Multi-cloud is interpreted as the use of more than one cloud service from more than one vendor. Technology leaders believe the multi-cloud allows them more flexibility to optimise performance and expand their cloud infrastructure. In this debate, the speakers discuss the reasons behind adopting a multi-cloud strategy and how this affects the business goals of the organisation.
With Chloe Tilley moderating, the speakers of this roundtable debate include:
- Paul Scott-Murphy, CTO, WANdisco
- Christina Scott, CPTO, OVO Energy
- Miloš Topić, VP, IT & CDO, Grand Valley State University
- Tom Tang, CIO, Alliant Energy
Adopting a multi-cloud strategy
Moderator Chloe Tilley asked the speakers to talk about the reasons why their organisation has or would consider adopting a multi-cloud strategy.
CPTO of OVO Energy, Christina Scott, stated that OVO was very much cloud-first in their approach. In the beginning, the organisation was set up as a cloud-native startup and acquired SSE three years ago. “I have every cloud environment so it’s very multi”, she said.
On the other hand, Christina argued that she will have to scale down in the future. They would still be multi-cloud, but with fewer cloud services. One of the reasons for this, she pointed out, is because there are specific tools that “lend itself to a better experience for different use cases”. While Christina believes that having a comparison between two clouds can be beneficial to the organisation, there are risks.
“With duplicate clouds comes a bigger security risk – a bigger overhead to manage”, she explained.
“Our position is really to look at cloud where appropriate”, said VP of IT and CDO at Grand Valley State University, Miloš Topić.
He stated that before joining the university two years ago, they were very data-centred. Miloš described how everything was in-house and they had large strategic partnerships with Amazon Web Services and Microsoft Azure. “I’m not a believer in cloud only and cloud at all costs… sometimes those costs are quite steep”, he said. He agreed with Christina’s earlier point about limiting the number of cloud services. He stated that technology leaders shouldn’t have all of their eggs in one basket. “You also don’t want them in 15”.
Multi-cloud resilience
In his view, Alliant Energy’s CIO, Tom Tang, believes we are multi cloud because everybody’s multi cloud these days. Going back to one of Miloš’ comments, he said, “Every cloud goes down”. He argued that it comes down to how you provide that resiliency and reliability through your strategy for customers.
When Chloe asked him how to provide that resilience, he stated that there are multiple answers to this question. “If you’re going to go with a single cloud, I would not necessarily advocate that”. Despite being able to span multiple geographical areas, he argued that if you want to use multi-cloud, leaders have to make sure that they balance their workload between those multiple clouds. This is something he described as “low latency, heightened secure fashion”.
Multi-cloud infrastructure
One subject WANdisco’s CTO, Paul Scott-Murphy, touched upon was the idea of thinking about the cloud as a service platform. He doesn’t want leaders to only think of the cloud as an infrastructure to build services or utilise applications.
Paul sees the cloud as a platform through which you deliver other services to organisations or to your customers directly. He pointed out that this may have an influence on whether technology leaders decide to adopt multiple cloud vendors.
Another aspect Paul believes factors into this is having customers who are located in regions close to the facilities provided by a particular cloud vendor. This, he argued, may be driving your organisation’s choice when considering cloud environments to adopt. Paul highlighted that the priority is to deliver those services to customers. “It’s not always about adopting services for your own use”, he said. For Paul, it’s about “potentially leveraging cloud to deliver functionality”.
Cloud and capability
Chloe asked the speakers which cloud provider is best for which capability.
There are a lot of different factors according to Christina. This doesn’t just include the capabilities, it’s also about the skills you have in your team or which regions you’re operating in. “Unfortunately it’s not a black and white answer”, she said.
Adding onto Christina’s previous point, Miloš said that skills and capabilities change. He stated: “I remember when we looked at Google’s platform several years ago it really struggled to handle backups overnight”. The platform, according to Miloš, would never finish tasks on time. He advised technology leaders not to select the cloud without looking at the entire stack.
In an example, Paul stated that in some instances, he has seen organisations become incapable of adopting their preferred cloud vendor because of demands imposed on certain regions. These demands could include restrictions on storage capacity or generic compute capacity. “While the cloud presents itself as an infinite pool of infrastructure and resources, it’s always the case”, he continued.
When choosing the best cloud provider for your capability, Paul argued that it depends on the workloads that are being brought to that cloud environment.
This roundtable was recorded at The Studio and made in partnership with WANdisco.
