIME CMD
19524 19550 ?01:11:38 java
19524 19551 ?01:33:55 java
19524 19552 ?01:38:32 java
19524 19557 ?01:30:12 java
19524 is the process id and these
You can enable profiling in tools like jvisualvm and see which methods
other select() show up high.
Thanks,
Jun
On Thu, Sep 17, 2015 at 5:58 PM, Jaikiran Pai
wrote:
> That's a good point, I missed that. Is there any other way to track down
> why it would consume this high CPU when it's idle? I
If you are on Linux, you can use the following approach.
1) Get the pid
2) Find the largest consumer of CPU via:
ps -L -p
3) Convert the LWP column to hex and then look for that value in the
thread dump on the "nid=0x.." field of each thread.
Cheers!
Brock
On Thu, Sep 17, 2015 at 3:58 PM, Jai
That's a good point, I missed that. Is there any other way to track down
why it would consume this high CPU when it's idle? I can send the thread
dumps (taken at intervals of a few seconds of each other), but each of
them has not much interesting (IMO) than these specific repeated
stacktraces.
The selector is probably not the issue. If there is no incoming traffic,
selector.select(300) won't return until after 300ms.
Thanks,
Jun
On Thu, Sep 17, 2015 at 1:13 PM, Jaikiran Pai
wrote:
> Sending this to the dev list since the Kafka dev team might have more
> inputs on this one. Can someo
Sending this to the dev list since the Kafka dev team might have more
inputs on this one. Can someone please take a look at the issue noted
below and whether the suggested change makes sense?
-Jaikiran
On Tuesday 15 September 2015 12:03 AM, Jaikiran Pai wrote:
We have been using Kafka for a whi
