Thanks will take a look..
On Wed, Feb 04, 2015 at 11:33:03PM -0800, Sumit Rangwala wrote:
> >
> > Any idea why you have session expirations? This is typically due to GC
> >> and/or flaky network. Regardless, we should be handling that scenario
> >> as well. However, your logs seem incomplete. Can
>
> Any idea why you have session expirations? This is typically due to GC
>> and/or flaky network. Regardless, we should be handling that scenario
>> as well. However, your logs seem incomplete. Can you redo this and
>> perhaps keep the set up running a little longer and send over those
>> logs?
>
On Wed, Feb 4, 2015 at 10:16 PM, Harsha wrote:
>
>whats your zookeeper.session.timeout.ms value
>
3 (30sec)
Sumit
>
> On Wed, Feb 4, 2015, at 09:35 PM, Sumit Rangwala wrote:
> > On Wed, Feb 4, 2015 at 6:14 PM, Joel Koshy wrote:
> >
> > > I took a look at your logs. I agree with
whats your zookeeper.session.timeout.ms value
On Wed, Feb 4, 2015, at 09:35 PM, Sumit Rangwala wrote:
> On Wed, Feb 4, 2015 at 6:14 PM, Joel Koshy wrote:
>
> > I took a look at your logs. I agree with Harsh that the logs seem
> > truncated. The basic issue though is that you have session
On Wed, Feb 4, 2015 at 6:14 PM, Joel Koshy wrote:
> I took a look at your logs. I agree with Harsh that the logs seem
> truncated. The basic issue though is that you have session expirations
> and controller failover. Broker 49554 was the controller and hosted
> some partition(s) of LAX1-GRIFFIN-
I took a look at your logs. I agree with Harsh that the logs seem
truncated. The basic issue though is that you have session expirations
and controller failover. Broker 49554 was the controller and hosted
some partition(s) of LAX1-GRIFFIN-r13-1423001701601. After controller
failover the new control
>
>
>> I have since stopped the container so I cannot say if
> LAX1-GRIFFIN-r45-142388317 was one of the topic in "marked for
> deletion" forever. However, there were many topics (at least 10 of them)
> that were perennially in "marked for deletion" state.
>
>
I have the setup to recreate the
On Tue, Feb 3, 2015 at 9:47 PM, Harsha wrote:
> Sumit,
> I grepped logs for this topic "LAX1-GRIFFIN-r13-1423001701601"
> it looks like topic partitions are getting deleted in
> state-change.log and this happens around 22:59 and server.log
> has data till 22:5
On Tue, Feb 3, 2015 at 9:47 PM, Harsha wrote:
> Sumit,
> I grepped logs for this topic "LAX1-GRIFFIN-r13-1423001701601"
> it looks like topic partitions are getting deleted in
> state-change.log and this happens around 22:59 and server.log
> has data till 22:5
On Tue, Feb 3, 2015 at 9:19 PM, Harsha wrote:
> you are probably handling it but there is a case where you call
> deleteTopic and kafka goes through delete topic process but your
> consumer is running probably made a TopicMetadataRequest for the same
> topic which can re-create the topic with the
Sumit,
I grepped logs for this topic "LAX1-GRIFFIN-r13-1423001701601"
it looks like topic partitions are getting deleted in
state-change.log and this happens around 22:59 and server.log
has data till 22:59.
I looked for other deleted topic "LAX1-GRIFFIN-r45-1423
you are probably handling it but there is a case where you call
deleteTopic and kafka goes through delete topic process but your
consumer is running probably made a TopicMetadataRequest for the same
topic which can re-create the topic with the default num.partitions and
replication.factor. Did you
On Tue, Feb 3, 2015 at 6:48 PM, Harsha wrote:
> Sumit,
>lets say you are deleting a older topic "test1" do you have any
>consumers running simultaneously for the topic "test1" while
>deletion of topic going on.
>
Yes it is the case. However, after a small period of time
Sumit,
lets say you are deleting a older topic "test1" do you have any
consumers running simultaneously for the topic "test1" while
deletion of topic going on.
-Harsha
On Tue, Feb 3, 2015, at 06:17 PM, Joel Koshy wrote:
> Thanks for the logs - will take a look tomorrow unless
Thanks for the logs - will take a look tomorrow unless someone else
gets a chance to get to it today.
Joel
On Tue, Feb 03, 2015 at 04:11:57PM -0800, Sumit Rangwala wrote:
> On Tue, Feb 3, 2015 at 3:37 PM, Joel Koshy wrote:
>
> > Hey Sumit,
> >
> > I thought you would be providing the actual ste
On Tue, Feb 3, 2015 at 3:37 PM, Joel Koshy wrote:
> Hey Sumit,
>
> I thought you would be providing the actual steps to reproduce :)
>
I want to but some proprietary code prevents me to do it.
> Nevertheless, can you get all the relevant logs: state change logs and
> controller logs at the ver
Hey Sumit,
I thought you would be providing the actual steps to reproduce :)
Nevertheless, can you get all the relevant logs: state change logs and
controller logs at the very least and if possible server logs and send
those over?
Joel
On Tue, Feb 03, 2015 at 03:27:43PM -0800, Sumit Rangwala wro
In my setup kafka brokers are set for auto topic creation. In the scenario
below a node informs other nodes (currently 5 in total) about a number of
new (non-existent) topics, and all the nodes almost simultaneously open a
consumer for each of those topics. Sometime later another node informs all
18 matches
Mail list logo
