Spark 3
You can run multiple versions of Apache Spark3 simultaneously within a single cluster.
This groundbreaking feature enables organizations to maintain backward compatibility, support diverse application requirements, and manage smooth version transitions—all within the same data platform.
The available Spark 3 versions include:
- Apache Spark 3.5.5 — ⭐ Recommended — Production workloads, new applications
- Apache Spark 3.5.1 — Supported — Compatibility with 3.5.x applications
- Apache Spark 3.3.3 — Supported — Legacy applications, backward compatibility
Info
** Spark 3 offers full compatibility with Java and Scala. However, Python 2 modules are not supported in Spark 3 version 3.1.1 and later.
For details, access the following pages.
- Key Benefits
- Spark 3 Use Cases
- Install Multiple Spark 3 Versions
- Using Multiple Spark Versions
- Manage Spark 3 Configurations
- Troubleshooting Spark 3
- Spark 3 Installation Conclusion
- Spark 3 Appendix: Quick Reference