spark #56

Description

Apache Spark is a fast and general engine for large-scale data processing.
This charm provides version 2.1.0 of the Spark application from Apache Bigtop.

Overview

Apache Spark is a fast and general purpose engine for large-scale data
processing. Learn more at spark.apache.org.

This charm deploys version 2.1.0 of the Spark component from Apache Bigtop.

Deploying

This charm requires Juju 2.0 or greater. If Juju is not yet set up, please
follow the getting-started instructions prior to deploying this charm.

This charm supports running Spark in a variety of modes:

Standalone

In this mode, Spark units form a cluster that can be scaled as needed.
Starting with a single node:

juju deploy spark

Scale the cluster by adding more spark units:

juju add-unit spark

When in standalone mode, Juju ensures a single Spark master is appointed.
The status of the unit acting as master reads ready (standalone - master),
while the rest of the units display a status of ready (standalone).
If the master is removed, Juju will appoint a new one. However, if a master
fails in standalone mode, running jobs and job history will be lost.

Standalone HA

To enable High Availability for a Spark cluster, simply add Zookeeper to
the deployment. To ensure a Zookeeper quorum, 3 units of the zookeeper
application are recommended. For instance:

juju deploy zookeeper -n 3
juju add-relation spark zookeeper

In this mode, the cluster can again be scaled as needed by adding/removing
units. Spark units report ready (standalone HA) in their status. To identify
the unit acting as master, query Zookeeper as follows:

juju run --unit zookeeper/0 'echo "get /spark/master_status" | /usr/lib/zookeeper/bin/zkCli.sh'

YARN

This charm leverages our pluggable Hadoop model with the hadoop-plugin
interface. This means that this charm can be related to an Apache Hadoop
cluster to run Spark jobs there. The suggested deployment method is to use the
hadoop-spark bundle:

juju deploy hadoop-spark

To switch among the above execution modes, set the spark_execution_mode
config variable:

juju config spark spark_execution_mode=<new_mode>

See the Configuring section below for supported mode options.

Network-Restricted Environments

Charms can be deployed in environments with limited network access. To deploy
in this environment, configure a Juju model with appropriate proxy and/or
mirror options. See Configuring Models for more information.

Verifying

Status

Apache Bigtop charms provide extended status reporting to indicate when they
are ready:

juju status

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

watch -n 2 juju status

The message column will provide information about a given unit's state.
This charm is ready for use once the status message indicates that it is
ready.

Smoke Test

This charm provides a smoke-test action that can be used to verify the
application is functioning as expected. Run the action as follows:

juju run-action spark/0 smoke-test

Watch the progress of the smoke test actions with:

watch -n 2 juju show-action-status

Eventually, the action should settle to status: completed. If it
reports status: failed, the application is not working as expected. Get
more information about a specific smoke test with:

juju show-action-output <action-id>

Spark Master web UI

Spark provides a web console that can be used to verify information about
the cluster. To access it, find the Public address of the spark application
and expose it:

juju status spark
juju expose spark

The web interface will be available at the following URL:

http://SPARK_PUBLIC_IP:8080

Spark Job History web UI

The Job History server shows all active and finished spark jobs submitted.
As mentioned above, expose the spark application and note the public IP
address. The job history web interface will be available at the following URL:

http://SPARK_PUBLIC_IP:18080

Using

Actions

Once Spark is ready, there are a number of actions available in this charm.

Run a benchmark (as described in the Benchmarking section):

juju run-action spark/0 pagerank
juju show-action-output <id>  # <-- id from above command

Run a smoke test (as described in the Verifying section):

juju run-action spark/0 smoke-test
juju show-action-output <id>  # <-- id from above command

Start/Stop/Restart the Spark Job History service:

juju run-action spark/0 [start|stop|restart]-spark-job-history-server
juju show-action-output <id>  # <-- id from above command

Submit a Spark job:

juju run-action spark/0 spark-submit \
  options='--class org.apache.spark.examples.SparkPi' \
  job='/usr/lib/spark/examples/jars/spark-examples.jar' \
  job-args='10'
juju show-action-output <id>  # <-- id from above command

Spark shell

Spark shell provides a simple way to learn the API, as well as a powerful
tool to analyze data interactively. It is available in either Scala or Python
and can be run from the Spark unit as follows:

juju ssh spark/0
spark-shell # for interaction using scala
pyspark     # for interaction using python

Command line

SSH to the Spark unit and manually run a spark-submit job, for example:

juju ssh spark/0
spark-submit --class org.apache.spark.examples.SparkPi \
 /usr/lib/spark/examples/jars/spark-examples.jar 10

Apache Zeppelin

Apache Zeppelin is a web-based notebook that enables interactive data
analytics. Make beautiful data-driven, interactive, and collaborative documents
with SQL, Scala and more. Deploy Zeppelin and relate it to Spark:

juju deploy zeppelin
juju add-relation spark zeppelin

To access the web console, find the Public address of the zeppelin
application and expose it:

juju status zeppelin
juju expose zeppelin

The web interface will be available at the following URL:

http://ZEPPELIN_PUBLIC_IP:9080

Configuring

Charm configuration can be changed at runtime with juju config. This charm
supports the following config parameters.

driver_memory

Amount of memory available for the Spark driver process (1g by default).
Set a different value with:

juju config spark driver_memory=4096m

executor_memory

Amount of memory available for each Spark executor process (1g by default).
Set a different value with:

juju config spark executor_memory=2g

Note: When Spark is in YARN mode, ensure the configured executor memory
does not exceed the NodeManager maximum (defined on each nodemanager as
yarn.nodemanager.resource.memory-mb in yarn-default.xml).

