Opera house shuns cloud for Nutanix and Rubrik hyper-converged infrastructure

The Royal Opera House considered moving all its infrastructure to the cloud, but a hybrid setup with Nutanix and Rubrik hyper-converged infrastructure came in 35% cheaper

The Royal Opera House (ROH) has been able to cover a storage and server refresh entirely from Opex budget, in a move that saw it replace a Dell EqualLogic-based iSCSI SAN infrastructure with Nutanix hyper-converged, Rubrik backup and disaster recovery (DR) and the cloud.

The refresh saw a significant portion of workloads moved to the cloud, but the organisation stopped well short of going all cloud because the costs would have been much more than the hybrid cloud approach it adopted.

The Royal Opera House puts on 420 performances a year and relies on ticketing and financial applications for revenue generation. It has recovery time objective (RTO) targets of 30 minutes with a recovery point objective (RPO) of five minutes.

ROH had undergone a £50m makeover and building upgrade at it Covent Garden site in 2018, and in the years following that it also sought to refresh an ageing IT infrastructure. This comprised two colo datacentres at Enfield and Slough, with Dell servers and EquallLogic iSCSI SAN storage linked to Covent Garden by WAN, plus AWS (the ROH website and shop) and Azure (Office 365, Sharepoint, file servers) cloud capacity.

The on-site equipment was near end-of-life by the time of the refresh, with head of technology operations Daniel Rubie describing it as “functional but inflexible”.

“If we wanted to add workloads or applications, it was a case of buying new equipment from Dell,” said Rubie. “It was maxed out. We wanted to move ticketing to AWS and connect it to the on-site infrastructure, but we lacked the flexibility to do that.”

Rubie said that to link the ticketing system to ROH’s events management software was a challenge. “You couldn’t just put it in AWS and everything would be taken care of automatically. And with the WAN, everything was costly and needed a lot of maintenance and a few members of staff to look after it,” he added.

After evaluating a number of options – including a move to all cloud that was rejected – ROH deployed a four-node Nutanix NX-1465 cluster with about 30TB of capacity over two datacentres. It has since added a further Nutanix node at Covent Garden to support streaming of products, which has become particularly important during the pandemic.

Meanwhile, it has also deployed a Rubrik backup appliance at its Farnborough datacentre that replicates to a site hosted by supplier ADP.

The hybrid cloud infrastructure now comprises a roughly 50/50 mix between on-site and remote, where previously it was just a few file servers and the website hosted in the cloud. But Rubie’s team discovered that moving everything to the cloud would not be worthwhile.

“Now it is the website, ticketing and Microsoft Office,” said Rubie. “We worked with Amazon and Microsoft to assess loads and see whether we needed to keep things, re-factor them or get rid, and we found that, while most of our infrastructure could work in the cloud, it was not going to be cost effective.”

Rubie calculated that if ROH had gone all cloud, costs would have been about 35% higher than the hybrid cloud environment with Nutanix and Rubrik.

“It would have been about 35% more expensive in cloud environment costs [capacity and data movement charges], but there would also be support,” said Rubie. “People think you just get AWS or Azure and just forget it, but there is configuration and maintenance and provisioning. Meanwhile, Nutanix is very easy to deploy and is almost maintenance-free.

“So, we made about 45% savings in Opex and were able to pay for it all from running costs, which is real bonus during lockdown when revenue is tight.”

Rubie said that with Nutanix and Rubrik, the organisation gets the best of both worlds in hybrid cloud: “The core environment is on-site and the cloud is used for backup and replication.”

In addition, the move has allowed the ROH technical and production team to bring all its data on-site and to centralise data and workflows, where previously “it had just been saving data where it could”.

Read more about hybrid cloud

Read more on Storage management and strategy