Seven pillars for constructing a modern data centre network

In order to successfully transform the network, businesses must be prepared to ask challenging questions that drive conversations around open networking, automation, modularity, scalability, segmentation and re-usability. Before moving forward, it is essential that organisations consider the following list of business and technical guiding principles:

1.       The network architecture should use standards-based protocols and services: Over the past few years, adoption of open source technology has increased significantly, as more organisations discover its considerable advantages which extend far beyond low costs. While proprietary protocols and closed ecosystems require highly specialised engineers, limit inter-operability, and force organisations into particular designs that are difficult to escape, standards-based protocols promote interoperability, competition and innovation.

2.       The network should be serviceable without downtime: It goes without saying that fault tolerance is a must. Service outages are always a risk and can occur for any reason, to any type or organisation, leading to financial and reputational damage. For example, a 2017 AWS outage cost publicly traded companies $150 million dollars, and the recent Google cloud outage generated negative headlines around the world.

To prevent outages, all compute nodes must be dual-connected to redundant upstream Leaf switches. Leaf switches should have redundant peer-link connections between each other, and to each Spine switch. Equal-cost multi-pathing ensures that all paths are active and forwarding. Inserting or removing a Leaf or Spine switch should not affect production traffic.

3.       The network architecture should promote automation: Manual configuration changes are time-consuming and prone to human error. When designing or monitoring a network, it’s important to ensure that it’s running as intended and adheres to set network and security policies. Automating tasks can make the network self-healing, more consumable, and easier to audit. Familiar Linux APIs allows DevOps engineers to integrate the network into automation engines without the friction of dealing with numerous, vendor-specific APIs. Having the same network operating system (NOS) on each device, regardless of the underlying hardware, opens the door for simplified network automation.

4.       The network should be consumable: Tied into automation is the concept of consumable self-service networks. Whether the data centre is private and serving a single organisation, or built for a busy IaaS platform, having the capability to empower administrators or customers with self-deployable networks should be a key consideration with new network designs. Creating networks in the public cloud is a fundamental feature everyone expects.

Customers should have the capability to deploy segmented networks on the fly, without the intervention of network engineers. A Linux NOS is ideal for orchestration solutions, due to native Linux modules and APIs. Deployments that harness EVPN with automation facilitate the deployment of new networks while simultaneously enabling customers to build their own on the fly.

5.       Physical boundaries should not restrict segmentation capabilities: Modular portability is critical when thinking about network design. Organisations can use EVPN to compartmentalise and segment tenant traffic across the data centre, providing an open and flexible architecture irrespective of physical boundaries, transporting network segments anywhere in the data centre or across data centres.

6.       The network must be scalable: A Leaf-Spine Clos architecture is ideal for data centres; with equal-cost multipathing of 128 links, Leaf-Spine pods can become massive. Additional pods can be added to grow horizontally, or new tiers to grow vertically, interconnecting indefinite numbers of pods. EVPN scales with the physical topology, providing the ultimate modularity for scale. If port-density or port-speeds in specific areas become insufficient, a disaggregated model allows data centre admins to swap hardware modularly, automating the NOS and network provisioning with ONIE, proving flexibility at the micro and macro scale.

7.       Network changes should be verifiably testable before implementation: Downtime and SLA violations can cost organisations significant dollars in the form of refunds or reputation. Organisations can reduce the risk of downtime by fully simulating network changes and upgrades before flipping the switch and making them live, assuring that simulated tested network changes will be successful on systems in production.

 

Modern IT demands automation, scalability and agility. The implications for businesses now are not just technological support but economical as well. An inflexible network becomes expensive to scale at the speed of customer expectations and business innovation. Business innovation puts pressure on data centres to offer extensive automation of the entire network life cycle, from provisioning and deployment to day-to-day management and upgrades.

When designing their next data centre network, organisations should carry the above guiding principles with them from project inception through to network deployment. While the list is far from all-encompassing, these ideas will help generate specific results for a highly effective and agile data centre, built to scale, and designed to lead.