BlueJeans achieved operational efficiency and reduced their TCO by using NVIDIA Cumulus Linux and Ansible for automation
Founded in November 2009, BlueJeans Network offers cloud-based video communications and is dedicated to making it easy. Starting from the ground up, BlueJeans built a video service that enables users to connect with each other any time, any place and from any device. Bluejeans believes that every phone call should be a video call and they are designing the technology to make that happen.
With over five data centers around the world, BlueJeans is helping customers like Facebook, Intuit and ADP encourage employee collaboration and boost productivity across their organizations.
As a cloud-based solution on which some of the largest global organizations rely, BlueJeans needed a solution that was reliable and scalable. With rapid business growth, the team wanted to reduce their Layer 2 complexities and adopt a network topology that was easy to design, manage and scale. Network outages at BlueJeans meant that their customers’ business communications would be hampered; hence network uptime was critical to the team.
With plans to expand their data center footprint across the world, BlueJeans wanted the ability to quickly replicate the environment without being locked into a particular vendor.
While cost was important, the team wanted a solution that would reduce the amount of time spent on troubleshooting and maintenance, and more time on supporting business critical applications.
To reduce Layer 2 complexities, BlueJeans adopted Layer 3 Clos architecture with NVIDIAⓇ Cumulus Linux. The solution allowed BlueJeans to easily scale for increasing business growth by making it easy to add additional server racks due to the simplicity of the spine-leaf architecture. Additionally, using a standardized configuration across all ToRs, built an easy-to-operate infrastructure as it allowed for easy automation. As each switch is cabled the same, they were able to leverage the same Ansible templates to provision every switch. Finally, using ECMP, MLAG and dual connections to every host provides a reliable network infrastructure capable of surviving unit failure. As a disaggregated model, this also offered the team the ability to gain supply chain flexibility while maintaining a standard solution across all their data centers.
The main criteria for selecting NVIDIA Cumulus Linux included:
- Supply chain flexibility: NVIDIA Cumulus Linux supports over 50 hardware platforms from eight partners. By building intelligence into the software layer, BlueJeans could pick any ONIE compatible hardware platform and run the same OS image across multiple data centers.
- Scalability: Growing from a few racks to over five data centers in under 2 years is an impressive feat for the BlueJeans team. UsingNVIDIA Cumulus Linux and automation through Ansible made that growth uncomplicated. Due to the native automation support of NVIDIA Cumulus Linux, BlueJeans was able to leverage automation scripts to quickly build out data centers and make fabric wide changes with little impact.
- Operational efficiency: BlueJeans was able to unify their data center by adopting standardized topography and model across the stack and by standardizing on common tools.
- Leveraging automation with NVIDIA Cumulus Linux, the team was able to provision new data centers in under 30 minutes, compared to the 2 days that it would take if using traditional vendors.
- OSPF unnumbered allowed ease of implementation by reducing the amount of configuration required to establish the networking fabric.
- Using the NVIDIA Cumulus Linux operating system meant Bluejeans could standardize their entire data center, from compute to network, with one standard topography and model. This resulted in a completely unified infrastructure that’s simple to configure, manage and monitor. This was an unprecedented benefit they couldn’t get with a traditional vendor.
- By leveraging standard tools across the stack, the team was able to triple their network operator to switch ratio and allowed the team to focus on business critical applications.
- Lower TCO: BlueJeans experienced lower CapEx by deploying a disaggregated solution, and they experienced dramatic savings on the OpEx front. The OpEx savings resulted from:
- Quicker time to production (deployed up to 95% quicker)
- Optimizing available team (3x operator to switch ratio)
- Leveraging common automation and monitoring tools across the stack
- Optimizing switches by picking the optimum vendor for the deployment
BlueJeans adopted a simple and standard network solution that helped them to achieve scalability and supply chain flexibility. With NVIDIA Cumulus Linux and Ansible for automation, the team achieved operational efficiency and reduced their TCO.