Juts to answer my own question,I found the source problem. It was because the brokers couldn't communicate with each other. By opening TCP ports (in my case EC2 security policy) the problem is solved.
On Wed, Aug 12, 2015 at 4:32 PM, shahab <shahab.mok...@gmail.com> wrote: > Sorry for posting the email again, because I still dealing with same > issue. > > > I have a kafka cluster consisting of two servers. I created a topic > "ReplicatedTopic" with 3 partitions and replication factor of 2. > > Next, I shutdown one the servers for 20 sec and return it back again. > Then, I looked at status of topic "ReplicatedTopic" and I got the > followings: > > Topic:ReplicatedTopic PartitionCount:3 ReplicationFactor:2 Configs: > Topic: ReplicatedTopic Partition: 0 Leader: 1 Replicas: 0,1 Isr: 1 > Topic: ReplicatedTopic Partition: 1 Leader: 1 Replicas: 1,0 Isr: 1 > Topic: ReplicatedTopic Partition: 2 Leader: 1 Replicas: 0,1 Isr: 1 > > Although broker zero is back. it seems it is not in sync with leader and > in fact it never became in sync again. > > Now question is how to make first broker in sync again so it appears both > in "isr" list and also it becomes leader for one of the partitions? > > best, > /Shahab > >
