everythingpossible - Fotolia
Oxford-based medical equipment manufacturer Penlon has saved about £30,000 by ditching its HPE server and SAN infrastructure for hyper-converged infrastructure from Scale Computing.
The move enables the company, which specialises in anesthesia systems, to avoid spend on physical servers and SAN hardware, as well as VMware licensing costs.
Penlon has about 200 employees at two Oxford sites and in Minnesota, US, and serves 90-plus countries.
Its IT infrastructure is centred on one of the Oxford sites, with key applications that include Microsoft Exchange, enterprise resource planning (ERP), financials and document management systems.
The company had reached a situation of “server sprawl”, said IT manager Tony Serratore, and moved to VMware-based virtualisation. But then it experienced “storage sprawl” as virtualisation made extra demands for storage capacity.
“We are a heavily regulated organisation and we need storage to retain records,” he said. “Adding more disks to expand the SAN was a pain, and because of downtime, it also cost the business.
“To expand the SAN was extra cost and also brought technical challenges with having to move data around.
“We were looking for a simple solution to which we could add services with minimum hassle.”
Serratore said his team evaluated hyper-converged products from Nutanix and Simplivity but at the time these required the use of VMware. Meanwhile, Scale Computing offered hyper-converged with its own hypervisor and following a proof-of-concept test period, nodes were deployed at primary and secondary locations.
Serratore said: “With Scale there was no VMware, with all its licence costs and complexity. That is one reason to get excited about it because costs were getting big with VMware, for high availability, for example. And Scale is so simple to set up. There’s no need to hire a VMware certified engineer. I could show a cleaner how to configure a new VM.”
Read more on hyper-converged infrastructure
- We look at the latest in hyper-converged infrastructure – server and storage hardware nodes that can be built into scale-out clusters – and the players in the market.
- What are the advantages of do-it-yourself hyper-converged infrastructure deployment and who are the key suppliers of HCI software? We run the rule over the DIY hyper-converged market.
Penlon started with a deployment of three HC4000 nodes totalling about 16TB (with two out of eight drives being flash) at its main site, with three HC1000 nodes for “limp mode” disaster recovery provision at a second location. To handle storage traffic, 10Gbps Ethernet switches were deployed.
Key benefits of the move to Scale hyper-converged appliances included savings on servers, storage and hypervisor licensing, as well as ease of deployment and management, and quicker recovery from outages.
“We saved a hell of a lot around admin with the physical servers,” said Serratore. “We can deploy VMs a lot quicker and our time to recover from outages is down from days to hours.
“We spent £12,000 on our last SAN, a server averaged about £5,000 and we were quoted £15,000 for a dedicated ERP server, so we have probably saved close to £30,000 over two years.”
Serratore added: “There are no licensing costs or ongoing support costs for VMware. Now it’s just about licensing Windows, not the servers or infrastructure. We have improved capacity to meet business needs and reduced management time and the ability to plan use of our IT budget.”
Now Penlon has only one throat to choke in terms of compute, storage and virtualisation, said Serratore.
“We used to ring up VMware and they’d say, ‘Oh, it’s Vizioncore [backup software] – call them’, but that can’t happen any more.”