Hello,
I am using 0.8.0-beta1, with two brokers. After
using bin/kafka-create-topic.sh created a topic named "ead_click",
under bin/kafka-list-topic.sh I can see the topic seems to be created
successfully:
[2013-11-01 15:49:35,388] INFO zookeeper state changed (SyncConnected)
(org.I0Itec.zkclient
I've got the following exception running produce/consume loop for several
hours,
that was just single exception, but during that time both producers and
consumers slowed down a lot. After that looks like everything works fine,
though i have
suspicion some messages were lost.
Can anyone explain what
When machine with kafka dies, most often broker cannot start itself with
errors about index files being corrupt. After i delete them manually broker
usually can boot up successfully. Shouldn't kafka try to delete/rebuild
broken index files itself?
Also this exception looks a bit weird:
java.lang.I
It is proportional to the number of topics but still seems too long. Did
the broker have high queue time on all requests? Also how many topics
existed on this cluster?
Thanks,
Neha
On Oct 31, 2013 10:56 PM, "Jason Rosenberg" wrote:
> In this case, it appears to have gone on for 104 seconds. Sho
Hello,
This might be a stupid question, but
I'm seeing a new tag 0.8.0 in the kafka git repository.
Does this mean, a final 0.8.0 release has been created and can be
fetched using this tag?
I couldn't find an official announcement though. Should one better wait
for it,
or did I overlook it?
Since
It is undergoing voting on release candidate 1 until November 4th. So yes,
very close to being released.
Thanks,
Neha
On Nov 1, 2013 3:10 AM, "Maier, Dr. Andreas"
wrote:
> Hello,
>
> This might be a stupid question, but
> I'm seeing a new tag 0.8.0 in the kafka git repository.
> Does this mean,
When you say the Kafka dies, did you mean a kill -9 or kill -15 ? I agree
that Kafka should try and delete/rebuild index files, but in this case, it
sounds like a bug. Can you file a JIRA?
Thanks,
Neha
On Fri, Nov 1, 2013 at 1:51 AM, Kane Kane wrote:
> When machine with kafka dies, most often
I think you are hitting https://issues.apache.org/jira/browse/KAFKA-824.
Was the consumer being shutdown at that time?
On Fri, Nov 1, 2013 at 1:28 AM, Kane Kane wrote:
> I've got the following exception running produce/consume loop for several
> hours,
> that was just single exception, but duri
The following exception seems suspicious and I think it is fixed in 0.8
HEAD. Do you mind trying 0.8 HEAD and see if this can still be reproduced ?
ava.util.
NoSuchElementException: key not found: /disk1/test/kafka-data
at scala.collection.MapLike$class.default(MapLike.scala:223)
at scal
Hi guys,
While running latest kafka trunk code I'm getting warning in server.log:
[2013-11-01 15:06:11,419] WARN Property log.cleanup.interval.mins is
not valid (kafka.utils.VerifiableProperties)
Do I need to ignore this message?
--
Thanks,
Viktor
I'm excited about Kafka but want to be sure it is ready (or will be soon)
for a critical data pipeline.
What are the showstoppers, if any? eg.
https://issues.apache.org/jira/browse/KAFKA-156
mentioned here:
http://stackoverflow.com/questions/12130481/is-kafka-ready-for-production-use/12764663#1
Priya, if you want you can look at RequestPurgatory.scala for some more
details.
The config is the size of the atomic requestCounter.
Basically the purge in the purgatory is a way to check if the request has
been satisfied and delayed and can get removed. It is a background scan
when the size re
Richard, KAFKA-156 is something that the client could take care of. It is
2 years old and a bit of a hack suggestion for what Replication in 0.8.0
has become and I would argue we close it (we should go through all the
ticket prior to 0.9 and chat about them, separate thread I will start on
dev).
Neha, yes when i kill it with -9, sure I will file a bug.
Thanks.
