Sure no problem. I actually did run the ‘localhost’ command to generate output
for the e-mail, but that was a typo: I always use the right IP. But I ran it
again, this time with the IP of the VM. I ran the command from my Mac (the
client). Thanks for taking a look! Here it is:
We are currently using Kafka 0.8.1.1 with log compaction in order to
provide streams of messages to our clients.
As well as constantly consuming the stream, one of our use cases is to
provide a snapshot, meaning the user will receive a copy of every message
at least once.
Each one of these
Yes, the topic I am producing to exists. I can produce to it fine when running
the kafa-console-producer.sh tool from the Kafka node (the VM) itself (in which
case I can either use localhost or the public-facing IP):
Here is where I produce messages (running on the Kafka node):
I think your log did show that your are connecting to localhost:9092:
[2015-02-17 20:43:32,622] WARN Fetching topic metadata with correlation id
0 for topics [Set(test)] from broker [id:0,host:localhost,port:9092]
failed (kafka.client.ClientUtils$)
java.nio.channels.ClosedChannelException
Can
I also tried running the producer from the Mac client again, but this time with
TRACE and DEBUG options un-commented from the log4j.properties file on the VM
server. It seems that the connection is established (on port 50045) and bytes
are being read from the client (192.168.241.1). Then
OK. Steve Miller helped me solve the problem. I needed to explicitly set
advertised.host.name to advertised.host.name=192.168.241.128. The logs showed
the producer could connect to 9092 but when it was told which hosts to connect
to to queue messages it got unresolvable hosts. By setting this
Reading offsets looks like it's compatible across 0.8.1 and 0.8.2.
However, we cannot use the update logic in ImportZkOffsets, since we want to
store offsets in the broker in 0.8.2.
It looks like SimpleConsumer.commitOffsets() would work with either version. Is
there a better way?
-Suren
Do you have to separate the snapshot from the normal update flow.
I've used a compacting kafka topic as the source of truth to a solr
database and fed the topic both with real time updates and snapshots from
a hive job. This worked very well. The nice point is that there is a
seamless transition
We have fixed the issue in KAFKA-1952. We will wait for a few more days to
see if any new issue comes up. After that, we will do an 0.8.2.1 release.
Thanks,
Jun
On Fri, Feb 13, 2015 at 3:28 PM, Jay Kreps jay.kr...@gmail.com wrote:
Hey all,
We found an issue in 0.8.2 that can lead to high
Does it make sense to wait, I don't think people will upgrade without the
patched version and I think we should release it to unblock people.
-Jay
On Wed, Feb 18, 2015 at 1:43 PM, Jun Rao j...@confluent.io wrote:
We have fixed the issue in KAFKA-1952. We will wait for a few more days to
see
If you catch up off a compacted topic and keep consuming then you will
become consistent with the log.
I think what you are saying is that you want to create a snapshot from the
Kafka topic but NOT do continual reads after that point. For example you
might be creating a backup of the data to a
Hi Guozhang,
Sorry for the delayed response. The code we use for the producer send call
looks like this:
We instantiate the producer like this:
ProducerEventType, EventType producer = new ProducerEventType,
EventType(config, context.getEventSerializer(), new
ProducerHandlerWrapper(config,
We are using the High Level Consumer API to interact with Kafka.
However, on restart in the case of failures, we want to be able to manually
reset offsets in certain situations.
What is the recommended way to do this?
Should we use the Simple Consumer API just for this restart case?
Ideally,
Hi list,
There were already 2 HTTP REST clients to Kafka (namely kafka-http [1] and
dropwizard-kakfa-http [2]).
I implemented another one in python with Flask (named flasfka):
https://github.com/travel-intelligence/flasfka
I tend to trust projects better when they come with tests and
After further investigation, I've figured out that the issue is caused by
the follower not processing messages from the controller until its
ReplicaFetcherThread has shutdown completely (which only happens when the
socket times out).
If the test waits for the socket to timeout, the logs show that
See https://cwiki.apache.org/confluence/display/KAFKA/System+Tools
and check the following:
GetOffsetShell (not very accurate - will set your offsets to much smaller
values than you really need; we log offsets frequently in application logs
and get it from there)
ImportZkOffsets
Kind regards,
You are also correct and perceptive to notice that if you check the end of
the log then begin consuming and read up to that point compaction may have
already kicked in (if the reading takes a while) and hence you might have
an incomplete snapshot.
