Databricks runtime support lifecycle

Databricks releases runtimes as Beta and GA versions. Databricks supports GA versions for six months, unless the runtime version is:

  • A Long Term Support (LTS) version. LTS versions are released every six months and supported for two full years.

  • A Databricks Light runtime. Light versions are supported until either 12 months after release or two months after the next Databricks Light release, whichever comes first.

Workloads on unsupported runtime versions may continue to run, but they receive no Databricks support or fixes.

For details, see the tables that follow.

Support lifecycle for Databricks Runtime versions

Phase

Guarantees

Beta

Support SLAs are not applicable. For more information, see Databricks Runtime preview releases.

Full Support

Major stability and security fixes are backported.

Full support for Databricks Runtime versions lasts for six months, with the exception of Long Term Support (LTS) versions, which Databricks releases every six months and supports for two years.

End of support (EOS)

Version is unsupported:

  • Workloads running on these versions receive no Databricks support

  • Databricks will not backport fixes

Supported releases along with their end-of-support dates are published at Supported releases.

End of life (EOL)

Databricks reserves the right to completely remove a release version from the API at any time after support ends, without prior notice.

Support lifecycle for Databricks Light versions

Phase

Guarantees

Beta

Support SLAs are not applicable. For more information, see Databricks Runtime preview releases.

Full Support

Major stability and security fixes are backported.

Full support lasts until either 12 months after release or two months after the next Databricks Light release, whichever comes first.

End of support (EOS)

Version is unsupported:

  • Workloads running on these versions receive no Databricks support

  • Databricks will not backport fixes

The end-of-support date is either 12 months after release or two months after the next Databricks Light release, whichever comes first.

End of life (EOL)

Databricks reserves the right to completely remove a release version from the API at any time after support ends, without prior notice.

Note

Products and services provided by third parties, including those listed in the Databricks integrations, are not covered by this policy.

For information about how to construct the Databricks runtime version string for REST API calls, see Runtime version strings.