Version 1.0.2

Released on 2017/01/09.

Note

If you are upgrading a cluster, you must be running CrateDB 0.57.0 or higher before you upgrade to 1.0.2.

If you want to perform a rolling upgrade, your current CrateDB version number must be Version 1.0.0 or higher. If you want to upgrade from a version prior to this, the upgrade will introduce all of the breaking changes listed for Version 1.0.0, and will require a full restart upgrade.

Warning

Before upgrading, you should back up your data.

Changelog

Changes

  • Do not display current statement in sys.jobs that disabled stats.

  • It is not possible any more to circumvent NOT NULL column constraint by omitting the column in the INSERT statement.

  • Updated crate-admin to 1.0.3 which includes the following change:

    • Added compatibility with future CrateDB versions which will serve the admin-ui from /admin/ instead of /_plugins/crate-admin/.

Fixes

  • Fixed a NPE when using a percentile aggregation in a multi-node environment.

  • Fixed INDEX constraint validation. Defining it on complex data types like e.g. object was silently ignored instead of throwing an error.

  • Fixed an issue that causes UnsupportedFeatureException for operations on a table where array access is used inside generated columns. e.g.:

    1. CREATE TABLE t1 (cola string, gencola AS (a[1]))

    Error thrown when: SELECT * from t1, DROP TABLE t1, etc.

  • Fixed an issue that causes hanging queries if a data node disconnects from the cluster.

  • CrateDB now throws a correct error if a user tries to create a table starting with _.

  • Fixed an issue that causes UnhandledServerException on joins when WHERE clause contains conditions on columns of array type of both tables which are separated with the OR operator. e.g.:

    1. SELECT * FROM t1 join t2 ON t1.id = t2.id
    2. WHERE 'foo' = ANY(a.strArray) OR 'bar' = ANY(b.strArray)

  • Scalar functions longitude and latitude no longer show misleading results.

  • Fixed an issue that causes ClassCastException on outer joins when WHERE clause contains a condition that prevents null values on the relation that may produce nulls. e.g.:

    1. SELECT * FROM t1 LEFT JOIN t2 ON t1.id = t2.id WHERE t2.txt = 'foo'

  • Fixed NPE that occurred when an UPDATE statement inserted NULL value into a column that did not exist before.

  • Fixed an issue that causes an error to be thrown if nested method calls are used in HAVING clause. e.g.:

    1. having sum(power(power(id, id), id)) > 1