I would suggest creating a JIRA and describing in detail what was going on
in the cluster when this happened, and posting the associated broker /
state change / controller logs.
Thanks,
Apurva
On Wed, Dec 14, 2016 at 3:28 AM, Mazhar Shaikh
wrote:
> Hi All,
>
> I am
[
https://issues.apache.org/jira/browse/KAFKA-4477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15746799#comment-15746799
]
Apurva Mehta commented on KAFKA-4477:
-
[~tdevoe], thanks for sharing all your extend broker logs
@Becket and @Rajini,
Thanks for those comments. You raise some very astute points. I will
address a subset of them here.
One common thread across your emails has to do with the notion of a 'batch'
of messages from the consumer's point of view. In particular, Rajini's
points 12 and 16, and
[
https://issues.apache.org/jira/browse/KAFKA-4477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15743529#comment-15743529
]
Apurva Mehta commented on KAFKA-4477:
-
Thanks [~tdevoe], These seem to be the broker logs, and nothing
[
https://issues.apache.org/jira/browse/KAFKA-4477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15743283#comment-15743283
]
Apurva Mehta commented on KAFKA-4477:
-
Hi [~tdevoe]: Thanks for sharing those logs. I had a look
[
https://issues.apache.org/jira/browse/KAFKA-4477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15734049#comment-15734049
]
Apurva Mehta commented on KAFKA-4477:
-
Also, could you share the values of the following metrics
[
https://issues.apache.org/jira/browse/KAFKA-4477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15733238#comment-15733238
]
Apurva Mehta commented on KAFKA-4477:
-
Hi [~tdevoe], [~jj83], [~michael.andre.pearce] : I had a look
[
https://issues.apache.org/jira/browse/KAFKA-4477?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Apurva Mehta reassigned KAFKA-4477:
---
Assignee: Apurva Mehta
> Node reduces its ISR to itself, and doesn't recover. Other nodes
+1 (non-binding)
On Wed, Dec 7, 2016 at 10:05 AM, Jason Gustafson wrote:
> +1 Thanks for the KIP!
>
> On Wed, Dec 7, 2016 at 2:53 AM, Ismael Juma wrote:
>
> > Thanks for the KIP, Vahid. +1 (binding)
> >
> > On Mon, Dec 5, 2016 at 6:16 PM, Vahid S
Hi Ben,
Now, on to your first question of how deal with consumer rebalances. The
short answer is that the application needs to ensure that the the
assignment of input partitions to appId is consistent across rebalances.
For Kafka streams, they already ensure that the mapping of input partitions
Hi Ben,
Those are both great questions. I will tackle the second one now, and
address the first one a bit later.
AppIds are prerequisite for using transactions, and must be consistent
across across application sessions. They are the mechanism by which
transaction recovery can occur across
Hi Jay,
Thanks for your comments. Answers to some of your points are below:
2. There have been long debates about the necessity of the initTransactions
method. Let's consider the options for doing without the initTransactions
method:
- If we do it on demand, we have to consider that the
ore work for the user, though
> of
> > course it would be transparent to for Kafka Streams users.
> >
> > One final note. I've described above how to get the strongest guarantees
> > that this work is capable of providing in an auto-scaling environment. We
> > al
Hi Ismael,
That is a good suggestion. We did not plan to move the design to a wiki,
but I think it is valuable to move at least the message format and RPC
changes to the wiki. We shall do so once the design is close to final so
that we do not have to edit multiple places as we iterate.
Thanks,
ucer with the same app-id, it creates a pid and appends (app-id,
> pid,
> > > epoch) into the transaction log.
> > >
> > > What about if the app-id/pid pair already exists and we increment the
> > > epoch? Should we append (app-id, pid, epoch++) to the transact
to the transaction log.
>
> What about if the app-id/pid pair already exists and we increment the
> epoch? Should we append (app-id, pid, epoch++) to the transaction log? I
> think we should, but step 2 doesn't mention this.
>
> On Wed, Nov 30, 2016 at 5:35 PM, Apurva Mehta <a
Thanks for your comments, let me deal with your second point regarding
merging the __consumer-offsets and transactions topic.