install-cuda

Provided by layer-nvidia-cuda, this option controls the installation
of NVIDIA CUDA packages if capable GPU hardware is present. When false (the
default), CUDA will not be installed or configured regardless of hardware
support. Set this to true to fetch and install CUDA-related packages from
the NVIDIA developer repository.

juju config spark install-cuda=true

Note: This option requires external network access to
http://developer.download.nvidia.com/. Ensure appropriate proxies are
configured if needed.

spark_bench_enabled

Controls the installation of the Spark-Bench benchmarking suite. When set
to true, this charm will download and install Spark-Bench from the URL
specified by the spark_bench_url config option. When set to false
(the default), Spark-Bench will not be installed on the unit, though any data
stored in hdfs:///user/ubuntu/spark-bench from previous installations will
be preserved.

Note: Spark-Bench has not been verified to work with Spark 2.1.x.

Note: This option requires external network access to the configured
Spark-Bench URL. Ensure appropriate proxies are configured if needed.

spark_execution_mode

Spark has four modes of execution: local, standalone, yarn-client, and
yarn-cluster. The default mode is standalone and can be changed by setting
the spark_execution_mode config option.

  • Local

    In Local mode, Spark processes jobs locally without any cluster resources.
    There are 3 ways to specify 'local' mode:

    • local

    Run Spark locally with one worker thread (i.e. no parallelism at all).

    • local[K]

    Run Spark locally with K worker threads (ideally, set this to the number
    of cores on the deployed machine).

    • local[*]

    Run Spark locally with as many worker threads as logical cores on the
    deployed machine.

  • Standalone

    In standalone mode, Spark launches a Master and Worker daemon on the Spark
    unit. This mode is useful for simulating a distributed cluster environment
    without actually setting up a cluster.

  • YARN-client

    In yarn-client mode, the Spark driver runs in the client process, and the
    application master is only used for requesting resources from YARN.

  • YARN-cluster

    In yarn-cluster mode, the Spark driver runs inside an application master
    process which is managed by YARN on the cluster, and the client can go away
    after initiating the application.

Benchmarking

This charm provides benchmarks to gauge the performance of the Spark cluster.
Each benchmark is an action that can be run with juju run-action:

$ juju actions spark
...
pagerank                          Calculate PageRank for a sample data set
sparkpi                           Calculate Pi
...

$ juju run-action spark/0 pagerank
Action queued with id: 339cec1f-e903-4ee7-85ca-876fb0c3d28e

$ juju show-action-output 339cec1f-e903-4ee7-85ca-876fb0c3d28e
results:
  meta:
    composite:
      direction: asc
      units: secs
      value: "83"
    start: 2017-04-12T23:22:38Z
    stop: 2017-04-12T23:24:01Z
  output: '{''status'': ''completed''}'
status: completed
timing:
  completed: 2017-04-12 23:24:02 +0000 UTC
  enqueued: 2017-04-12 23:22:36 +0000 UTC
  started: 2017-04-12 23:22:37 +0000 UTC

Issues

Apache Bigtop tracks issues using JIRA (Apache account required). File an
issue for this charm at:

https://issues.apache.org/jira/secure/CreateIssue!default.jspa

Ensure Bigtop is selected as the project. Typically, charm issues are filed
in the deployment component with the latest stable release selected as the
affected version. Any uncertain fields may be left blank.

Contact Information

Resources

Configuration

spark_bench_url
(string) URL (including hash) of a Spark-Bench tarball. By default, this points to a pre-built Spark-Bench binary based on sources in the upstream repository. This option is only valid when 'spark_bench_enabled' is 'true'.
https://s3.amazonaws.com/jujubigdata/ibm/noarch/SparkBench-2.0-20170403.tgz#sha256=709caec6667dd82e42de25eb8bcd5763ca894e99e5c83c97bdfcf62cb1aa00c8
bigtop_version
(string) Apache Bigtop release version. The default, '1.2.0' will use the current GA release, Bigtop 1.2.0, for all hiera data, puppet recipes, and installable packages. Set this to 'master' to use the latest upstream bits.
1.2.0
spark_execution_mode
(string) Options are "local", "standalone", "yarn-client", and "yarn-cluster". Consult the readme for details on these options.
standalone
driver_memory
(string) Specify gigabytes (e.g. 1g) or megabytes (e.g. 1024m). If running in 'local' or 'standalone' mode, you may also specify a percentage of total system memory (e.g. 50%).
1g
spark_bench_enabled
(boolean) When set to 'true', this charm will download and install the Spark-Bench benchmark suite from the configured URLs. When set to 'false' (the default), Spark-Bench will not be installed on the unit, though any data stored in hdfs:///user/ubuntu/spark-bench from previous installations will be preserved. Note that Spark-Bench has not been verified to work with Spark 2.1.x.
executor_memory
(string) Specify gigabytes (e.g. 1g) or megabytes (e.g. 1024m). If running in 'local' or 'standalone' mode, you may also specify a percentage of total system memory (e.g. 50%).
1g
install-cuda
(boolean) Install the CUDA binaries if capable hardware is present. This option is False by default, as not all applications support GPU acceleration.
cuda-version
(string) The cuda-repo package version to install.
8.0.61-1