Version 2.2.0

Released on 2017/09/19.

Note

If you are upgrading a cluster, you must be running CrateDB Version 1.1.3 or higher before you upgrade to 2.2.0.

You cannot perform a rolling upgrade to this version. Any upgrade to this version will require a full restart upgrade.

Warning

Before upgrading, you should back up your data.

Table of contents

Changelog

Breaking Changes

  • GROUP BY now executes against the real columns and fallbacks to substituting possible aliases in the case that the columns do not exist.

    This will cause statements that used alias values that shadowed multiple columns with the same name to fail (eg. alias name in a join statement where multiple tables have the real column name).

    Statements with aliases that shadowed a real column and also retrieve that real column will no longer fail as the real column will be used in the GROUP BY clause (this will not be ambiguous anymore).

  • Change semantics of IN operator to behave like = ANY.

    The argument list for IN now has to be comprised of the same type. For example, this is now an illegal IN query because the list mixes integer and double type:

    SELECT * FROM t1 WHERE id IN (1, 1.2, 2)
    

    The above would get translated into the following and throw an error:

    SELECT * FROM t1 WHERE id = ANY([1, 1.2, 3])
    

Changes

  • Added the shards plugin for the Enterprise Edition in the Admin UI.

    The shards plugin is a visualization that displays information about shards by table, partition and node.

  • Added new tab in Enterprise Edition of the Admin UI to display users and their privileges.

  • General Admin UI improvements.

  • Updated Crash to 0.22.1, which includes the following changes:

    • Added a status toolbar that prints the current session info.
    • Start autocompletion for non-command keys at the third character.
  • Allow IPv6 addresses in IP column type.

  • Upgraded Elasticsearch to 5.5.2.

  • Added node cluster check for cluster-name folder in data path directory.

  • Introduce support for single column subselects in ANY and IN, e.g.:

    SELECT * FROM t1 WHERE id = ANY(SELECT id FROM t2)
    SELECT * FROM t1 WHERE id IN (SELECT id FROM t2)
    
  • Improved resiliency of the table rename operation.

  • Relaxed column naming restrictions.

  • Improved resiliency of drop, close and open table operations.

  • Added empty tables KEY_COLUMN_USAGE and REFERENTIAL_CONSTRAINTS to INFORMATION_SCHEMA to be more compliant with the SQL99 standard.