Edge Computing

Cloud databases therefore only play a subordinate role in edge computing. An additional limiting dependency is the constant need for online connections. Even a brief loss of connection could bring entire production lines to a standstill. Resilience looks different. Therefore, sensor data is processed in real time directly on-site with edge databases, which have reduced functionality tailored to edge needs (e.g., analytical capabilities to evaluate peripheral data). On the other hand, edge databases do not need to be particularly powerful or highly scalable, nor do they need full-text search capability.

Cloud databases enter the scene again for consolidation downstream and to store data no longer needed at the edge. For this to happen, you need a database that maps the decentralized architecture of edge computing and supports its special features. The pared-down edge database and the full-format database in the cloud need to understand each other and complement each other in terms of functionality.

The Cloud Is Everywhere

That said, cloud databases can be useful in virtually any conceivable deployment scenario. In terms of functionality, they are by no means inferior to in-house databases. They come in the SQL and NoSQL database flavors that harmonize well with virtualized and containerized environments (keyword: cloud native). They typically tie up fewer resources and are readily available, highly scalable, and usually less expensive to operate.

The Author

Gregor Bauer is Senior Solutions Engineer Central Europe and a technical team leader and presales engineer at Couchbase, a US enterprise database provider. He focuses on NoSQL databases, web technologies, device management, edge computing, and mobile applications.