.. _version_6.3.2:

=============
Version 6.3.2
=============

Released on 2026-05-13.

.. NOTE::

    If you are upgrading a cluster, you must be running CrateDB 5.0.0 or higher
    before you upgrade to 6.3.2.

    We recommend that you upgrade to the latest 6.2 release before moving to
    6.3.2.

    A rolling upgrade from >= 6.2.0 to 6.3.2 is supported.
    Before upgrading, you should `back up your data`_.

.. WARNING::

    Tables that were created before CrateDB 5.x will not function with 6.x
    and must be recreated before moving to 6.x.x.

    You can recreate tables using ``COPY TO`` and ``COPY FROM`` or by
    `inserting the data into a new table`_.

.. _back up your data: https://cratedb.com/docs/crate/reference/en/latest/admin/snapshots.html
.. _inserting the data into a new table: https://cratedb.com/docs/crate/reference/en/latest/admin/system-information.html#tables-need-to-be-recreated

.. rubric:: Table of contents

.. contents::
   :local:


See the :ref:`version_6.3.0` release notes for a full list of changes in the 6.3
series.

Fixes
=====

- Fixed an issue that could cause queries against ``sys.allocations`` to fail if
  concurrently resizing the shards of a table.

- Fixed a performance regression introduced in 6.0.0 that could cause ``INSERT
  INTO <table> VALUES (...)`` bulk statements that hit many shards to be less
  aggressive than possible, taking longer than necessary.

  To mitigate the issue in earlier versions it is also possible to increase the
  size of the :ref:`write thread pool <thread_pool.<name>.queue_size>`.

- Fixed an issue that caused the ``statement_timeout`` setting to expire a query
  if a PostgreSQL client was idling between creating a prepared statement
  (sending a ``parse`` message) and using the prepared statement (sending
  ``bind`` or ``execute`` messages).

- Fixed an issue that could lead to data loss for due to corruption of the
  metadata for tables created before :ref:`version_5.5.0`, when the cluster is
  upgraded to versions :ref:`version_6.0.0` or newer, and certain actions take
  place on these tables, e.g.: ``ALTER TABLE SET <config_param>=value``,
  dropping a partition, etc. The corruption will cause all columns to return
  ``NULL`` values, for all existing data in the table, whereas new data (via
  ``INSERT`` or ``UPDATE``), can be retrieved normally.

- Fixed :ref:`INSERT FROM VALUES <sql-insert>` statements that create new
  partitions are no longer interrupted by :ref:`SWAP <alter_cluster_swap_table>`
  or :ref:`RENAME <sql-alter-table-rename-to>` table statements.

- Fixed an issue that lead to a ``NullPointerException`` error on selects from
  the :ref:`sys.snapshots <sys-snapshots>` table and simultaneously deleting
  tables or partitions and running ``CREATE SNAPSHOT``

- Fixed an issue that could lead to an ``OutOfMemoryError`` when running a
  distributed query with aggregations or joins under memory pressure.

- Fixed an authorization issue for ``_blobs`` HTTP endpoint.

- Fixed an issue where :ref:`RESTORE SNAPSHOT <sql-restore-snapshot>` allowed
  restoring into existing closed tables, which is not a supported operation and
  could leave the restored table empty or unavailable. CrateDB now rejects such
  restores.
