apache kafka #24

Supports: trusty
Add to new model


A single Kafka broker can handle hundreds of megabytes of reads and writes per
second from thousands of clients.

Kafka is designed to allow a single cluster to serve as the central data
backbone for a large organization. It can be elastically and transparently
expanded without downtime. Data streams are partitioned and spread over a
cluster of machines to allow data streams larger than the capability of any
single machine and to allow clusters of co-ordinated consumers.

Messages are persisted on disk and replicated within the cluster to prevent
data loss. Each broker can handle terabytes of messages without performance

Distributed by Design
Kafka has a modern cluster-centric design that offers strong durability and
fault-tolerance guarantees.


Apache Kafka is an open-source message broker project developed by the Apache
Software Foundation written in Scala. The project aims to provide a unified,
high-throughput, low-latency platform for handling real-time data feeds. Learn
more at kafka.apache.org.


Kafka requires the Zookeeper distributed coordination service. Deploy and
relate them as follows:

juju deploy apache-zookeeper zookeeper
juju deploy apache-kafka kafka
juju add-relation kafka zookeeper

Once deployed, we can list the zookeeper servers that our kafka brokers
are connected to. The following will list <ip>:<port> information for each
zookeeper unit in the environment (e.g.:

juju action do kafka/0 list-zks
juju action fetch <id>  # <-- id from above command

We can create a Kafka topic with:

juju action do kafka/0 create-topic topic=<topic_name> \
 partitions=<#> replication=<#>
juju action fetch <id>  # <-- id from above command

We can list topics with:

juju action do kafka/0 list-topics
juju action fetch <id>  # <-- id from above command

We can write to a topic with:

juju action do kafka/0 write-topic topic=<topic_name> data=<data>
juju action fetch <id>  # <-- id from above command

We can read from a topic with:

juju action do kafka/0 read-topic topic=<topic_name> partition=<#>
juju action fetch <id>  # <-- id from above command

And finally, we can delete a topic with:

juju action do kafka/0 delete-topic topic=<topic_name>
juju action fetch <id>  # <-- id from above command

Status and Smoke Test

Kafka provides extended status reporting to indicate when it is ready:

juju status --format=tabular

This is particularly useful when combined with watch to track the on-going
progress of the deployment:

watch -n 0.5 juju status --format=tabular

The message for each unit will provide information about that unit's state.
Once they all indicate that they are ready, you can perform a "smoke test"
to verify that Kafka is working as expected using the built-in smoke-test

juju action do kafka/0 smoke-test

After a few seconds or so, you can check the results of the smoke test:

juju action status

You will see status: completed if the smoke test was successful, or
status: failed if it was not. You can get more information on why it failed

juju action fetch <action-id>


Creating a cluster with many brokers is as easy as adding more units to your
Kafka service:

juju add-unit kafka

After adding additional brokers, you will be able to create topics with
replication up to the number of kafka units.

To verify replication is working you can do the following:

juju add-unit kafka -n 2
juju action do kafka/0 create-topic topic=my-replicated-topic \
    partitions=1 replication=2

Query for the description of the just created topic:

juju ssh kafka/0
kafka-topics.sh --describe --topic my-replicated-topic \
    --zookeeper <zookeeperip>:2181

You should get a response similar to:

Topic: my-replicated-topic PartitionCount:1 ReplicationFactor:2 Configs:
Topic: my-replicated-topic Partition: 0 Leader: 2 Replicas: 2,0 Isr: 2,0

Connecting External Clients

By default, this charm does not expose Kafka outside of the provider's network.
To allow external clients to connect to Kafka, first expose the service:

juju expose kafka

Next, ensure the external client can resolve the short hostname of the kafka
unit. A simple way to do this is to add an /etc/hosts entry on the external
kafka client machine. Gather the needed info from juju:

user@juju-client$ juju run --unit kafka/0 'hostname -s'
user@juju-client$ juju status --format=yaml kafka/0 | grep public-address
public-address: 40.784.149.135

Update /etc/hosts on the external kafka client:

user@kafka-client$ echo "40.784.149.135 kafka-0" | sudo tee -a /etc/hosts

The external kafka client should now be able to access Kafka by using
kafka-0:9092 as the broker.

Deploying in Network-Restricted Environments

This charm can be deployed in environments with limited network access. To
deploy in this environment, you will need a local mirror to serve the packages
and resources required by this charm.

Mirroring Packages

You can setup a local mirror for apt packages using squid-deb-proxy.
For instructions on configuring juju to use this, see the
Juju Proxy Documentation.

Mirroring Resources

In addition to apt packages, this charm requires a few binary resources
which are normally hosted on Launchpad. If access to Launchpad is not
available, the jujuresources library makes it easy to create a mirror
of these resources:

sudo pip install jujuresources
juju-resources fetch --all /path/to/resources.yaml -d /tmp/resources
juju-resources serve -d /tmp/resources

This will fetch all of the resources needed by this charm and serve them via a
simple HTTP server. The output from juju-resources serve will give you a
URL that you can set as the resources_mirror config option for this charm.
Setting this option will cause all resources required by this charm to be
downloaded from the configured URL.

Contact Information



(string) Specify the hostname that kafka will set as its advertised hostname. Note that all clients connecting to kafka will use this hostname. We\'ll setup a good default hostname for you, based on your cloud provider\'s DNS settings. You should set this manually only if you know what you\'re doing!
(string) A string containing the name of a network interface, or a CIDR range. For split network environments, or for other secure environments, you may wish to bind to a specific network interface. You may either name the interface here, or specify a CIDR range that contains the IP of the network interface. The charm will translate that into a specific IP address to bind to, and drop that into the Kafka config. To reset the bindings, pass in
(string) URL used to fetch resources (e.g., Kafka binaries) instead of the location specified in resources.yaml.