Spark Standalone Mode

In addition to running on top of Mesos, Spark also supports a standalone mode, consisting of one Spark master and several Spark worker processes. You can run the Spark standalone mode either locally (for testing) or on a cluster. If you wish to run on a cluster, we have provided a set of deploy scripts to launch a whole cluster.

Getting Started

Compile Spark with sbt package as described in the Getting Started Guide. You do not need to install Mesos on your machine if you are using the standalone mode.

Starting a Cluster Manually

You can start a standalone master server by executing:

./run spark.deploy.master.Master

Once started, the master will print out a spark://IP:PORT URL for itself, which you can use to connect workers to it, or pass as the “master” argument to SparkContext to connect a job to the cluster. You can also find this URL on the master’s web UI, which is http://localhost:8080 by default.

Similarly, you can start one or more workers and connect them to the master via:

./run spark.deploy.worker.Worker spark://IP:PORT

Once you have started a worker, look at the master’s web UI (http://localhost:8080 by default). You should see the new node listed there, along with its number of CPUs and memory (minus one gigabyte left for the OS).

Finally, the following configuration options can be passed to the master and worker:

-i IP, --ip IP IP address or DNS name to listen on
-p PORT, --port PORT IP address or DNS name to listen on (default: 7077 for master, random for worker)
--webui-port PORT Port for web UI (default: 8080 for master, 8081 for worker)
-c CORES, --cores CORES Number of CPU cores to use (default: all available); only on worker
-m MEM, --memory MEM Amount of memory to use, in a format like 1000M or 2G (default: your machine's total RAM minus 1 GB); only on worker
-d DIR, --work-dir DIR Directory to use for scratch space and job output logs (default: SPARK_HOME/work); only on worker

Cluster Launch Scripts

To launch a Spark standalone cluster with the deploy scripts, you need to set up two files, conf/ and conf/slaves. The conf/ file lets you specify global settings for the master and slave instances, such as memory, or port numbers to bind to, while conf/slaves is a list of slave nodes. The system requires that all the slave machines have the same configuration files, so copy these files to each machine.

In conf/, you can set the following parameters, in addition to the standard Spark configuration settongs:

Environment VariableMeaning
SPARK_MASTER_IP Bind the master to a specific IP address, for example a public one
SPARK_MASTER_PORT Start the master on a different port (default: 7077)
SPARK_MASTER_WEBUI_PORT Port for the master web UI (default: 8080)
SPARK_WORKER_PORT Start the Spark worker on a specific port (default: random)
SPARK_WORKER_CORES Number of cores to use (default: all available cores)
SPARK_WORKER_MEMORY How much memory to use, e.g. 1000M, 2G (default: total memory minus 1 GB)
SPARK_WORKER_WEBUI_PORT Port for the worker web UI (default: 8081)
SPARK_WORKER_DIR Directory to run jobs in, which will include both logs and scratch space (default: SPARK_HOME/work)

In conf/slaves, include a list of all machines where you would like to start a Spark worker, one per line. The master machine must be able to access each of the slave machines via password-less ssh (using a private key). For testing purposes, you can have a single localhost entry in the slaves file.

Once you’ve set up these configuration files, you can launch or stop your cluster with the following shell scripts, based on Hadoop’s deploy scripts, and available in SPARK_HOME/bin:

Note that the scripts must be executed on the machine you want to run the Spark master on, not your local machine.

Connecting a Job to the Cluster

To run a job on the Spark cluster, simply pass the spark://IP:PORT URL of the master as to the SparkContext constructor.

To run an interactive Spark shell against the cluster, run the following command:

MASTER=spark://IP:PORT ./spark-shell

Job Scheduling

The standalone cluster mode currently only supports a simple FIFO scheduler across jobs. However, to allow multiple concurrent jobs, you can control the maximum number of resources each Spark job will acquire. By default, it will acquire all the cores in the cluster, which only makes sense if you run just a single job at a time. You can cap the number of cores using System.setProperty("spark.cores.max", "10") (for example). This value must be set before initializing your SparkContext.

Monitoring and Logging

Spark’s standalone mode offers a web-based user interface to monitor the cluster. The master and each worker has its own web UI that shows cluster and job statistics. By default you can access the web UI for the master at port 8080. The port can be changed either in the configuration file or via command-line options.

In addition, detailed log output for each job is also written to the work directory of each slave node (SPARK_HOME/work by default). You will see two files for each job, stdout and stderr, with all output it wrote to its console.

Running Alongside Hadoop

You can run Spark alongside your existing Hadoop cluster by just launching it as a separate service on the machines. To access Hadoop data from Spark, just use a hdfs:// URL (typically hdfs://<namenode>:9000/path, but you can find the right URL on your Hadoop Namenode’s web UI). Alternatively, you can set up a separate cluster for Spark, and still have it access HDFS over the network; this will be slower than disk-local access, but may not be a concern if you are still running in the same local area network (e.g. you place a few Spark machines on each rack that you have Hadoop on).