Filter on tags All Bring down costs Hardware Performance Scaling Self-healing


  • Q: Is any kind of design more suitable when looking at cost?

    A: From a cost perspective, a horizontally scaling design is the cornerstone for a high availability and high performance cluster, especially based on commodity hardware.

    Performance Bring down costs

  • Q: Do you have to build using expensive hardware?

    A: With no single points of failure and no performance bottlenecks to protect, it is possible to chose cheaper off-the-shelf, low energy hardware.

    Hardware Bring down costs

  • Q: What makes the cluster scale linearly?

    A: To fully scale horizontally, the Compuverde software in the storage cluster is designed to make the storage nodes independent of each other, but at the same time cooperate to form a homogenous storage cluster.


  • Q: What happens when hardware eventually fails?

    A: The Compuverde software is designed from the perspective that everything will break sooner or later, and to handle these errors transparently to the user.


  • Q: How is scaling up the storage cluster affected?

    A: Multicast enables higher cluster elasticity compared to a static DHT cluster. There is no need for moving data within the cluster when changing the size of the cluster, which is needed by design in a DHT cluster.