| clear query|facets|time |
Search criteria: .
Results from 81 to 90 from
1163 (0.233s).
|
|
|
Loading phrases to help you
refine your search...
|
|
Re: configuring kafka - Kafka - [mail # user]
|
|
...1. Take a look at #6 in http://kafka.apache.org/faq.html2. Typically, you want each batch to have 10s-100s KB.3. In the latest 0.8, we have 3 network threads and 8 io threads. Theseshould be...
|
|
|
Author: Jun Rao,
2013-04-26, 04:05
|
|
|
Re: reassign Partition leader - Kafka - [mail # user]
|
|
...Yes, there is an admin tool PreferredReplicaLeaderElectionCommand that youcan run.Thanks,JunOn Thu, Apr 25, 2013 at 12:49 AM, Eureka wrote:> Hi All,>> Here is my test scenario...
|
|
|
Author: Jun Rao,
2013-04-25, 15:06
|
|
|
Re: LeaderNotAvailable Exception - Kafka - [mail # user]
|
|
...Hmm, then it's weird. Could you try the latest code in 0.8? If the problempersists, could you file a jira and attach the broker log?Thanks,JunOn Wed, Apr 24, 2013 at 11:00 PM, Yin Yin ...
|
|
|
Author: Jun Rao,
2013-04-25, 14:21
|
|
|
Re: Kafka wiki Documentation conventions - looking for feedback - Kafka - [mail # user]
|
|
...For #5, if you always start from an offset returned by getOffsetBefore, itwon't happen since getOffsetBefore will always return offset at thecompressed messageSet boundary. However, if you s...
|
|
|
Author: Jun Rao,
2013-04-25, 04:08
|
|
|
Re: acknowledged mode for performance test - Kafka - [mail # user]
|
|
...It typically means wait until all replicas have received the message. Fordetails, seehttp://www.slideshare.net/junrao/kafka-replication-apachecon2013 (-1 ==wait until message is committed).T...
|
|
|
Author: Jun Rao,
2013-04-25, 04:05
|
|
|
Re: LeaderNotAvailable Exception - Kafka - [mail # user]
|
|
...What output do you get if you add the --unavailable-partitions option inlist topic?Thanks,JunOn Wed, Apr 24, 2013 at 5:20 PM, Yin Yin wrote:> This exception happens only once. But t...
|
|
|
Author: Jun Rao,
2013-04-25, 03:58
|
|
|
Re: Exception from kafka server 0.8 - Kafka - [mail # user]
|
|
...Actually, replication factor 2 means a total of 2 replicas (a leader and afollower). If the leader is down, another replica should automatically takeover as the leader. There could be some t...
|
|
|
Author: Jun Rao,
2013-04-25, 03:52
|
|
|
Re: Advice on cacheing producers - Kafka - [mail # user]
|
|
...You don't want to create a new producer for every batch of messages.Keeping the producer connection open should be fine.Thanks,JunOn Wed, Apr 24, 2013 at 11:45 AM, Karl Kirch wrote:>...
|
|
|
Author: Jun Rao,
2013-04-25, 03:49
|
|
|
Re: Broker List vs Zookeeper - Kafka - [mail # user]
|
|
...Typically, if you use broker list, you will set up a VIP in a load balancerin front of all brokers.Thanks,JunOn Wed, Apr 24, 2013 at 11:41 AM, Karl Kirch wrote:> Any advice on using...
|
|
|
Author: Jun Rao,
2013-04-25, 03:47
|
|
|
Re: Multiple log directories in 0.7.2 - Kafka - [mail # user]
|
|
...Anand,We try not to do too much development on 0.7 except for critical bug fixes.Thanks,JunOn Wed, Apr 24, 2013 at 10:43 AM, anand nalya wrote:> Hi,>> I'm aware of KAFKA-188 w...
|
|
|
Author: Jun Rao,
2013-04-25, 03:45
|
|
|
|
