Tearfund saves £950,000 by deploying Hypergrid hyper-converged in place of HP servers and SAN, cutting costs on space, power, staff and hardware, and avoiding VMware
UK charity Tearfund has deployed Hypergrid hyper-converged infrastructureand saved nearly £1m. The company replaced its traditional server, SAN, and networking equipment, and has saved money on equipment, staff, power and space.
Tearfund has around 500 UK staff, with 1,000 worldwide, and runs two datacentres, one near London for its primary production infrastructure and a secondary disaster recovery location in south Wales.
Key applications are a bespoke customer relationship management (CRM) platform and accounting software that run in a Microsoft Hyper-V virtualisation environment on around 86 virtual machines, with physical servers retained for Active Directory, backup and the PBX.
Tearfund previously had HP blade servers and LeftHand SAN storage with around 70TB of capacity, but this was nearing end of life and its performance was suffering, said Stuart Hall, lead infrastructure engineer at Tearfund.
“It didn’t go fast enough and didn’t store enough,” he said. “We have to deliver decision support via a balanced scorecard approach that depends on OLAP processing. It was taking 12 hours and even when we optimised it to within an inch of its life it took two, but that expanded to four or five if other processes were running.”
“The bottom line is that reports were not being generated quickly enough,” he added.
Hall was exposed to the idea of hyper-converged infrastructure at a trade show but, when it came to evaluating products, only two were in the running – Nutanix and Hypergrid – because Tearfund wanted to retain Hyper-V for cost reasons.
Hyper-V was already in place and, because of its charity status and eye for cost savings, VMware was ruled out, said Hall.
“Obviously we absolutely cannot engage with VMware. It’s a cost issue. Hyper-V was what we had in place and we had too much collateral, with PowerShell playbooks in place that we depend on.”
The challenge for the two hyper-converged suppliers was to bring down overnight processing to less than one hour. Both were subject to proof-of-concept testing, and Hypergrid managed to carry out the tasks in less than 20 minutes “day-in, day-out”, said Hall, while Nutanix came in at around one hour.
“Both suppliers did what was asked,” said Hall. “But it came down to cost and there was a clear winner in terms of performance, which was important in terms of having room to grow.”
Tearfund deployed three nodes of Dell servers at each site with Hypergrid on board and 28TB of flash storage in each box. Space occupied has been cut from seven 42U racks to just half a rack at each site.
Hall said the advantages are that as well as getting key processing tasks done in good time, the charity has also saved around £950,000.
“I’m running around 80% to 90% utilisation and Hypergrid doesn’t break a sweat,” said Hall. “Power consumption is around 2.2kWh where it used to be 22kWh – that’s a saving of around £200,000 over three years, which means the solution pays for itself.”
Tearfund now only uses around 0.5 FTE in human resources where it used to take four people to run the previous infrastructure.
“We’ve come out with exactly the same processes and tool sets and moved to the new platform overnight with no downtime at all,” said Hall. Total savings are approximately £950,500 over the five-year lifetime of the system, he added.