关注微信公众号
第一手干货与资讯
加入官方微信群
获取免费技术支持
Location: San Diego, CA
Industry: biotechnology
Company Size: 6,750+
Ilumina is a life sciences company and a market leader in genetic sequencing. The ground-breaking research that comes out of Illumina’s offices drives work on disease, drug reaction, agriculture, and much more.
At Illumina, massive amounts of data are a critical part of the equation. The company has to figure out ways to move data through a pipeline, and how to perform copious amounts of analytics, while also managing all of those administrative tasks that you have as a cutting-edge 21st-century company.
One of the biggest challenges Illumina faced is an environment where custom-compiled codebases were not accessible or fully organized for company and employee use. Container implementation was in itself a challenge: in trying to migrate processes to a new container system, the company found that using a container model took quite a bit of research and staff-hours.
Another challenge was efficiency of management: 10 people had to support 300 developers and petabytes of data.
Through it all, Illumina leaders kept trying to enhance automation and create a better environment in which to run all of these applications. The overall architecture had to support containers running on vendor systems like AWS, as well as bare metal on-premises.
“We needed to make everything self-healing, automated, and really low touch. Rancher made this all happen.” Lou Sacco, Principal IT Engineer
“We needed to make everything self-healing, automated, and really low touch. Rancher made this all happen.”
Lou Sacco, Principal IT Engineer
The company chose Rancher as their Kubernetes platform in order to accelerate their DevOps process. The team at Illumina worked to build a Rancher-based system that works in different environments, for example, using AWS where that makes sense, and putting other components on internal systems on-premise for less cost and more control.
Illumina has also been able to use Rancher functionality toward new goals in deep learning, a much-pursued technological capability. Using Kubernetes as the orchestration layer, teams deployed their first two production machine learning environments in Rancher, where managers have the ability to schedule against GPUs.
The automation department also uses Rancher orchestrated container models to host persistent microservices related to HTML messaging, identity management, and service desk creation.
“We’ve been able to plug a lot of different things into Rancher and just kind of have them go. This has been powerful for us.” Lou Sacco, Principal IT Engineer
“We’ve been able to plug a lot of different things into Rancher and just kind of have them go. This has been powerful for us.”