This post is also available in: Italian

Reading Time: 7 minutes

Scale Computing has an interesting storage hyperconverged solution: its HC3® platform can be used for small- and medium-sized businesses but also in some enterprise departments with a simple (and different) approach to virtualization and storage.

As data requirements at organizations of every size continue to require more storage assets and additional compute power, virtualization becomes a more-attractive option. For IT staffs at smaller organizations, the introduction of a virtualization layer can add complexity and management issues beyond what they are often prepared to handle. The new user interface deployed in Scale Computing’s version 6 upgrade features an intuitive design with almost no learning curve that allows administrators to employ a “set it and forget it” mentality where they only need to periodically log in to make changes to the system.

As written the solution use an hyperconverged approach for compute and storage, but the unique aspect (or at least quite different from other similar approaches) is that this product include also the virtualization and management layers.

Scale Computing integrates storage, servers, and virtualization software into an all-in-one appliance based system that is scalable, self-healing and as easy to manage as a single server. Scale Computing’s HC3 virtualization platform is a complete ‘data center in a box’ with server, storage and virtualization integrated into a single appliance to deliver simplicity, availability and scalability at a fraction of the cost of similar solutions.

Some days ago I’ve got the opportunity to talk with Jason Collier (CTO and founder) about the announce of the new release of HyperCore 6.0 and how their products are different from other similar.

First they are focusing a clear market segment that is the SMB one and trying to solve common issues and problems, first to all the complexity of similar environment both for the deployment and the maintaining phase.

Then the management part: usually you require dedicated server and some kind of infrastructure… in this case there is a web management distribuited on all the nodes in order to provide a simple management. The idea is not new (for example the XenCenter works in a distribuited way, but required a thick client, in this case you need just a web client) or is not the only way to simplify the management (other vendors, for example, are using a SaaS approach for this aspect). But considering the target, this approach is good and simple for the user point of view (there are several complication at design level, like how replicate the data, but in this case a custom database replication solution is used).

The web interface is portable enough to be used also from a mobile device, but actually the full support for mobile device will be added in future release.

Where is almost different different from other solutions it the built in hypervisor, that is KVM and it’s fully integraded in the management interface and included in each appliance. This choice is to simplify the deployment: each appliance is complete and ready to virtualize workload. From the other side… this mean no other hypervisors support… But just remember that the main target is the SMB market share.

Of course the virtualization platform include most of the common features (for the SMB market) like HA, live migration, … and now with latest release also a built-in remote disaster recovery feature to allow users to set up continuous replication on a VM by VM basis between two HC3 clusters with space-efficient snapshot technology replicating to a secondary site, tracking only the blocks unique to each manual or automatic snapshot and sending the changed blocks.

Testing a DR infrastructure plan is simple as cloning a snapshot on the target cluster and starting a VM with no disruption to ongoing replication. In case of disaster, users can simply “clone” a snapshot on the target cluster for the manual failover of a VM that is immediately bootable or simply replicate changed data at a DR site back to the primary site for simple failback.

Technically the HyperCore platform is based on a 64-bit hardened and proven OS kernel and leverage a mixture of patented proprietary and adapted open source components. As a result, Scale Computing has full control over the entire hardware and software stack enabling Scale to optimize and enhance all software layers for simplicity and reliability and tune the software for optimal performance.

The HyperCore filesystem is the Scale Computing Reliable Independent Block Engine (SCRIBE), an enterprise class clustered block storage layer, purpose built to be consumed by the HC3 embedded KVM based hypervisor directly. SCRIBE discovers and aggregates all block storage devices across all nodes of the system into a single managed pool of storage. All data written to this pool is immediately available for read or write by any and every node in the storage cluster, allowing for sophisticated data redundancy and load balancing schemes to be used by higher layers of the stack such as the HC3 compute layer.

The data redoundacy is similar to a distribuited RAID10 and actually each appliance has only 10k disks, but in the future a different placement algoritm could be introduced in order to supporto different type of disks… but note that will be only a placenment, not a tiering or caching solution.

The hardware platforms HC3 are converged appliance all with 1U form factor and Dell OEM PowerEdge: you can start from minimum 3 nodes to a maxium of 8 nodes (this is not a technical limit, but mostly a suggested limit to fit in the SMB target).

Each node has different specifications and different cost (cost is just per appliance, for more information see the sales brochure). In the future probably new type of nodes will be introduces (like a 2U with 3,5″ form factor for huge disks). Also the number of nodes may change in the future, for example for branch office user case, a two node cluster could be interesting, but actually is not yet supported.

Network topology is quite simple and it’s based by a front-end network (for management and VM connectivity) and a back-end network (for data replication and internal communications):

Fault domains or similar concept are not yet implemented… technically the failure grouping concept already exist, but is not visible in the GUI, so probably new version of this product may introduce this concept.

For P2V, there is a partnership with Double Take (with a rebranded HC3 Move product) in order to provide this kind of feature. For backup actually you can use traditional solutions, or replication or cloning features (although backup is quite different). Maybe in the future some automatically export functions will be provided.

See also:


Virtualization, Cloud and Storage Architect. Tech Field delegate. VMUG IT Co-Founder and board member. VMware VMTN Moderator and vExpert 2010-20 and vExpert Pro. Dell TechCenter Rockstar 2014-15. Microsoft MVP 2014-16. Veeam Vanguard 2015-19. Nutanix NTC 2014-20. Several certifications including: VCDX-DCV, VCP-DCV/DT/Cloud, VCAP-DCA/DCD/CIA/CID/DTA/DTD, MCSA, MCSE, MCITP, CCA, NPP.