Liverpool School of Tropical Medicine has replaced its legacy SAN and servers with a hyperconverged infrastructure system from Scale Computing, enabling it to improve the resilience of its data centre equipment and deploy disaster recovery capabilities.
“Our main priority was to upgrade our existing infrastructure and to make it more resilient,” says Matthew Underhill, IT leader at Liverpool School of Tropical Medicine (LSTM). “We can now test our backups more regularly and can recover VMs within five minutes, ensuring we will be back up and running again in the event of a disaster.”
Underhill is responsible for maintaining IT infrastructure at the 120 year old institution, which teaches over 600 students from 68 countries each year, conducting research into tropical diseases such as malaria and tuberculosis.
With research projects spanning years in many cases, reliability of infrastructure is vital for the institution in order to ensure data is available at all times. LSTM previously relied on an IBM Server BladeCenter chassis and MSA20-40 SAN to support its applications and workloads - from file servers and finance systems to data analytics - but the infrastructure became difficult to manage and to upgrade. Also, the BladeCenter servers did not have replication capabilities, meaning the institution had to look elsewhere for disaster recovery options.
It has now replaced the legacy equipment with Scale Computing’s HC3 hyperconverged infrastructure platform, setting up both a primary and secondary cluster for disaster recovery purposes.
One of the advantages of the preconfigured hyperconverged infrastructure platform was that it could be set up in less than a day. Underhill says that the hardware was also easier to manage than its previous data centre equipment.
Since the HC3 rollout was completed, management time has been reduced by 50 percent. “Previously if there was an issue with any of our servers we had to spend hours troubleshooting… but now it is a completely streamlined process.”
Cloud too costly
Underhill says that while the institution is moving certain systems to the cloud, it made sense to keep certain applications on premise for cost reasons, as well as the time taken to complete the migration of data.
"We have migrated our lot of servers to the cloud - so email servers and SharePoint," he explains. "We are heavily invested in Office 365 we are moving things up to Azure. We also have very large servers and amounts of data so it would take too long and would be too expensive to move that up to Azure at the moment.
"Simple things like print servers we thought it would be better to invest on-site servers. But because Scale was affordable we didn’t see it as too much of an investment [compared to] in five years’ time having to migrate everything to the cloud."
Choosing Scale over Nutanix
According to Underhill, the LSTM did look at other hyperconverged infrastructure and software vendors such as VMware and Nutanix before deciding on Scale.
“We like to think we are at the forefront of technology and it has come back to bite us a couple of times but we like to go with the latest and greatest tech,” he said.
However he said that Nutanix proved difficult to work with as a smaller buyer.
"I didn’t find [the Nutanix hardware] as user friendly as the Scale system and in all honesty they refused to quote us. We said what we wanted and went through our reseller to get it because obviously we have to have a price...but they never actually got back to us. Whether they thought we were too small, I am not quite sure."
The infrastructure rollout - which began last year with the initial purchase of an HC3 cluster - is now in its final stages and will be fully completed this summer, with LSTM adding more storage and server capacity to the existing infrastructure.