skip to Main Content

All aboard the Databus!


While every DBA has a business imperative to migrate data to the cloud, the task is full of risk. Production must continue without disruption and data must remain secure. Cloud data must be verified to be consistent with on-prem data before making any changes to legacy on-prem databases.

Several tools and products claim to fit the requirements for migrating and modernizing data. However, a product’s underlying design philosophy can make the difference between successful project completion and empty claims.

Introducing Databus

In the last decade, the engineers and architects at LinkedIn created several enterprise data management systems.  These systems included Databus for transactional data and Kafka for behavioral and click data.

From its inception, Databus has focused on transactional data. Databus guarantees timeline consistency between sources and destination transactional databases. This means that the relative order of changes on the target always matches the source. In addition, the Databus pull-based architecture enables extremely high durability and nearly infinite scale for synchronization of the largest enterprise databases.

After hardening Databus in production for several years, LinkedIn released the core technology to open source. More information about Databus is available on SlideShare or the original ACM paper

Enter Griddable

To address the demanding requirements of synchronizing transactional data, Griddable selected Databus as its core foundational technology. Griddable then extended this open core with several valuable management and usability enhancements.  

Using Databus, Griddable creates a resilient scale-out grid architecture that supports any topology of heterogeneous databases. With Griddable, you can easily deploy hyper-connected topologies across clouds in minutes using a graphical UI. Further, Griddable synchronizes connected source and destination databases in real time with sub-second latency using this high-performance grid infrastructure.

Intelligent in-line services connect Griddable end points and selectively filter, mask, and transform data while guaranteeing transaction consistency. The Griddable data pipeline filters and transforms data at both the source and destination. Consequently, users have fine-grained control of the data in each target database, regardless of database or cloud type. 

Griddable policies are easy to setup and change, eliminating the need for complex coding across a myriad of single-function tools. Users can configure all grid services using do-it-yourself policies written in JSON or created with a graphical schema browser. Define and tag groups of similar data, like personal data, and use tags to simplify policy definitions. 

Next step

Over the following years, through the emergence of many competing designs, Databus has proven its value. Transactional data requires a grid specifically designed for its demanding semantic requirements. Databus meets these demands by applying the best of distributed system design principles. 

Griddable delivers these important design principles in a product that excels in flexibility and ease of use. See for yourself. Click the “Live demo” button for a guided tour of what Griddable can do for you.

Back To Top