How to run databases at scale

By design, all the FlashDrive infrastructure is serverless, meaning your app is not installed on a specific server but on a network of physical servers grouped inside a cluster.

That design doesn’t allow for horizontal scalability of databases (adding more nodes using the same databases and the same disks) because that would create collisions and data loss if several nodes write the same file at the same time.

The best way to achieve high performance and scalability with a high traffic database (hundreds of concurrent connections) is to create a replicated database composed of several nodes with independent disks. While this is not yet directly available from our marketplace we will offer this setup very soon. It can be done directly using Docker in the meantime.

The other way is to scale your database vertically by increasing the instance size. Big Data instances, for example, are specifically made for high-density databases with dedicated resources and can absorb a lot of data. But in this case, you will remain dependant on a single instance database with potential data loss between backups.

At this time FlashDrive doesn’t offer databases as a service but it’s compatible with all majors providers. You can connect a FlashDrive to any provider offering databases as a service like AWS or Google Cloud.

Was this helpful?

0 / 0