Like many companies, when Covid-19 hit, the world suddenly felt very big. And the ability to bridge continents, felt too small.
This is the story of how a Tier-1 European Service Provider, 2 DriveNets engineers, 16 Networking White Boxes, and a 1,000+ KM car journey deployed the cutting edge DriveNets Network Cloud, against the odds, during the height of the crisis. It is also a story that illustrates the agility of a disaggregated network solution that can be simply transported, installed and provisioned with the leanest team imaginable.
Deploying a Disaggregated Network
Let’s step back in time to November 2019. DriveNets is finalizing a POC with their European customer. The timeline is carefully planned and a June date is set for the on-site installation.
The team of salespeople, engineers and installers are lined up to travel to Europe in the spring. Then Covid-19 hits. The demands on the network are greater than ever and the customer is keen to test the disaggregated solution that can lower their costs and substantially scale their network. Nobody knows when the situation will improve or how long it will take for the skies to open.
No More Forklifts
Enter Cristian Matei, Delivery Manager and Liviu Bleotu, Tier 2 Customer Support Team Leader. The R&D center in Bucharest had just begun hiring 50 additional engineers after reaching its Q2 2020 recruitment target a quarter early. Sending just 2 members of the European center-of-excellence team seemed like the perfect solution to the predicament. These engineers were free to travel in Europe and were fluent in the technology. Preparations were made to assemble the hardware and ship everything directly to the customer. The Cluster boxes were extremely easy to maneuver and ship with minimal staffing. A huge asset compared to incumbent vendors which require elaborate preparations and a certain size of plane to fly their router chassis. Not to mention, coordinate with a forklift operator and a large support team on the customer site – a big challenge in a time of social distancing.
Meanwhile, Cristian and Liviu packed their bags, filled up their car with gas and left for an early night. At 3am on Sunday the 4th of July they began the 13 hour car journey west. After a good night’s sleep they were ready to meet the white box delivery bright and early on Monday morning at 9am. The equipment had already been provisioned in the DriveNets lab (in Raanana, Israel) and the team went straight to the lab.
The whole process of building the new architecture took just over two days. The hardware itself was light enough for a tiny team to handle and the architecture, with its zero touch deployment capabilities, was simple and extremely fast to set up. The aim was to build a medium cluster that comprised 16 white boxes, able to support up to 96Tbps. The main stages to accomplish were, installation, racking, cabling and software provisioning.
Let’s recap the highlights of the deployment.
- Monday, day one: The engineers quickly got to work in the lab, installing and cabling 16 white boxes with a combination of 100Gig and 400Gig ports. The cabling was all aligned by the end of the day and the DriveNets team and the customer engineers could go out to discuss the work so far.
- Tuesday, day two: With the simple hardware in place it was time to install the software. The cloud-native software runs networking services in independent containers, allowing services like core and edge routing to scale independently from one another. By close of day, DriveNets Network Cloud passed the testing requirements and was installed correctly.
- Wednesday, day three: The final stage in the process was the installation of the medium cluster, followed by the DriveNets Network Orchestrator (DNOR) which ensures the automation of the full solution life cycle and the management of DriveNets Network Operating System (DNOS), a full featured networking stack, ready to run on any Network Cloud certified hardware platform.
- Thursday/Friday, days four and five: The Cluster is fully integrated into the customer’s network topology. This is achieved with no interoperability issues. The system passes a thorough set of sanity checks and functional verifications to ensure full functionality and interoperability
The work was done and the team awaited a final team call after the weekend. The consensus was that the verification process had been successful and the scope was achieved right on schedule.
Disaggregated Network that can Scale and Make Customers Happy
Beyond the technical achievement, another highlight of the visit was the excellent relationships built on site. The customer had originally planned to just send an engineer as a point of contact, but there was such growing interest in the cutting edge system that other engineers at the service provider were drawn out to the site to see the Network Cloud system for themselves. As the POC came together, it generated buzz and excitement. The customer’s engineers were enthusiastic to start playing with the technology.
In the following weeks since, the timeline progressed well. Further use cases were tested, with multiple configurations and functionalities explored. All were successful. The customer feels confident about the solution and the DriveNets staff can easily answer any questions remotely. Now, the cluster is being integrated into the main network and technical protocols are being run.
The beauty of the DriveNets Network Cloud has always been the agility, flexibility and scalability of the solution. This deployment was carried out under the strangest of circumstances and succeeded because of these essential characteristics. Seeing really is believing.
A Report by AvidThink
Network, Meet Cloud - Scaling Disaggregated Network Infrastructure using Cloud Principles