Elastic Data Processing (EDP)

Overview

Sahara’s Elastic Data Processing facility or EDP allows the execution of jobs on clusters created from Sahara. EDP supports:

• Hive, Pig, MapReduce, MapReduce.Streaming and Java job types on Hadoop clusters
• Spark jobs on Spark standalone clusters
• storage of job binaries in Swift or Sahara’s own database
• access to input and output data sources in
  • HDFS for all job types
  • Swift for all types excluding Spark and Hive
• configuration of jobs at submission time
• execution of jobs on existing clusters or transient clusters

Interfaces

The EDP features can be used from the Sahara web UI which is described in the Sahara (Data Processing) UI User Guide.

The EDP features also can be used directly by a client through the Sahara REST API v1.1 (EDP).

EDP Concepts

Sahara EDP uses a collection of simple objects to define and execute jobs. These objects are stored in the Sahara database when they are created, allowing them to be reused. This modular approach with database persistence allows code and data to be reused across multiple jobs.

The essential components of a job are:

• executable code to run
• input data to process
• an output data location
• any additional configuration values needed for the job run

These components are supplied through the objects described below.

Job Binaries

A Job Binary object stores a URL to a single script or Jar file and any credentials needed to retrieve the file. The file itself may be stored in the Sahara internal database or in Swift.

Files in the Sahara database are stored as raw bytes in a Job Binary Internal object. This object’s sole purpose is to store a file for later retrieval. No extra credentials need to be supplied for files stored internally.

Sahara requires credentials (username and password) to access files stored in Swift unless Swift proxy users are configured as described in Sahara Advanced Configuration Guide. The Swift service must be running in the same OpenStack installation referenced by Sahara.

There is a configurable limit on the size of a single job binary that may be retrieved by Sahara. This limit is 5MB and may be set with the job_binary_max_KB setting in the sahara.conf configuration file.

Jobs

A Job object specifies the type of the job and lists all of the individual Job Binary objects that are required for execution. An individual Job Binary may be referenced by multiple Jobs. A Job object specifies a main binary and/or supporting libraries depending on its type:

Job type	Main binary	Libraries
Hive	required	optional
Pig	required	optional
MapReduce	not used	required
MapReduce.Streaming	not used	optional
Java	not used	required
Spark	required	optional

Data Sources

A Data Source object stores a URL which designates the location of input or output data and any credentials needed to access the location.

Sahara supports data sources in Swift. The Swift service must be running in the same OpenStack installation referenced by Sahara.

Sahara also supports data sources in HDFS. Any HDFS instance running on a Sahara cluster in the same OpenStack installation is accessible without manual configuration. Other instances of HDFS may be used as well provided that the URL is resolvable from the node executing the job.

Job Execution

Job objects must be launched or executed in order for them to run on the cluster. During job launch, a user specifies execution details including data sources, configuration values, and program arguments. The relevant details will vary by job type. The launch will create a Job Execution object in Sahara which is used to monitor and manage the job.

To execute Hadoop jobs, Sahara generates an Oozie workflow and submits it to the Oozie server running on the cluster. Familiarity with Oozie is not necessary for using Sahara but it may be beneficial to the user. A link to the Oozie web console can be found in the Sahara web UI in the cluster details.

For Spark jobs, Sahara uses the spark-submit shell script and executes the Spark job from the master node. Logs of spark jobs run by Sahara can be found on the master node under the /tmp/spark-edp directory.

General Workflow

The general workflow for defining and executing a job in Sahara is essentially the same whether using the web UI or the REST API.

