CrateDB allows ingesting large amounts of data, from hybrid sources and at scale, while allowing real-time queries with a familiar SQL interface. With the release of CrateDB v4.6 we continue to improve CrateDB to achieve these goals even better, based on our customer and user feedback, adding improvements in performance, SQL compatibility, and security.
At a glance
- Improved security for node-to-node communication
- Performance and resource usage optimization for SQL queries
- Added data types and SQL compatibility improvements
- New functions for array aggregation
- Continue below for more details or find the full release notes here
Multi-node set-ups are a key part of CrateDB, allowing it to effectively scale, provide redundancy, and overall deliver the best performance. In most cases, all nodes are located in the same secure local network, but there are scenarios where a CrateDB cluster is distributed across multiple geographic locations and communication between the nodes could take place over a public network. Until now, this node-to-node communication would happen unencrypted over the public network. Securing this connection could be achieved with a VPN tunnel between the locations, but this requires additional setup and introduces a potential point of failure. Therefore we are introducing node-to-node encryption as part of the CrateDB v4.6 release. Node-to-node encryption allows all nodes in a CrateDB cluster to securely communicate over a TLS/SSL secured connection, removing the need for any additional measures to secure the node-to-node communication in a distributed cluster.
High performance on high-density data is at the core of CrateDB and we always strive to find more angles to improve CrateDB and enabling our users with the best possible performance.
In this release, we added additional optimizations for specific SELECT and INSERT FROM SQL commands that will allow these to be executed faster and better utilize the available resources in a cluster without impacting overall cluster performance.
Offering a familiar interface to accessing data is a core principle we follow with CrateDB and as such we are keen to enhance it’s the SQL capabilities of CrateDB. With CrateDB v4.6, we are introducing a bit data type with bit(n). This data type allows for storing binary information in a string-like format.
The COPY FROM SQL command also got extended to allow using wildcards for folders in the URI passed to the command. With this, importing from multiple files in a folder structure can be done in a more convenient way and you can easily import a larger amount of data.
Scalar functions for arrays
CrateDB already offers a number of scalar functions and specifically aggregate functions. With the release of CrateDB v4.6, we are introducing additional aggregate functions specifically for array operations. With these new functions, you will be able to easily get to the aggregate values contained in an array directly within a query.