As Docker container use continues to grow, there is a need for a distributed database technology that is built for the container era, which is precisely what Crate Technology is now aiming to deliver.
Crate Technology today announced a $4 million seed funding round to help build, develop and extend its open-source-based container database technology. Among Crate Technology's investors is Solomon Hykes, the founder of Docker Inc.
Christian Lutz, CEO of Crate, noted that Hykes is often interested in technologies that build on Docker technology, expanding the utility for everyone in the container ecosystem. "We started from the ground up to build a database that is distributed and designed to run in a master-less environment," Lutz told eWEEK.
In a traditional database structure, there is often the concept of master and slave nodes to help distribute and scale data. In the master-less approach, any node can be the master, which is intended to enable a faster, more scalable approach. As part of Crate, Lutz and his team have also built a distributed SQL query engine that allows for quick searches and data retrieval.
"Crate feels like a very normal single-node database, but actually it's a very powerful distributed system," Lutz said.
The Crate technology uses some existing open-source components to help enable the system. Crate uses the open-source Apache Lucene and ElasticSearch technologies as the underlying cluster management tools, Lutz explained. On top of that, Crate also uses Facebook's open-source Presto distributed SQL query engine.
From a container perspective, Crate uses Docker to enable its database deployment. Lutz explained that a user can simply deploy a multi-node container-native Crate database with Docker.
"Every node can be a master so there is no dedicated master node," Lutz explained. 'We're utilizing automatic sharding and replication."
Lutz explained that as long as the individual container nodes can see each other, they will automatically join the crate database cluster. He added that when a developer or application sends data to the crate database, it doesn't matter with which specific crate database container they are directly communicating. "Any node can take the master role, and the Crate system will figure out how to most efficiently process the queries that you throw against it," Lutz said.
At this early stage for Crate, Lutz is already seeing the technology used as a replacement for an existing database, where an organization has scalability or performance issues. Crate has helped security vendor Skyhigh Networks with a replacement for the open-source MySQL database.
From a data perspective, there are some in the container community that advocate for the use of some form of plug-in to enable persistent data storage, but that's not the approach that Lutz is taking with Crate. Lutz added that the persistence of the data volume is provided by Crate
"We don't think plug-ins make sense as they don't scale," Lutz said. "Crate is a native container, so you don't need to use a plug-in to create a data volume and move it around."
Part of the new funding that Crate has raised will be used to help build out an enterprise Crate technology that Lutz expects will hit the market in the second quarter of this year. The enterprise features will provide additional authentication capabilities as well as advanced data center replication features.
Sean Michael Kerner is a senior editor at eWEEK and InternetNews.com. Follow him on Twitter @TechJournalist.