On Fri, Nov 1, 2013 at 3:43 AM, Neha Narkhede wrote:
> When you say the Kafka dies, did you mean a kill -9 or kill -15 ? I agree
> that Kafka should try and delete/rebuild index files, but in this case, it
> sounds like a bug. Ca
Hello Neha, I think i might be hitting this. I didn't shutdown the consumer
(at least intentionally). Basically it was just an attempt to pipe ~1T
through kafka, i would wild guess it's related to log expansion. Because
around the time it happened i saw messages about expanding log, is it
possible
Did you check the error code associated with each partition in the fetch
response?
Thanks,
Jun
On Thu, Oct 31, 2013 at 9:59 PM, Lu Xuechao wrote:
> No. The simple consumer does receive some responses and can iterate the
> loop:
>
> for (MessageAndOffset messageAndOffset : fetchResponse.messag
It seems in trunk log.cleanup.interval.mins is renamed to
log.retention.check.interval.ms. We should put in the new name in
config/server.properties. Could you file a jira?
Thanks,
Jun
On Fri, Nov 1, 2013 at 6:10 AM, Viktor Kolodrevskiy <
viktor.kolodrevs...@gmail.com> wrote:
> Hi guys,
>
> Wh
Are you using the latest code in the 0.8 branch?
Thanks,
Jun
On Fri, Nov 1, 2013 at 7:36 AM, Kane Kane wrote:
> Neha, yes when i kill it with -9, sure I will file a bug.
>
> Thanks.
>
>
> On Fri, Nov 1, 2013 at 3:43 AM, Neha Narkhede >wrote:
>
> > When you say the Kafka dies, did you mean a
Hello Jun Rao, not it's not the head, I've compiled it a couple of weeks
ago. Should i try with latest?
On Fri, Nov 1, 2013 at 7:58 AM, Jun Rao wrote:
> Are you using the latest code in the 0.8 branch?
>
> Thanks,
>
> Jun
>
>
> On Fri, Nov 1, 2013 at 7:36 AM, Kane Kane wrote:
>
> > Neha, yes w
Jun,
I've seen this happen in the latest code in trunk, need to remember how to
reproduce it.
Thanks,
Neha
On Nov 1, 2013 7:59 AM, "Jun Rao" wrote:
> Are you using the latest code in the 0.8 branch?
>
> Thanks,
>
> Jun
>
>
> On Fri, Nov 1, 2013 at 7:36 AM, Kane Kane wrote:
>
> > Neha, yes when
Filed a bug meanwhile, https://issues.apache.org/jira/browse/KAFKA-1112
On Fri, Nov 1, 2013 at 8:15 AM, Neha Narkhede wrote:
> Jun,
>
> I've seen this happen in the latest code in trunk, need to remember how to
> reproduce it.
>
> Thanks,
> Neha
> On Nov 1, 2013 7:59 AM, "Jun Rao" wrote:
>
> >
Currently the index files will only be deleted on startup if there are any
.swap file indicating the server crashed while opening the log segments. We
should probably change this logic.
Guozhang
On Fri, Nov 1, 2013 at 8:16 AM, Kane Kane wrote:
> Filed a bug meanwhile, https://issues.apache.org
Richard - I can't speak about 0.8x, but we've been using Kafka in a most
critical capacity on production since the early days @ Datadog.
And...
- It never failed (!)
- It scaled predictably
- Its performance was consistent
- The simplicity of its design made it really easy to reason about
We
Could you send the entire consumer log?
On Fri, Nov 1, 2013 at 7:45 AM, Kane Kane wrote:
> Hello Neha, I think i might be hitting this. I didn't shutdown the consumer
> (at least intentionally). Basically it was just an attempt to pipe ~1T
> through kafka, i would wild guess it's related to log
Done!
https://issues.apache.org/jira/browse/KAFKA-1113
--
Thanks,
Viktor
2013/11/1 Jun Rao :
> It seems in trunk log.cleanup.interval.mins is renamed to
> log.retention.check.interval.ms. We should put in the new name in
> config/server.properties. Could you file a jira?
>
> Thanks,
>
> Jun
>
>
I'm a fan of kafka as well. We've been using 0.7.2 for about a year.
I recommend it strongly.
