The public cloud has expanded how organizations can deliver value to customers in their respective markets. Organizations today have taken on Multi-Cloud, and with that, see a proliferation of where data lives. The post-pandemic hybrid workforce has accelerated the need for organizations to upgrade their data protocols to be centered around the cloud rather than on-premises.
Instead of having a few data centers, organizations look at the arduous task of transitioning to "centers of data." While this may sound simple, it increases the number of ways data can be accessed that may introduce risky patterns into operations. These include:
- On-premises to Cloud Endpoints (clouds)
- Cloud Endpoints to on-premises (data centers)
- Internet ingress and egress (multiple clouds and on-premises)
- Significant data center backhaul (for scrubbing or access to data)
- Loss in visibility hampering day-2 operational efforts
Building for the future
Technical practitioners tend to provide solutions for the problem at hand. However, public cloud consumption expands quickly, and those solutions need to do far more with less overhead. Some of the strategic factors include:
- Applications extend beyond single availability zones and require multi-regional capabilities to meet DR requirements.
- Lines of business, mergers, and acquisitions extend that single cloud to multiple clouds.
- SaaS adoption leads the charge for organizations advancing timelines for transformational initiatives.
- Hybrid work means enabling employees to work when and where they are most productive.
- Many large organizations still have a large portion of their application portfolio running on-premises and need to pivot quickly.
Bridging old and new
The requirements and building blocks for the network of 2022 and beyond are complicated. Connecting data centers to the cloud for application migrations is one thing, but managing complex networking inside and across the public cloud providers sets the bar higher for network practitioners.
Two completely different worlds that were typically handled separately now collide, and getting clarity, standards, and building patterns may seem impossible with existing network technologies and practices. To arm organizations with the means to segue into the cloud era, let's reimagine networking built for the cloud and managed like the cloud.
Thinking 'as-a-service'
Traditional network design deploys virtual appliances, routers, firewalls, or other types of devices in each network that requires peering. Furthermore, cloud-native solutions do not easily enable "transitive routing," which adds more difficulty in provisioning networks heterogeneously. Fortunately, there are new ways to "normalize" provisioning and operating networks inside and across multiple cloud environments without leaving on-premises networking behind.
In using Cloud Network as a Service, organizations can take advantage of consistent network policy, security, and governance across the entire network. If you are already using a 3rd party vendor, integrating best-of-breed products comes natively. Organizations of 2022 and beyond will require new features, but they will also need the flexibility to navigate the complicated transformation landscape. Taking the great innovation that the public cloud ushered in, and applying it to the full domain of networking, brings a whole new set of benefits.
Cloud scale
Elasticity is a system's ability to scale up or down as demand increases or subsides. How does a system adapt to heavier demand? Historically, vertical scaling has solved this problem for the network. Practitioners have met the capacity needs by adding larger nodes that accommodate more throughput. As enterprises have scaled in the past, costly hardware refreshes replace gear not meeting demand with higher-capacity network gear. A transition from vertical to horizontal scaling is taking off.
Instead of adding additional resources to a box or replacing that box with a bigger box, you work from a larger pool of resources. As demand increases and the pool of resources needs more capacity, more routing functions are added horizontally to accommodate the scale required. As demand for capacity subsides, routing functions are removed from the pool of resources.
Horizontally scaling also lends itself to the idea of consuming what you use. If your network usage fluctuates, you don't want to pay for the max cost continually. Having the ability to pay only for what you are using carries tremendous advantages.
Embracing agile
By design, the as-a-service approach increases speed-to-market. With no physical or virtual appliances to deploy and no patching or upgrades, the energy of network practitioners is focused on outcomes that enable the business. This also reduces friction between network, development, and compute teams as new products are planned and rolled out.
The network is where data meets services and is the first step in creating a cohesive strategy. This strategy must meet the business where it is today while also providing future capabilities. One of the best ways to establish a cohesive strategy is by finding a way to normalize consumption and streamline operations. There must be a grounded understanding of the network across clouds, on-premises, applications, and data to bring consistency to the data pipeline.
