Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
Kafka >> mail # user >> Replicas for partition are dead


+
Jason Huang 2013-03-19, 20:35
+
Jun Rao 2013-03-20, 04:31
+
Jason Huang 2013-03-20, 13:58
+
Jun Rao 2013-03-20, 14:39
Copy link to this message
-
Re: Replicas for partition are dead
The 0.8 version I use was built from trunk last Dec. Since then, this
error happened 3 times. Each time we had to remove all the ZK and
Kafka log data and restart the services.

I will try newer versions with more recent patches and keep monitoring it.

thanks!

Jason

On Wed, Mar 20, 2013 at 10:39 AM, Jun Rao <[EMAIL PROTECTED]> wrote:
> Ok, so you are using the same broker id. What the error is saying is that
> broker 1 doesn't seem to be up.
>
> Not sure what revision of 0.8 you are using. Could you try the latest
> revision in 0.8 and see if the problem still exists? You may have to wipe
> out all ZK and Kafka data first since some ZK data structures have been
> rename a few weeks ago.
>
> Thanks,
>
> Jun
>
> On Wed, Mar 20, 2013 at 6:57 AM, Jason Huang <[EMAIL PROTECTED]> wrote:
>
>> I restarted the zookeeper server first, then broker. It's the same
>> instance of kafka 0.8 and I am using the same config file. In
>> server.properties I have: brokerid=1
>>
>> Is that sufficient to ensure the broker get restarted with the same
>> broker id as before?
>>
>> thanks,
>>
>> Jason
>>
>> On Wed, Mar 20, 2013 at 12:30 AM, Jun Rao <[EMAIL PROTECTED]> wrote:
>> > Did the broker get restarted with the same broker id?
>> >
>> > Thanks,
>> >
>> > Jun
>> >
>> > On Tue, Mar 19, 2013 at 1:34 PM, Jason Huang <[EMAIL PROTECTED]>
>> wrote:
>> >
>> >> Hello,
>> >>
>> >> My kafka (0.8) server went down today for unknown reason and when I
>> >> restarted both zookeeper and kafka server I got the following error at
>> >> the kafka server log:
>> >>
>> >> [2013-03-19 13:39:16,131] INFO [Partition state machine on Controller
>> >> 1]: Invoking state change to OnlinePartition for partitions
>> >> (kafka.controller.PartitionStateMachine)
>> >> [2013-03-19 13:39:16,262] INFO [Partition state machine on Controller
>> >> 1]: Electing leader for partition
>> >> [topic_a937ac27-1883-4ca0-95bc-c9a740d08947, 0]
>> >> (kafka.controller.PartitionStateMachine)
>> >> [2013-03-19 13:39:16,451] ERROR [Partition state machine on Controller
>> >> 1]: State change for partition
>> >> [topic_a937ac27-1883-4ca0-95bc-c9a740d08947, 0] from OfflinePartition
>> >> to OnlinePartition failed (kafka.controller.PartitionStateMachine)
>> >> kafka.common.PartitionOfflineException: All replicas for partition
>> >> [topic_a937ac27-1883-4ca0-95bc-c9a740d08947, 0] are dead. Marking this
>> >> partition offline
>> >>         at
>> >>
>> kafka.controller.PartitionStateMachine.electLeaderForPartition(PartitionStateMachine.scala:300)
>> >> .....
>> >> Caused by: kafka.common.PartitionOfflineException: No replica for
>> >> partition ([topic_a937ac27-1883-4ca0-95bc-c9a740d08947, 0]) is alive.
>> >> Live brokers are: [Set()], Assigned replicas are: [List(1)]
>> >> .......
>> >>
>> >> I am using one single server to host kafka and zookeeper. Replication
>> >> factor is set to 1.
>> >>
>> >> This happened for all the existing topics. Not sure how this happened
>> >> but it appeared to be a bug. I did some search and the only possible
>> >> fix for this bug seems to be KAFKA-708.
>> >>
>> >> Any comments on this?  Thanks!
>> >>
>> >> Jason
>> >>
>>

 
+
Jun Rao 2013-03-20, 16:15
+
Jason Huang 2013-03-22, 13:51
+
Jun Rao 2013-03-22, 15:07
+
Jason Huang 2013-03-22, 15:10
+
Neha Narkhede 2013-03-22, 15:17
+
Jason Huang 2013-03-22, 16:09