But I will point one thing. Not an issue with Kafka itself but when the
client side has failed, choosing what offset to reset to is not an exact
science. You will have to decide how much data loss or dat
Hmm,
It looks as though adding multicast support is more an issue with the Metrics
code rather than with kafka-ganglia. The only way I could see to tell
GangliaReporter to use multicast when sending messages was to pass it a
GangliaMessageBuilder that constructed GangliaMessages using a proper
Hello Neha, rest of the log goes to my consumer code, are you interested
in? It's a little bit modified version of ConsoleConsumer.
Thanks.
On Fri, Nov 1, 2013 at 9:24 AM, Neha Narkhede wrote:
> Could you send the entire consumer log?
>
>
> On Fri, Nov 1, 2013 at 7:45 AM, Kane Kane wrote:
>
>
checked fetchResponse.hasError() but has no error.
On Fri, Nov 1, 2013 at 7:45 AM, Jun Rao wrote:
> Did you check the error code associated with each partition in the fetch
> response?
>
> Thanks,
>
> Jun
>
>
> On Thu, Oct 31, 2013 at 9:59 PM, Lu Xuechao wrote:
>
> > No. The simple consumer do
Neha,
This cluster has on the order of 750 topics.
It looks like if I add a 20 second delay before placing a broker into the
vip for metadata requests, it never seems to have this issue. So I'm not
sure about the 104 seconds number, other than that was how long the flood
of "Topic creation" log
Guozhang,
I have to agree with Priya the doc isn't very clear. Although the
configuration is documented, it is simply rewording the name of the config,
which isn't particularly useful if you want more information about what the
purgatory is. I searched the whole wiki and doc and could not find any
In http://kafka.apache.org/documentation.html#monitoring,
ISR expansion rate
"kafka.server":name="ISRShrinksPerSec",type="ReplicaManager"See above
I believe this should be
"kafka.server":name="IsrExpandsPerSec",type="ReplicaManager"
-Andrew Otto
Marc, thanks for writing that up. I think it is worth adding some
details on the request-purgatory on a wiki (Jay had started a wiki
page for kafka internals [1] a while ago, but we have not had time to
add much to it since.) Your write-up could be reviewed and added
there. Do you have edit permiss
Unclean shutdown could result in data loss - since you are moving
leadership to a replica that has fallen out of ISR. i.e., it's log end
offset is behind the last committed message to this partition.
>>> But if data is written with 'request.required.acks=-1', no data s
Thanks for catching that. Pushed the fix.
-Neha
On Fri, Nov 1, 2013 at 12:58 PM, Andrew Otto wrote:
> In http://kafka.apache.org/documentation.html#monitoring,
>
> ISR expansion rate
> "kafka.server":name="ISRShrinksPerSec",type="ReplicaManager"See above
>
>
> I believe this should be
>
>
Sorry no clue about that - anyone else know?
On Mon, Oct 28, 2013 at 10:41 AM, Roger Hoover wrote:
> Joel,
>
> Thank you! This is very helpful.
>
> What I notice now is that it works for Test classes that
> extend org.scalatest.junit.JUnit3Suite. There are other tests in the
> codebase that us
The mbeans are explained here -
http://kafka.apache.org/documentation.html#monitoring. Look for *QueueTimeMs
Thanks,
Neha
On Fri, Nov 1, 2013 at 12:14 PM, Jason Rosenberg wrote:
> Neha,
>
> This cluster has on the order of 750 topics.
>
> It looks like if I add a 20 second delay before placing
For supporting more durability at the expense of availability, we have a
JIRA that we will fix on trunk. This will allow you to configure the
default as well as per topic durability vs availability behavior -
https://issues.apache.org/jira/browse/KAFKA-1028
Thanks,
Neha
On Fri, Nov 1, 2013 at 1
Hello Kane, do you mind attach your index file causing the issue if you
still have it?
Guozhang
On Fri, Nov 1, 2013 at 8:42 AM, Guozhang Wang wrote:
> Currently the index files will only be deleted on startup if there are any
> .swap file indicating the server crashed while opening the log seg
Hello, yes, I can easily reproduce it, will send it to you asap.
Thanks.
