There’s still a cultural difference between devs and ops, says OpenStack User Committee member Saverio Proto. Now’s the time to improve it…

image

There are five members on the OpenStack’s User Committee who strive to provide oversight and guidance to working groups that target specific areas for improvement.

Here we talk to one of recently elected members of the UC, Saverio Proto, about cultural differences between dev and ops, being part of a Stacker family and deployment. Proto is a cloud engineer at SWITCH, a national research and education network in Switzerland, which runs a public cloud for national universities.

How did you get involved with OpenStack?

At my previous job, I was a network engineer. The company planned to migrate the data center to a big OpenStack cloud. Helping out the IT operations folks to design the network for the cloud data center, I ended up learning the whole thing. I liked it so much that I moved to a new job where OpenStack is the core business.

What are some of the most pressing issues facing OpenStack users now? In the next year?

OpenStack is always evolving. While this evolution is a positive thing that drives innovation and business, it’s a challenge for existing deployments to be upgraded and maintained. There is already an open discussion if the current release cycle model fits the needs of the users. Deployments often rely also on deployment tools like Puppet modules or Ansible playbooks. When these tools get deprecated in favor of new deployments methods, the issue is what to do with the existing production deployment. I think the priority for the next few years is to find a strategy to support deployments in long term of production operations, so new users will feel that OpenStack is the right tool to start a long-term project.

What’s the role of research entities like SWITCH in the OpenStack community?

OpenStack is important to entities like SWITCH, because there’s a need to build a cloud for researchers that continuously evolves following the changes of workloads of its users. In our OpenStack experience, we deployed heterogeneous hardware, we experimented with compute nodes with GPUs, we have run multiple storage back ends at the same time and we deployed complex network setups.
OpenStack benefits from the research users because all our experience running the platform is shared with the wider community. Pushing OpenStack to the limit, makes possible to identify scalability issues. I think the best example is looking at the experience of CERN — the contributions arrived from their deployment to the community is massive.

Your wife is also a Stacker — what’s that like?

I have to say that having my wife working with OpenStack helped both of us learning the platform quickly. This is always true when there is exchange of information among OpenStack operators who work in different teams. It just helps to chat with someone who works in a different company using OpenStack,  you gain an immediate win-win learning process.
But don’t worry, to run an OpenStack cloud you don’t need to get your whole family involved! We have great community events like the OpenStack Operators Meetup, the Project Teams Gathering (PTG) and the Summit, where anyone can exchange ideas and learn from the others. I also think it’s very useful to join your local OpenStack User group.

Anything else you’d like to add?

My experience as an OpenStack operator is that there is still cultural difference between devs and ops. We see this difference in the release cycle, in the upgrade procedures and how things are tested before going to production. We can still improve in terms of working together in the OpenStack community. I will do my best during this year at the UC to improve communication between developers and operators.