Isn't it sufficient to just repeat the check
I am running a cluster of 5 brokers with 40G ms/mx for each. I found one of
the brokers is constantly using above ~90% of memory for jvm.heapUsage. I
checked from lsof output that the size of the index files for this broker
is too large.
Not sure what is going on with this one broker in the
Well, I guess what I was thinking is that since we have the long timeout on
the vote anyway, no reason not to call the vote now, should anything else
pop up we can cancel the vote.
-Jay
On Wed, Feb 18, 2015 at 4:04 PM, Jun Rao j...@confluent.io wrote:
Well, KAFKA-1952 only introduces high CPU
Hi,
My name is Deepak and I work for salesforce. We are using kafka 8.11 and
have a question about starting kafka with same broker id.
Steps:
Start a kakfa broker with broker id =1 - it starts fine with external ZK
Start another kafka with same broker id =1 .. it does not start the kafka
which
+1
On Feb 18, 2015 7:23 PM, Matt Narrell matt.narr...@gmail.com wrote:
+1
On Feb 18, 2015, at 7:56 PM, Jun Rao j...@confluent.io wrote:
This is the first candidate for release of Apache Kafka 0.8.2.1. This
only fixes one critical issue (KAFKA-1952) in 0.8.2.0.
Release Notes for the
Do you have to separate the snapshot from the normal update flow.
We are trying to avoid using another datasource if possible to have one
source of truth.
I think what you are saying is that you want to create a snapshot from the
Kafka topic but NOT do continual reads after that point. For
This is the first candidate for release of Apache Kafka 0.8.2.1. This
only fixes one critical issue (KAFKA-1952) in 0.8.2.0.
Release Notes for the 0.8.2.1 release
https://people.apache.org/~junrao/kafka-0.8.2.1-candidate1/RELEASE_NOTES.html
*** Please download, test and vote by Saturday, Feb 21,
40G is really huge, generally you would want more like 4G. Are you sure you
need that? Not sure what you mean by lsof and index files being too large,
but the index files are memory mapped so they should be able to grow
arbitrarily large and their memory usage is not counted in the java heap
(in
Well, KAFKA-1952 only introduces high CPU overhead if the number of
partitions in a fetch request is high, say more than a couple of hundreds.
So, it may not show up in every installation. For example, if you have 1000
leader replicas in a broker, but have a 20 node cluster, each replica fetch
Deepak,
You should getting following error and kafka server will shutdown
itself it it sees same brokerid registered in zookeeper.
Fatal error during KafkaServer startup. Prepare to shutdown
(kafka.server.KafkaServer)
java.lang.RuntimeException: A broker is already registered on the
2. Make the log end offset available more easily in the consumer.
Was thinking something would need to be added in LogCleanerManager, in the
updateCheckpoints function. Where would be best to publish the information
to make it more easily available, or would you just expose the
Why would you want to ever do that?
On Feb 18, 2015, at 15:16, Deepak Dhakal ddha...@salesforce.com wrote:
Hi,
My name is Deepak and I work for salesforce. We are using kafka 8.11 and
have a question about starting kafka with same broker id.
Steps:
Start a kakfa broker with broker
+1
On Feb 18, 2015, at 7:56 PM, Jun Rao j...@confluent.io wrote:
This is the first candidate for release of Apache Kafka 0.8.2.1. This
only fixes one critical issue (KAFKA-1952) in 0.8.2.0.
Release Notes for the 0.8.2.1 release
Yes, that makes sense. I will try to roll out 0.8.2.1 for vote later today.
Thanks,
Jun
On Wed, Feb 18, 2015 at 4:15 PM, Jay Kreps jay.kr...@gmail.com wrote:
Well, I guess what I was thinking is that since we have the long timeout on
the vote anyway, no reason not to call the vote now,
Yeah I was thinking either along the lines Joel was suggesting or else
adding a logEndOffset(TopicPartition) method or something like that. As
Joel says the consumer actually has this information internally (we return
it with the fetch request) but doesn't expose it.
-Jay
On Wed, Feb 18, 2015 at
31 matches
Mail list logo