The following is the comment in the code. Do you see any ZK session
expiration in the broker log around that time?
// NOTE: the above write can fail only if the current controller
lost its zk session and the new controller
// took over and initialized this partition. This can happen if
the current controller went into a long
// GC pause
Thanks,
Jun
On Sun, Feb 16, 2014 at 3:10 AM, Bae, Jae Hyeon <metac...@gmail.com> wrote:
> Hi
>
> I am getting the following errors
>
> state-change.log:[2014-02-16 10:23:20,708] ERROR Controller 0 epoch 2
> encountered error while changing partition [request_trace,18]'s state from
> New to Online since LeaderAndIsr path already exists with value
> {"leader":0,"leader_epoch":0,"isr":[0,1]} and controller epoch 2
> (state.change.logger)
> state-change.log:[2014-02-16 10:23:20,708] ERROR Controller 0 epoch 2
> initiated state change for partition [request_trace,18] from NewPartition
> to OnlinePartition failed (state.change.logger)
>
> In the first try, I pre-created the topic, started sending the traffic but
> I didn't see any metric report information. So, I started to doubt
> something went wrong in the broker.
>
> I tried with automatic topic creation but the same error happened. I guess
> this might be a timing problem because another topic with small traffic
> didn't have any problem.
>
> If this is already know issue, I will fetch from trunk and deploy and test
> again. Otherwise, could you let me know how to avoid this error?
>
> Thank you
> Best, Jae
>