1. Launch a cluster from Sahara if there is not one already available
2. Create all of the Job Binaries needed to run the job, stored in the Sahara database or in Swift
   • When using the REST API and internal storage of job binaries, there is an extra step here to first create the Job Binary Internal objects
   • Once the Job Binary Internal objects are created, Job Binary objects may be created which refer to them by URL
3. Create a Job object which references the Job Binaries created in step 2
4. Create an input Data Source which points to the data you wish to process
5. Create an output Data Source which points to the location for output data

(Steps 4 and 5 do not apply to Java or Spark job types. See Additional Details for Java jobs and Additional Details for Spark jobs)

6. Create a Job Execution object specifying the cluster and Job object plus relevant data sources, configuration values, and program arguments
   • When using the web UI this is done with the Launch On Existing Cluster or Launch on New Cluster buttons on the Jobs tab
   • When using the REST API this is done via the /jobs/<job_id>/execute method

The workflow is simpler when using existing objects. For example, to construct a new job which uses existing binaries and input data a user may only need to perform steps 3, 5, and 6 above. Of course, to repeat the same job multiple times a user would need only step 6.

Specifying Configuration Values, Parameters, and Arguments

Jobs can be configured at launch. The job type determines the kinds of values that may be set:

Job type	Configration Values	Parameters	Arguments
Hive	Yes	Yes	No
Pig	Yes	Yes	Yes
MapReduce	Yes	No	No
MapReduce.Streaming	Yes	No	No
Java	Yes	No	Yes
Spark	Yes	No	Yes

• Configuration values are key/value pairs.
  • The EDP configuration values have names beginning with edp. and are consumed by Sahara
  • Other configuration values may be read at runtime by Hadoop jobs
  • Currently additional configuration values are not available to Spark jobs at runtime
• Parameters are key/value pairs. They supply values for the Hive and Pig parameter substitution mechanisms.
• Arguments are strings passed as command line arguments to a shell or main program

These values can be set on the Configure tab during job launch through the web UI or through the job_configs parameter when using the /jobs/<job_id>/execute REST method.

In some cases Sahara generates configuration values or parameters automatically. Values set explicitly by the user during launch will override those generated by Sahara.

Generation of Swift Properties for Data Sources

If Swift proxy users are not configured (see Sahara Advanced Configuration Guide) and a job is run with data sources in Swift, Sahara will automatically generate Swift username and password configuration values based on the credentials in the data sources. If the input and output data sources are both in Swift, it is expected that they specify the same credentials.

The Swift credentials can be set explicitly with the following configuration values:

Name
fs.swift.service.sahara.username
fs.swift.service.sahara.password

Additional Details for Hive jobs

Sahara will automatically generate values for the INPUT and OUTPUT parameters required by Hive based on the specified data sources.

Additional Details for Pig jobs

Sahara will automatically generate values for the INPUT and OUTPUT parameters required by Pig based on the specified data sources.

For Pig jobs, arguments should be thought of as command line arguments separated by spaces and passed to the pig shell.

Parameters are a shorthand and are actually translated to the arguments -param name=value

Additional Details for MapReduce jobs

Important!

If the job type is MapReduce, the mapper and reducer classes must be specified as configuration values. Note, the UI will not prompt the user for these required values, they must be added manually with the Configure tab. Make sure to add these values with the correct names:

Name	Example Value
mapred.mapper.class	org.apache.oozie.example.SampleMapper
mapred.reducer.class	org.apache.oozie.example.SampleReducer

Additional Details for MapReduce.Streaming jobs

Important!

If the job type is MapReduce.Streaming, the streaming mapper and reducer classes must be specified.

In this case, the UI will prompt the user to enter mapper and reducer values on the form and will take care of adding them to the job configuration with the appropriate names. If using the python client, however, be certain to add these values to the job configuration manually with the correct names:

Name	Example Value
edp.streaming.mapper	/bin/cat
edp.streaming.reducer	/usr/bin/wc

Additional Details for Java jobs

Java jobs use two special configuration values:

• edp.java.main_class (required) Specifies the class containing main(String[] args)
• edp.java.java_opts (optional) Specifies configuration values for the JVM

A Java job will execute the main(String[] args) method of the specified main class. There are two methods of passing values to the main method:

• Passing values as arguments

  Arguments set during job launch will be passed in the String[] args array.

• Setting configuration values

  Any configuration values that are set can be read from a special file created by Oozie.

