Apache Spark 4.0 End of Life: November 23, 2026 (New 6-Month Policy)
Apache Spark 4.0 reaches end of life on November 23, 2026, eighteen months after its release on May 19, 2025. It is the next Spark line to lose maintenance. It is also one of the last to get eighteen months: the project's versioning policy now says that, starting with Spark 4.3, releases ship every three months and most are maintained for six months only.
Spark lines and their dates
| Line | Released | End of life | Which rule applies |
|---|---|---|---|
| Spark 4.2 | July 14, 2026 | January 11, 2028 | 18 months. The policy says 4.2.x is not affected by the change |
| Spark 4.1 | December 11, 2025 | June 11, 2027 | 18 months |
| Spark 4.0 | May 19, 2025 | November 23, 2026 | 18 months |
| Spark 3.5 | September 9, 2023 | November 30, 2027 | Extended LTS: security fixes only, to November 2027 |
| Spark 3.4 | April 7, 2023 | October 21, 2024 | Ended |
The Spark project states the 3.5 extended period by month; we show the last day of it.
The new rules, from Spark 4.3 onward
| Release type | Maintenance window in the policy |
|---|---|
| Major (x.0.0), once a year | 6 months |
| Feature (x.1.0, x.2.0), quarterly | 6 months |
| LTS: the last feature release of each major line | 18 months |
- For the 4.x line the policy names Spark 4.5 as the LTS release, as an exception during the transition.
- The policy's illustration for the following year is Spark 5.0 as the annual major, then 5.1, 5.2 and 5.3, with 5.3 as the 5.x LTS.
- Critical security patches are backported to all actively maintained branches.
- The 3.5 extended period covers the main Apache Spark project only, and security fixes only. The policy also warns that a dependency fix that needs a newer Java than 8 may be marked as not to be fixed on 3.5.
What to do
- On Spark 3.4 or older: out of maintenance. Spark has an entry on CISA's Known Exploited Vulnerabilities list (a command injection flaw added in March 2023), so unmaintained clusters are not a theoretical risk.
- On Spark 3.5: security fixes continue to November 2027. That is the project's migration window to 4.x; use it rather than waiting it out.
- On Spark 4.0: move to 4.1 or 4.2 before November 23, 2026. Both keep the eighteen-month window; 4.2 runs to January 11, 2028.
- Planning past 4.2: decide now whether you will follow the quarterly releases or sit on LTS releases (4.5, then 5.3 in the policy's illustration). A six-month window does not fit an annual upgrade habit.
- Using a vendor's Spark (Databricks, EMR, Dataproc, Cloudera and others): the vendor's runtime has its own support dates, which are the ones that apply to you.
Frequently Asked Questions
When is Apache Spark 4.0 end of life?
November 23, 2026, the end of the 18-month maintenance window that applies to Spark 4.0. Spark 4.1 follows on June 11, 2027 and Spark 4.2 on January 11, 2028.
How long is an Apache Spark release maintained?
Releases up to Spark 4.2 keep an 18-month maintenance window. Starting with Spark 4.3, the project ships a release every three months and maintains major and feature releases for six months each; the last release of each major line is the LTS release and is maintained for 18 months. For the 4.x line the project says the LTS will be Spark 4.5.
Is Apache Spark 3.5 still supported?
Yes, with security fixes only. The Spark project gave 3.5 an extended LTS period that ends in November 2027, to allow migrations to 4.x to be completed.
Related
- Apache Spark — every line, live status and EOL Risk Score
- The 2026 end-of-life calendar
- How we verify our dates