On linux,  "echo $!  > pid.txt" will save the PID of the previous background 
job to a file.
If you run java to launch kafka directly rather than using the provided 
scripts, it 
can save process pid into the file. 

Regards,

Libo


-----Original Message-----
From: Jun Rao [mailto:[email protected]] 
Sent: Friday, July 05, 2013 12:00 AM
To: [email protected]
Subject: Re: PID files for monitoring two different instances of Kafka on same 
machine

We don't have PID files now. You can figure out the instance from the config 
dir in the command line.

Thanks,

Jun


On Thu, Jul 4, 2013 at 7:20 AM, Hanish Bansal < 
[email protected]> wrote:

> Hi all,
>
>
> I was just trying to create a monitoring service of my own for Kafka. 
> I wanted  to know if Kafka provides any sort of functionality for pid 
> files which I can see to check if Kafka is up or not.
>
> My scenario is:
>
> I will have 2 Kafka instances running on same machine with different 
> configurations i.e. their home directories will be different.
> I want to start a specific Kafka instance if any one of the two gets 
> killed. So I wanted to know which Kafka instance is down.
>
>
> So, how I can I get the PID of a particular Kafka instance ?
>
> Thanks in advance.
>
> --
> *Thanks & Regards*
> *Hanish Bansal*
>

Reply via email to