Practically every organization uses the public cloud for some part of their applications. Larger organizations — whether to meet diverse business needs or through mergers and acquisitions — use multiple cloud platforms. Most of these organizations also have applications on-prem, which makes it a hybrid multi-cloud.
The 22nd edition of Cloud Field Day focused on the challenges of connecting the disparate networks that make up a hybrid multi-cloud deployment, and delivering applications directly to end customers. One of the clear messages that came out of the discussions is that the fundamentals still matter in the public cloud.
Linking Disparate Networks
Cloud networks like on-premises virtual networks have their technology roots in ethernet and TCP/IP. Yet, every cloud provider and virtual network platform has a different view on describing, building and utilizing their network. Linking these systems and operating applications across locations is challenging. A third-party managed service may provide the best solution for joining all these networks.
Unlocking Multi-Cloud Agility
Often, organizations use the public cloud for the agility and speed of change it provides. This dynamic nature is even more of a challenge when operating a hybrid multi-cloud network, as changes in one cloud might require updates in another. For this, a software-defined mechanism for joining the networks and adding and removing sub-networks can unlock multi-cloud agility.
Abstraction and Integration
Modern IT has so many abstraction layers that it is easy to think that the underlying layers don’t matter, yet understanding the lower layers is vital for reliable deployment and troubleshooting. Each public cloud uses different services and abstraction models to provide fundamentals such as IP address assignment and name resolution. Modern, serverless applications can escape this infrastructure management until they must connect to existing servers.
Large organizations bring a legacy of virtual machines, database servers and load balancers, all of which live on IP networks; even serverless services rely on DNS to find each other. Cloud providers don’t usually offer an easy federation of these fundamentals across multiple clouds and on-premises data centers, leaving the integration to customers. Once again, a third-party tool with software-defined capabilities can aid companies in achieving the integration and agility they seek.
Risk and Responsibility
You may have heard the saying, “The cloud means I don’t need to care.” Nothing could be further from the truth. Every cloud provider operates a shared responsibility model and carries the minimum risk. When a cloud tenant’s application is compromised, it won’t be the cloud provider that suffers the loss.
Data and networks in the public cloud belong to and are configured by tenant companies. As a result, the fundamentals of securing and protecting them remain vital.
Data residing in SaaS applications is often a blind spot for businesses. There is an assumption that the SaaS provider is protecting everything. Again, if a tenant accidentally deletes data, it will not be the SaaS provider that is in trouble. Customers remain responsible for ensuring that protection meets their business requirements. A software-defined third-party service can provide the best-fitted data and network protection solution in the cloud.
In Conclusion
The nirvana of cloud computing, where everything works with little effort, is far from the messy and complex enterprise reality. Organizations using hybrid multi-cloud must ensure that the fundamentals of connectivity and protection are well-understood and appropriately handled. Ultimately, the public cloud is a tool to deliver value to the business, not as a way to outsource all risk and responsibility.
