Skip to main content

Connect Pyspark to remote clusters

Project description

Pypsark Gateway is a library to seamlessly connect to remote spark clusters.

Quick Start

Install the pysparkgateway package on both the remote Spark cluster you are connecting to and the local machine.

pip install pysparkgateway

Start the Pyspark Gateway server on the cluster.

pyspark-gateway start

Pyspark Gateway communicates over 3 ports, 25000, 25001, 25002. Currently the client only supports connecting to these ports on localhost so you’ll need to tunnel them.

ssh myuser@foo.bar.cluster.com -L 25000:localhost:25000 -L 25001:localhost:25001 -L 25002:localhost:25002

Now you’re ready to connect. The main thing to keep in mind is the Pyspark Gateway import needs to come before any other import. Pypsark Gateway needs to patch your local pyspark in order to function properly.

The way that your local Python connects to the remote cluster is via a custom py4j gateway. Pyspark Gateway will create and configure automatically, you just need to pass it into the SparkContext options.

Also to enable all pyspark functions to work, spark.io.encryption.enabled needs to be set to true.

# This import comes first!
from pyspark_gateway import PysparkGateway
pg = PysparkGateway()

from pyspark import SparkContext, SparkConf

conf = conf.set('spark.io.encryption.enabled', 'true')
sc = SparkContext(gateway=pg.gateway, conf=conf)

Now you have a working spark context connected to a remote cluster.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

PysparkGateway-0.0.16.tar.gz (9.3 kB view details)

Uploaded Source

File details

Details for the file PysparkGateway-0.0.16.tar.gz.

File metadata

  • Download URL: PysparkGateway-0.0.16.tar.gz
  • Upload date:
  • Size: 9.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/1.12.1 pkginfo/1.4.2 requests/2.22.0 setuptools/41.0.0 requests-toolbelt/0.8.0 tqdm/4.26.0 CPython/2.7.13

File hashes

Hashes for PysparkGateway-0.0.16.tar.gz
Algorithm Hash digest
SHA256 e84cd5fb8770043cb2a892bafb7aeff3eed3a95c194958c1405d476640994257
MD5 f823392aa8b4a82bbec600008d778f8f
BLAKE2b-256 1823f07e65c2ab0b1e77ee710a136ba8aac53780fc894e60812a207ed43e0cb9

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page