The unveiling of Cisco HyperFlex Systems

On the 1st March 2016 Cisco will announce a new HyperConverged solution based on their existing UCS C & B Class compute, existing UCS Fabric Interconnects and a new Software Defined Storage approach that allows Cisco to offer that “Complete Solution” inclusive of Compute, Storage and networking where as other HyperConverged are missing that element that Cisco do really well, Networking!

“True Hyper-convergence”

Before I dive into the deep information regarding this new range from Cisco I want to touch on the current state of Hyper-convergence, a caveat from myself here is that I for one do not do much in this space and I am working on resources that I have read since learning of this new range from Cisco.

The story so far with HCI, make it simple to deliver and bring to market as fast as possible then try and innovate to bring up to market speed, but ultimately its about delivering that modular way of increasing performance and capacity to an environment leveraging new technology and possibly software defined where possible.

Back to Cisco HyperFlex and the entrance of True Hyper-convergence, what does that mean? Well other main vendors within the HCI space only really bring Compute and Storage to the table. Cisco will bring Networking along with them.

Why does this make Cisco a bigger differentiator in this specific market? Basically the reason is simply because Customers need more from these HCI offerings, Simplicity is still key, having the ability for not having to have separate silos of hardware for different roles within the data centre, an easy scaling model for all entities. Being able to add resource efficiencies specifically around the storage. Today we have our applications but who knows what is coming and how our worlds will change with the likes of containers and other movements within the market.

  • Deliver Simplicity
  • Easy Scaling
  • Smash The Silo’s
  • Resource Efficiencies
  • Todays Applications and Tomorrow’s World

The Missing Part of the puzzle – Introducing Cisco HyperFlex Systems

Let’s move deeper into the stack, we have mentioned the “Complete Hyper-convergence” above being able to use Software Defined “Everything” Compute, Network and Storage. The HyperFlex system also brings additional “Always-on” storage optimisations and flexible scaling.  The final piece of the puzzle is the ability to bring a “Future Ready Architecture” to market that is built for our applications today but also emerging workloads.

The fundamental backbone to this offering is the Cisco UCS platform, the already proven High performance network capabilities in the Fabric Interconnects. The virtualisation aware compute platform, and the ability to automate management of all hardware pieces.

The difference between first gen HCI and HyperFlex

Taken from Cisco press briefing
“Built on legacy file systems intended for local-machine write-in-place operations” in reference to first generation HCI systems.

The Cisco HyperFlex has been built from the ground up for Hyper-convergence. The key areas to mention here are as shown below, In the best endeavour to keep this post as short as possible I will look to dive deeper into each piece when the time allows.

  • Log-Structured File System.
  • Designed for distributed scale-out storage.
  • Inline data optimisation with low performance impact.
  • Superior flash endurance.

In particular, the feature rich software defined storage brings a lot of great benefits to the table

  • Independent Scaling of compute and capacity
  • Dynamic Data Distribution
  • High Resiliency / Fast Recovery
  • Continuous Data optimization
  • Integrated Management and Data Services
  • Native Snapshots
  • Native Clones

The HyperFlex Portfolio

The entry point is 3 nodes, VMware only on day 1 but with a roadmap to ensure other virtualisation platforms and future workloads. With the ability to add existing UCS platforms that are already in your Data Centre this allows for a seamless integration joining the picture and flexibility together.

I will be looking to look at how Veeam can sit well here and assist in delivering the availability for the new platform, this will include backup but also a key feature missing from the software defined storage is the ability to replicate to a secondary site or cluster, Veeam brings that to the table along with all the other great benefits around application consistent backups.

Resources

Tech Field Day – Springpath Corporate and Data Platform Overview – https://www.youtube.com/watch?v=Eq8J6OOLG5s

Tech Field Day – Springpath HALO Architecture Deep Dive – https://www.youtube.com/watch?v=krmkywnz970

Cisco Champions Call

Leave a Reply

Your e-mail address will not be published. Required fields are marked *