One-click deployments allow researchers to run the numbers.


As those lab coats crunch vast numbers and loads of data, they’re often doing it with open source. And, like everything in tech, it needs a catchy name: lab-as-a-service.

“Lab-as-a-service is a community resource, it’s an opportunity for developers and open-source users to have access to resources that they might not have. A lot of the stuff we’re working on today as a community with OPNFV, OpenStack and ONAP — and you can’t you can’t really fire those up on your laptop,” says Lincoln Lavoie, senior engineer, Interoperability Lab at the University of New Hampshire (UNH-IOL) in an interview with TelecomTV. “It provides compute resources and networking resources to those users.”

Users login, make a reservation to the resource they need for a set amount of time, then work from remote via VPN. The service is currently powered by 54 servers, both Intel based and ARM-based systems, with a “fairly large” networking setup, a combination of 10 gig or 40 gig links to the servers “so you have plenty of resources.”

In the latest 2.0 version, users can perform more design and provisioning, configuring what the network looks like between multiple servers. “So if I’ve got two or three nodes that I’m provisioning and I want specific layer 2 networks between them, I could do an actual OpenStack deployment.  You can fit all that into what you’re trying to design.”

Lavoie says LaaS is a valuable community resource to make it easier for end users to try out open-source projects. For instance, in the latest version, there’s a one-click virtual deployment button for OpenStack that can run small VNFs on one node, giving new users to OpenStack or OPNVF a chance to take a test run.  Developers, on the other hand, get access to the bare-metal system, including the lights-out management and users can do a lot of low-level checking and low-level installs “it’s a pretty awesome resource,” he adds.

For more on what features are coming next, catch his full interview here.