A software-defined future will save ForViva more than £300,000 over the next 5 years
February 3, 2020

I can think of countless organisations that could learn a thing or two from Quadris!

Danny Codling

Group Assistant Director of ICT, ForViva

ForViva is a forward-thinking social organisation with the single-minded goal of improving lives by creating opportunities for people, families, homes, and communities. 

Each of the 3 members of the ForViva Group has its own unique mission yet work closely together in order to have a lasting, collective impact. 

  • ForHousing: a progressive landlord that owns and manage more than 24,000 homes and delivers housing management services for other landlords. 
  • Liberty: the go to people for property services, thinking ahead to deliver personal, smart and altogether better property services across gas, repairs, maintenance and construction.
  • ForLiving: offering high-quality homes for sale, shared ownership and private rent across the North West. 

The challenge.  

When you have an existing IT infrastructure that appears to be working adequately, how do you justify replacing it with an entirely new solution? 

At the end of the day, much of it comes down to finding an alternative that is more efficient and cost-effective. In the case of ForViva, a progressive organisation that fully embraces the latest technology, they also wanted a solution that would deliver 24/7 availability with automated failover in case of a Disaster Recovery (DR) event. 

So, with no specific agenda in mind (Quadris is after all completely vendor agnostic) we were asked to take a long, hard look at ForViva’s legacy HCI platform, and this is what our deliberations revealed. 

Firstly, we noted that each node in the cluster required a controller VM with the sole responsibility for managing the storage in the system. The problem with this set up is that each controller VM has a number of CPU cores assigned to it. As a result, the VMs were using 15% – 20% of the overall CPU capacity in the cluster. 

This means that with a 10-node cluster, the controller VMs are using at least one full node of CPU resource. When you add an additional node all you are really doing is adding enough compute to run the controller VMs on the cluster; a significant detail that contributes considerably to the overall cost. 

The legacy Disaster Recovery design of the cluster is based on an Active/Passive configuration, using Asynchronous replication. Consequently, ForViva could only use the computing power of one data centre. The other side simply sat there almost redundant; its sole purpose being to wait for a disaster to happen. 

For every node the system required, ForViva would actually have to buy two. And they didn’t even get to use all of that compute power, because as we already highlighted the controller VM was eating up a large proportion of those resources. All in all, a very inefficient means of achieving a clearly defined set of goals. 

Due to the continued growth of the ForViva Group, additional resources needed to be added to the primary data centre in order to cope with increased demand. In order to overcome the inefficiencies explained above, at least 4 nodes were required, two at the primary data centre and two at the DR data centre. 

Another important point that needed careful consideration with regard to the existing infrastructure was the fact that much of the equipment was fast approaching end of life. Simply adding extra nodes would merely be playing catch up and ultimately result in throwing good money after bad. 

The solution. 

After completing a thorough review of the existing system, we designed a future-focused solution that would deliver ForViva with a true Private Hybrid Cloud Software Defined Data Centre. 

In order to achieve this aim, we started by recommending moving to a software-defined storage system using VxRailAs VxRail doesn’t require controller VMs it would automatically require fewer nodes to achieve the same aims. In addition, it meant that ForViva could also go All-Flash and enjoy the many benefits of a smart storage system at greatly reduced cost. 

We also proposed greatly improving their DR capability by employing a stretched cluster configuration, combined with an ingenious GSLB software defined network solution. This would provide ForViva with a fully Synchronous DR as opposed to their previous Asynchronous setup; delivering automated Active/Active DR capabilities with zero RTO/RPO recovery times and supported by fully redundant point-to-point 10Gbps links. 

The advantage of this solution means that both data centres are live, so there is no longer a disaster recovery site. Workloads are now spread evenly and resiliently across both data centres and are free to move between them, as and when required. In the event that an entire data centre goes down for any reason, workloads simply restart in the second data centre automatically and within seconds. All external network services seamlessly failover using the software defined network design. This is a Disaster Avoidance solution! 

This new configuration also offers potentially huge savings; not only with regard to the operational benefits to the business, but also in the reduction in the administration of replication jobs and testing, as well as many associated risks. 

Furthermore, Quadris’ innovative new solution would provide ForViva with many additional advantages, such as broad vendor support for the hypervisor, together with the reduction in risk to future projects. 

The cost analysis and projected savings. 

Unsurprisingly, ForViva wanted to see what the predicted cost would be if they stayed with their existing infrastructure, or if they employed Public Cloud DRaaS, compared to upgrading to our new Private Hybrid Cloud Software Defined Data Centre solution. 

So Quadris’ senior technical team produced a comprehensive consultation piece that would clearly spell out the various alternatives and their respective costs. This involved the creation of a 5-year cost analysis showing all the milestone investments they would have to make, including ongoing hardware and software replacements etc.  

The findings were quite astonishing, with Quadris’ SDDC solution proving the clear winner. 

In addition to upgrading the entire IT infrastructure, the calculations showed it would deliver a staggering £300,000 saving over the alternatives. And not just in Total Cost of Ownership (TCO), but in actual money. If you add TCO savings from administration, DR testing, documentation, and all the complexities that come with updating the infrastructure, the eventual savings will probably be doubled. 

Based on Quadris’ in-depth knowledgeexpert advice, analysis, and the clear cost-saving advantages offered by this masterful future-focussed solution, Danny Codling Group Assistant Director of ICT at ForViva gave the project the green light. 

The implementation. 

The implementation of the complete solution was divided into two distinct phases: 

Phase 1. 

The first phase involved moving ForViva to a software-defined storage system.  

But while we were on schedule for the agreed timescale, the installation was brought forward owing to a major problem that occurred with the existing infrastructure. Fortunately, we already had the new system in Predeploy so with typical Quadris flexibility and speed of response we went into action. 

With less than 24 hours’ notice the first tranche of VxRail nodes were installed, commissioned, and workloads were being pushed live. To add to an already impressive feat, there was no impact to ForViva’s day-to-day operations; no downtime, no incidents and everything remained online while we undertook the migration.  

The performance benefits were immediate and impressive; with login times slashed by a staggering two thirds. 

Phase 2. 

The second phase is due to commence in just a few months. 

This will entail the adding of the additional VxRail nodes into the system and setting them up into stretched cluster; delivering the fully automated Active/Active DR capabilities.  

This will ensure that ForViva not only make maximum use of all their compute power, they will also enjoy the peace of mind that comes with knowing that in the event of a disaster the automated failover will ensure that they will continue to deliver 24/7 availability. 

The key project benefits at a glance. 

  • Migrate from outdated legacy infrastructure to a future-focused SDDC solution.
  • More efficient use of all available compute power across both data centres.
  • Upgrade to All-Flash storage.
  • Fully automated Active/Active DR capabilities which delivers with zero RTO/RPO recovery times.
  • Automatic failover delivering Disaster Avoidance as opposed to Disaster Recovery. 
  • Second data centre site added with fully redundant point-to-point 10Gbps links.
  • 24/7 availability of all services.
  • Cost savings of more than £300,000 over the next 5 years.

What challenges do you face?

To discuss your IT challenges and how we can help, just pick a slot in the calendar and Peter will be happy to help. Alternatively, feel free to call 0161 537 4980 and ask for Peter Grayson or email peter.grayson@quadris.co.uk 

More Articles