The whole pitch is the idea that “backups suck”, and they want to give administrators a much better back and restore experience, similar to Apple’s ‘Time Machine’ feature.
Rubrik are offering two “Brik” configurations initially in their r300 series:
- r330 – a 2U 3-node appliance with 10TB of usable disk capacity.
- r340 – a 2U 4-node appliance with 15TB of usable disk capacity.
Rubrik state that the cluster can be deployed in less than 15 minutes. They also tell me that there is guaranteed availability against disk or node failure. This is achieved by distributing multiple copies of the data around the cluster. They also have the ability to instantly mount any copy of the data, so if you need a copy of a VM from a point in time, this is possible on Rubrik. The system exposes itself as a scale-out NFS server to any host when a snapshot is mounted.
The appliances come with a Unified Console (Rubrik Reflect) which manages your data. The console provides visibility into protected and unprotected VMs, snapshots, SLA policies, storage usage, and throughput. The following screenshot shows the console:
So how does it work with vSphere? Once the user enters the credentials for its virtualized environment (e.g., vCenter username/password), Rubrik auto-discovers details of the entire virtualized environment, such as ESXi hosts, VMs and even applications. Rubrik utilizes VMware APIs (vStorage APIs for Data Protection) to discover VMware environments.
From the list of discovered virtual machines (VMs), the user selects which VMs to protect and what SLA policies to apply for recovery. Some pre-configured SLA policies are provided by Rubrik, but the user can create new SLAs if they so wish. A policy could contain requirements such as the desired snapshot capture frequency and/or data retention policy for example.
For VMware environments, Rubrik utilize VMware’s Changed Block Tracking (CBT) to identify and copy only the changed blocks from the previous operation. Deduplication and compression is applied before the data is stored in the appliance.
Users can also select where data is stored, whether on-premise in the Rubrik appliance or, as mentioned earlier, in a public cloud (e.g., Amazon S3). The user uses the UI to specify the time at which data should be stored in the public cloud (e.g. say after 30 days). Only deduplicated data is transferred to the cloud. Data in-flight and at-rest in the cloud utilize AES 256-bit encryption.
Rubrik are bringing together the backup software and backup storage into a single appliance. They are addressing common backup pain points, such as ease of deployment, ease of manageability and usage, ease of a scalability, as well as allowing users to utilize the economics of using cloud over tape for archiving.
These guys have a very nice solution, and I think they have identified a market that has not yet been tapped into. I expect to hear a lot more about them in 2015. Check Rubrik out here.