Data Source objects are not used with Java job types. Instead, any input or output paths must be passed to the main method using one of the above two methods. Furthermore, if Swift data sources are used the configuration values listed in Generation of Swift Properties for Data Sources must be passed with one of the above two methods and set in the configuration by main.

The edp-wordcount example bundled with Sahara shows how to use configuration values, arguments, and Swift data paths in a Java job type.

Additional Details for Spark jobs

Spark jobs use a special configuration value:

• edp.java.main_class (required) Specifies the class containing the Java or Scala main method:
  • main(String[] args) for Java
  • main(args: Array[String] for Scala

A Spark job will execute the main method of the specified main class. Values may be passed to the main method through the args array. Any arguments set during job launch will be passed to the program as commandline arguments by spark-submit.

Data Source objects are not used with Spark job types. Instead, any input or output paths must be passed to the main method as arguments. Remember that Swift paths are not supported for Spark jobs currently.

The edp-spark example bundled with Sahara contains a Spark program for estimating Pi.

Special Sahara URLs

Sahara uses custom URLs to refer to objects stored in Swift or the Sahara internal database. These URLs are not meant to be used outside of Sahara.

Sahara Swift URLs passed to running jobs as input or output sources include a ”.sahara” suffix on the container, for example:

swift://container.sahara/object

You may notice these Swift URLs in job logs, however, you do not need to add the suffix to the containers yourself. Sahara will add the suffix if necessary, so when using the UI or the python client you may write the above URL simply as:

swift://container/object

Sahara internal database URLs have the form:

internal-db://sahara-generated-uuid

This indicates a file object in the Sahara database which has the given uuid as a key

EDP Requirements

The OpenStack installation and the cluster launched from Sahara must meet the following minimum requirements in order for EDP to function:

OpenStack Services

When a Hadoop job is executed, binaries are first uploaded to a cluster node and then moved from the node local filesystem to HDFS. Therefore, there must be an instance of HDFS available to the nodes in the Sahara cluster.

If the Swift service is not running in the OpenStack installation

• Job binaries may only be stored in the Sahara internal database
• Data sources require a long-running HDFS

If the Swift service is running in the OpenStack installation

• Job binaries may be stored in Swift or the Sahara internal database
• Data sources may be in Swift or a long-running HDFS

Cluster Processes

Requirements for EDP support depend on the EDP job type and plugin used for the cluster. For example a Vanilla Sahara cluster must run at least one instance of these processes to support EDP:

• For Hadoop version 1:
  • jobtracker
  • namenode
  • oozie
  • tasktracker
  • datanode
• For Hadoop version 2:
  • namenode
  • datanode
  • resourcemanager
  • nodemanager
  • historyserver
  • oozie

EDP Technical Considerations

There are a several things in EDP which require attention in order to work properly. They are listed on this page.

Transient Clusters

EDP allows running jobs on transient clusters. In this case the cluster is created specifically for the job and is shut down automatically once the job is finished.

Two config parameters control the behaviour of periodic clusters:

• periodic_enable - if set to ‘False’, Sahara will do nothing to a transient cluster once the job it was created for is completed. If it is set to ‘True’, then the behaviour depends on the value of the next parameter.
• use_identity_api_v3 - set it to ‘False’ if your OpenStack installation does not provide Keystone API v3. In that case Sahara will not terminate unneeded clusters. Instead it will set their state to ‘AwaitingTermination’ meaning that they could be manually deleted by a user. If the parameter is set to ‘True’, Sahara will itself terminate the cluster. The limitation is caused by lack of ‘trusts’ feature in Keystone API older than v3.

If both parameters are set to ‘True’, Sahara works with transient clusters in the following manner:

1. When a user requests for a job to be executed on a transient cluster, Sahara creates such a cluster.
2. Sahara drops the user’s credentials once the cluster is created but prior to that it creates a trust allowing it to operate with the cluster instances in the future without user credentials.
3. Once a cluster is not needed, Sahara terminates its instances using the stored trust. Sahara drops the trust after that.