The DriveScale Composable Platform
The DriveScale Composable Platform enables you to create your own server platform on the fly with the exact compute and storage resources required for an application or workload using heterogeneous, low-cost compute, storage and network fabric. With DriveScale, compute and storage are disaggregated from the traditional direct-attached storage server and connected to a standard Ethernet network and composed into adaptable servers, then clustered into a high scale, high-performance platform for bare metal or containerized applications.
Compose a Server
A server is created in minutes using Composer, the DriveScale orchestration application. In an easy-to-use interface, users provide requirements and policies and the server is automatically created from pools of disaggregated, low-cost compute and storage resources. When a workload completes, resources can be returned to the pool to be composed for the next workload.
Servers are composed then clustered to create an application-ready platform.
Compute nodes or drives can be replaced in seconds.
Rapidly Replace Failed Elements
Today, when a storage drive fails, someone needs to physically pull and replace it. With DriveScale Composer, when a drive fails, it can be detached from the server and replaced by another drive from the available pool – in seconds. In addition, traditionally when a server fails the only option has been to order a new one. Like a drive, using Composer, a failed compute node can be detached from its drives and a spare compute node from the pool can be attached in seconds.
When a DriveScale server is created, drives are mounted to compute nodes over an Ethernet fabric. After a workload completes, data on the drives can remain in place and compute resources can be detached and returned to the resource pool to be re-used with a separate application or workload. This unique capability means compute doesn’t sit idle and network congestion is minimized.
Compute nodes in a server can be detached from the cluster and re-used.
Architected for high availability, performance, security and cloud scale