Needless to say, we considered doing this, but chose to keep them separate
for the following reasons:
1. Your assumption that group.id and transaction.app.id can be
+1 (non-binding)
On Wed, Nov 30, 2016 at 1:20 PM, Xavier Léauté wrote:
> FYI, Based on internal feedback I renamed AssignedReplicasCount to simply
> be called ReplicasCount.
>
> On Tue, Nov 29, 2016 at 7:56 PM Neha Narkhede wrote:
>
> > This seems
+1 (non-binding)
On Wed, Nov 30, 2016 at 2:00 PM, Jason Gustafson wrote:
> +1. Thanks for the KIP!
>
> On Wed, Nov 30, 2016 at 1:47 PM, Gwen Shapira wrote:
>
> > +1 (binding)
> >
> > On Wed, Nov 30, 2016 at 1:34 PM, Xavier Léauté
> >
If the bug is in Kafka, here is the process for submitting a fix:
http://kafka.apache.org/contributing
If the bug is in the mesos-kafka framework, I think you should look up that
project and find out how to commit a bugfix there. I think it should not be
more complicated than submitting a PR.
Not sure if this answers your question, but the entry point for handling
the FetchRequest is here:
https://github.com/apache/kafka/blob/trunk/core/src/main/scala/kafka/server/KafkaApis.scala#L436
That file has the entry points for the handlers of other requests as well.
Thanks,
Apurva
On Sun,
[
https://issues.apache.org/jira/browse/KAFKA-4215?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Apurva Mehta resolved KAFKA-4215.
-
Resolution: Won't Fix
This is the expected behavior. When the replication factor is 1, and when
[
https://issues.apache.org/jira/browse/KAFKA-4215?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15519566#comment-15519566
]
Apurva Mehta commented on KAFKA-4215:
-
Another observation: the consumer only misses messages when
[
https://issues.apache.org/jira/browse/KAFKA-4215?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15519348#comment-15519348
]
Apurva Mehta commented on KAFKA-4215:
-
Here is [~junrao]'s theory for what is going on:
{quote
[
https://issues.apache.org/jira/browse/KAFKA-4214?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Apurva Mehta updated KAFKA-4214:
Description:
Due to KAFKA-4204, we never realized that the existing system test for testing
Apurva Mehta created KAFKA-4215:
---
Summary: Consumers miss messages during partition reassignment
Key: KAFKA-4215
URL: https://issues.apache.org/jira/browse/KAFKA-4215
Project: Kafka
Issue Type
Apurva Mehta created KAFKA-4214:
---
Summary: kafka-reassign-partitions fails all the time when brokers
are bounced during reassignment
Key: KAFKA-4214
URL: https://issues.apache.org/jira/browse/KAFKA-4214
[
https://issues.apache.org/jira/browse/KAFKA-4213?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Work on KAFKA-4213 started by Apurva Mehta.
---
> Add system tests for replication throttling (KIP
Apurva Mehta created KAFKA-4213:
---
Summary: Add system tests for replication throttling (KIP-73)
Key: KAFKA-4213
URL: https://issues.apache.org/jira/browse/KAFKA-4213
Project: Kafka
Issue Type
[
https://issues.apache.org/jira/browse/KAFKA-4204?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15517207#comment-15517207
]
Apurva Mehta edited comment on KAFKA-4204 at 9/23/16 6:36 PM:
--
Another issue
[
https://issues.apache.org/jira/browse/KAFKA-4204?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15517207#comment-15517207
]
Apurva Mehta commented on KAFKA-4204:
-
Another issue with `KafkaService.verify_reassign_partitions
[
https://issues.apache.org/jira/browse/KAFKA-4204?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Apurva Mehta updated KAFKA-4204:
Assignee: Apurva Mehta
> KafkaService.verify_reassign_partitions is a no
Apurva Mehta created KAFKA-4204:
---
Summary: KafkaService.verify_reassign_partitions is a no-op
Key: KAFKA-4204
URL: https://issues.apache.org/jira/browse/KAFKA-4204
Project: Kafka
Issue Type
401 - 433 of 433 matches
Mail list logo