End of support for Spark version 2.0 in Analytics for Apache Spark

5 min read

By: Rohan Vaidyanathan

End of support for Spark version 2.0 in Analytics for Apache Spark

We’d like to inform you about the withdrawal of support for Apache Spark v2.0 in IBM Analytics for Apache Spark which will end on April 30, 2018. However, this service will continue to support Apache Spark versions under the Spark 2.1.x branch. As part of IBM’s effort to provide a current, secure and GDPR compliant cloud service, several security vulnerabilities are being identified and repaired. A number of them present in Spark v2.0 have been fixed in the Spark v2.1 branch. In addition, Spark 2.1 API is compatible with that of Spark 2.0 with minimal upgrade updates needed to move from version 2.0 to 2.1.x.

For a more complete list of changes added to Apache Spark, consult the release notes for the Spark release at (Spark 2.1). The support for Apache Spark v2.0 will be withdrawn on April 30, 2018. Here’s what you need to know:

  • End of Support Announce Date: February 27, 2018

  • End of Support Withdrawal Date: April 30, 2018

For the period after the End of Support Announce Date, through April 30, 2018 all new or existing instances of Analytics for Apache Spark will have the ability to submit and execute jobs using a Spark v2.0 kernel or driver.

As of April 30, 2018, new or existing kernels or jobs cannot be submitted to execute with Spark v2.0 driver.

  • Please remember to update your applications and notebooks to use Spark v2.1.2 APIs before the End of Support Withdrawal Date, i.e. April 30, 2018.

We strongly recommend you to move your applications to Spark v2.1.2 in order to get access to the latest APIs and capabilities of Apache Spark in order to stay compliant with the GDPR requirements. Also, you may follow these migration guidelines to upgrade.

As always, we’re here to help! Reach out to us in the support forums.

Be the first to hear about news, product updates, and innovation from IBM Cloud