apache core batch processing #43

  • 6 machines, 6 units

Overview

The Apache Hadoop software library is a framework that allows for the
distributed processing of large data sets across clusters of computers
using a simple programming model.

It is designed to scale up from single servers to thousands of machines,
each offering local computation and storage. Rather than rely on hardware
to deliver high-avaiability, the library itself is designed to detect
and handle failures at the application layer, so delivering a
highly-availabile service on top of a cluster of computers, each of
which may be prone to failures.

This bundle provides a complete deployment of the core components of the
Apache Hadoop 2.7.1
platform to perform distributed data analytics at scale. These components
include:

  • NameNode (HDFS)
  • ResourceManager (Yarn)
  • Slaves (DataNode and NodeManager)
  • Client (example and node for manually running jobs from)
    • Plugin (colocated on the Client)

Deploying this bundle gives you a fully configured and connected Apache Hadoop
cluster on any supported cloud, which can be easily scaled to meet workload
demands.

Deploying this bundle

In this deployment, the aforementioned components are deployed on separate
units. To deploy this bundle, simply use:

juju quickstart u/bigdata-dev/apache-core-batch-processing

See juju quickstart --help for deployment options, including machine
constraints and how to deploy a locally modified version of the
apache-core-batch-processing bundle.yaml.

The default bundle deploys three slave nodes and one node of each of
the other services. To scale the cluster, use:

juju add-unit slave -n 2

This will add two additional slave nodes, for a total of five.

Verify the deployment

The services provide extended status reporting to indicate when they are 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 charms for each master component (namenode, resourcemanager)
also each provide a smoke-test action that can be used to verify that each
component is functioning as expected. You can run them all and then watch the
action status list:

juju action do namenode/0 smoke-test
juju action do resourcemanager/0 smoke-test
watch -n 0.5 juju action status

Eventually, all of the actions should settle to status: completed. If
any go instead to status: failed then it means that component is not working
as expected. You can get more information about that component's smoke test:

juju action fetch <action-id>

Deploying in Network-Restricted Environments

The Apache Hadoop charms 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 these charms.

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, the Apache Hadoop charms require 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 a 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 that charm.
Setting this option will cause all resources required by the charm to be
downloaded from the configured URL.

You can fetch the resources for all of the Apache Hadoop charms
(apache-hadoop-namenode, apache-hadoop-resourcemanager,
apache-hadoop-slave, apache-hadoop-plugin, etc) into a single
directory and serve them all with a single juju-resources serve instance.

Contact Information

Resources