On Fri, Nov 1, 2013 at 2:05 PM, Guozhang Wang wrote:
> Hello Kane, do you mind attach your index file causing the issue if you
> still have it?
>
> Guozhang
>
>
> On Fri, Nov 1, 2013 at 8:42 AM, Guozhang Wang wrote:
>
>
Which offset did you use for fetching? Is there data in the kafka log dir?
Thanks,
Jun
On Fri, Nov 1, 2013 at 11:48 AM, Lu Xuechao wrote:
> checked fetchResponse.hasError() but has no error.
>
>
> On Fri, Nov 1, 2013 at 7:45 AM, Jun Rao wrote:
>
> > Did you check the error code associated wi
I think
addSbtPlugin("com.eed3si9n" % "sbt-assembly" % "0.8.8")
Should be updated to 0.9.0 at least to successfully compile. I've had
an issue with assembly-package-dependency.
The consumer starts from offset 0. Yes, in the log dir.
On Fri, Nov 1, 2013 at 4:06 PM, Jun Rao wrote:
> Which offset did you use for fetching? Is there data in the kafka log dir?
>
> Thanks,
>
> Jun
>
>
> On Fri, Nov 1, 2013 at 11:48 AM, Lu Xuechao wrote:
>
> > checked fetchResponse.hasError(
Hi Joel,
I used to have edit to the wiki, I made a few additions to it a while ago
but it's seem I don't have it anymore. It might have been lost in the
confluence update. I would be glad to add what I have written if I get it
back. Otherwise, feel free to paste my words in one of the pages, I don
Yes, could you send around the log4j file of the consumer around the time
of the error in question.
Thanks,
Neha
On Fri, Nov 1, 2013 at 11:35 AM, Kane Kane wrote:
> Hello Neha, rest of the log goes to my consumer code, are you interested
> in? It's a little bit modified version of ConsoleConsu
To edit the Wiki you need to send an ICLA
http://www.apache.org/licenses/#clas to Apache and then once that is done
an email to priv...@kafka.apache.org (or to me and I will copy private)
with your Wiki username and that you sent the ICLA to Apache.
Then, I can add you to edit the Wiki.
/
Does the problem exist with trunk? If so, could you open a jira and submit
a patch?
Thanks,
Jun
On Fri, Nov 1, 2013 at 4:14 PM, Kane Kane wrote:
> I think
> addSbtPlugin("com.eed3si9n" % "sbt-assembly" % "0.8.8")
>
> Should be updated to 0.9.0 at least to successfully compile. I've had
> an i
Did you make sure the fetch size in the fetch request is larger than the
size of a single message?
Thanks,
Jun
On Fri, Nov 1, 2013 at 5:07 PM, Lu Xuechao wrote:
> The consumer starts from offset 0. Yes, in the log dir.
>
>
> On Fri, Nov 1, 2013 at 4:06 PM, Jun Rao wrote:
>
> > Which offset d
Yes, I've had problem, which resolved with updating sbt-assembly. Will
open a ticket and provide a patch.
On Fri, Nov 1, 2013 at 8:43 PM, Jun Rao wrote:
> Does the problem exist with trunk? If so, could you open a jira and submit
> a patch?
>
> Thanks,
>
> Jun
>
>
> On Fri, Nov 1, 2013 at 4:14 PM
I had only 1 topic with 45 partitions replicated across 3 brokers.
After several hours of uploading some data to kafka 1 broker died with
the following exception.
I guess i can fix it raising limit for open files, but I wonder how it
happened under described circumstances.
[2013-11-02 00:19:14,86
In response to Joel's point, I think I do understand that messages can be
lost, if in fact we have dropped down to only 1 member in the ISR at the
time the message is written, and then that 1 node goes down.
What I'm not clear on, is the conditions under which a node can drop out of
the ISR. You
Unfortunately I've recompiled and installed kafka from scratch today,
deleting all data. I'm still doing these tests, will send you
stacktrace if I would be able to repro it.
Thanks for help!
On Fri, Nov 1, 2013 at 6:13 PM, Neha Narkhede wrote:
> Yes, could you send around the log4j file of the
52 matches
Mail list logo
