|CockroachDB Production Ready|
|Written by Kay Ewbank|
|Friday, 19 May 2017|
CockroachDB 1.0 has been released as the first open source, cloud-native SQL database. The announcement included details of $27 million new funds invested in the company behind CockroachDB.
The developers say CockroachDB provides distributed SQL to accommodate ever-larger data sizes and multi-active availability, a new model for high availability (HA). They describe it as being cloud native because it offers horizontal scalability, no single points of failure, survivability, automatable operations, and no platform-specific encumbrances. In practical terms, this comes down to distributed SQL to make it usable for small and large requirements, with the ability of scaling between them; multi-active availability for always-consistent high availability; and flexible deployment supporting a wide range of environments.
The distributed SQL element of this comes from support for fully-distributed ACID transactions, zero-downtime schema changes, and support for secondary indexes and foreign keys. It works out of the box with many popular ORM frameworks because of its support for an industry standard SQL dialect. Release 1.0 also introduces a distributed query execution engine, so you can set up distributed JOINs to support analytics queries that speed up linearly as nodes are added to your cluster.
The developers recently demonstrated an application with Baidu processing 2 billion inserts a day for one of the top ten largest global internet companies. This rate of transactions continued regardless of artificially-induced “chaos” events, which caused significant concurrent re-replication and rebalancing across the cluster.
The most important part of CockroachDB's high availability element is consistency, according to the developers. Unlike other products that rely on eventual consistency, it employs strongly-consistent consensus-based replication, which uses three or more active replicas, any of which can begin serving read/write client traffic. All the resources are utilized dynamically, and the documentation says that unlike the more common active-active replication, this type of availability will not read or write inconsistently and doesn’t require conflict resolution.
CockroachDB can be managed either with datacenter orchestration technologies like Kubernetes and Docker Swarm, or and even manually. It can run on premise or on any public cloud, deployed in hybrid cross-cloud configurations, and migrated between clouds with zero downtime. At the OpenStack Interop conference in Boston, CockroachDB was deployed live across 15 private clouds. A video of the demo is below:
Alongside the open source free version, the team announced a feature available specifically as part of the CockroachDB Enterprise offering: distributed, incremental backup and restore.
This feature is intended to serve customers with large data sets. The distributed backup / restore parallelizes backup and restore tasks across all nodes in the cluster. Data can be written to and restored from any configurable storage sink, such as blob stores offered by AWS, GCP, and Azure. A non-distributed option for backup / restore is available for free in the CockroachDB Core offering.
or email your comment to: firstname.lastname@example.org
|Last Updated ( Friday, 19 May 2017 )|