The private cloud is the fake cloud

Telecoms.com periodically invites third parties to share their views on the industry’s most pressing issues. In this pieceDanielle Royston, CEO of Optiva argues in favour of the public, as opposed to private cloud for telcos.

Confusion reigns when it comes to cloud and telecoms. CTOs are looking to cloud architectures to increase their processing power, scalability and savings. Yet they are in the dark ages! For starters, there is confusion about private and public clouds and how they differ. Some CTOs think that they can get the same benefits of public cloud with private cloud. Sure, private clouds are ideal if you have money to burn – it is a lot of effort with no rewards.

Often when most telco CTOs refer to the cloud, they mean private cloud deployments, hosting telecom infrastructures on-premises. Most are unaware that private clouds can be complex and costly. When using private cloud, compute and database resources still need to be pre-purchased and provisioned — typically up to 20-30% above peak capacity. Even when the data is virtualized, there is little or no elasticity. Instead, it’s being held in an over-provisioned, private cloud. Basically — a fake cloud.

Moving to the public cloud, on the other hand, can unlock huge cost savings by using external compute resources and it can quickly deliver benefits only possible with a true cloud deployment. With the public cloud, operators can provision for average capacity and scale up or down as needed. That can in turn make a significant difference to the bottom line.

Learning from the enterprise

Many enterprises have deployed Google Cloud, AWS and Azure. For example, Salesforce uses Google Cloud and AWS to host their application services, and Twitter recently moved to AWS. But the telecoms industry hardly hears about the major cloud moves. Why? Most telcos are stuck in the 90s of application management — virtualizing servers, rather than managing IT stacks in the public cloud.

These are some of the key benefits in enterprise that operators can replicate:

How not to go ‘cloud-native’

The public cloud’s architecture cannot be replicated with an on-premise cloud. Taking existing software architecture and dropping it into Google Cloud or AWS (known as a “lift and shift”) will not deliver the benefits operators hope to gain from the cloud. While “lift and shift” is a quick and dirty way to move to the cloud, it’s not cloud-native, and it’s not the same as an application architected to run “cloud-natively.”

Cloud-native means that an application is built from the ground up to work in the cloud. It is the responsibility of CTOs to dig into vendor terminology that might be misleading, such as “cloud-ready.” Doing so helps to ensure that what you are considering is genuinely fit for the carrier’s purpose and allows the operator to realize the real, full benefits of the public cloud.

If an application claims to be “cloud-ready” or “cloud-native” but requires an old-school Oracle relational database, you’ll have serious problems — and vast license fees too. Instead, a genuine cloud-native database will be faster, more powerful and cost-effective. A good sign that your application is architected and fully suited for the cloud is also when it utilizes containerization, like Kubernetes.

A turning point

It’s surprising that more CTOs have not connected the dots when it comes to the benefits of public cloud. Given the cost pressures facing the industry, the savings alone of moving to the public cloud should be a no brainer. 2019 has been a turning point for telco engagement with the public cloud — one of the biggest technological innovations of the past decade. Don’t take the easy route and go with a fake cloud.

Danielle Royston is CEO of Optiva Inc. (TSX: OPT) and has close to 20 years of executive experience in the technology industry with an emphasis on turning around enterprise software companies. Before joining Optiva, she served as Portfolio CEO for the ESW Capital family of companies, leading over 15 turnarounds during her tenure. Royston holds a B.S. in computer science from Stanford University.