Author: Kaylee Mae Tucker (OST Account Manager)
I often find myself trying hard to understand Client A’s definition of DevOps versus the definition I heard from Client B the day prior. As is the nature of technology, DevOps seems to be constantly changing in concept but also quite different from technologist to technologist.
In an effort to greater understand this part of the world I service, I thought I’d interview one of my favorite developers, Matthew Mousseau. He’s a strong developer who by happenstance and curiosity landed himself in the DevOps world.
Meet Matthew Mousseau:
“Matthew Mousseau is a software developer passionate about learning. After moving to southern California, he started Continio Software Solutions in 2015 to apply those passions and help businesses expect more from their applications and data. When he isn’t nose deep in code, he enjoys deep sea fishing and reading a good book.”
Kaylee: How did you find Continuous Integration/DevOps?
Matthew: I thought it would be great if my code could copy itself so I looked for tools/technologies to do this and stumbled into CI/DevOps. Continuous Integration and DevOps is a natural consequence of codifying procedures, policies, ideas, instructions for a computer to follow versus instructions for humans to follow – it’s more reliable.
Kaylee: How can someone identify a DevOps environment?
Matthew: The most identifying attribute of DevOps is when the organization’s operations team is involved in the planning of systems. It’s not separate teams but more of a continuum working together actively. Working to understand each other. But this can be disruptive to an established work flow. Introducing DevOps to a company can bring a certain amount of mistrust. It’s introducing automation to something that wasn’t automated before. This is scary to an operations person.
Kaylee: How did you develop an interest in this?
Matthew: I had a desire to get to the end result faster, to make a change faster, to get feedback faster. A huge benefit is the ability to deliver software more regularly. You can react to new ideas and features faster.
Kaylee: Do you think DevOps or Continuous Integration environments required an Agile methodology or work flow versus traditional Waterfall?
Matthew: It doesn’t necessarily require it. By nature, Waterfall methodology doesn’t seem to “need” DevOps or Continuous Integration as it is inherently a more structured, step by step process – and much slower. So Agile lends itself to a DevOps environment.
In Matthew’s current project, he’s helping the organization take baby steps towards a more integrative approach to system/software development and infrastructure. He started by categorizing the applications so it’s easier to standardize the types of questions required by the ops team to be answered by the dev team. Seems like the first step towards an integrated DevOps to me!
They are also building a “system in the middle of all systems and teams” that basically coordinates the infrastructure. They are using Bamboo, Puppet, and some customized software that is similar to Amazon Web Services. Matthew truly believes for an effective DevOps to exist; the environment cannot be pigeonholed into one tool or technology. He’s a perfect example of this – he’s a multifaceted developer, infrastructure manager, scripter, and “DevOps implementer”.
Matthew mentioned an unexpected benefit is the ability to ensure great security across the business. DevOps and CI truly impact this area. Traditional development and deployment models cannot respond quickly enough to the security demands our cyber world is facing. But integrating and seamlessly streamlining infrastructure and software can help organizations to quickly respond to security threats and ensure ongoing security is maintained.
He said the three most important areas of assessment are: Application Health, Operations Logging, and Metrics. The health of an organizations’ software or application is obviously imperative but takes on new meaning when associated with DevOps. Some areas of “application health” are the ability to connect to the applications resources (databases, etc.) and the storage of the application. Operations Logging basically measures meaningful history; it’s a historical log or description of what’s going on in the environment. Metrics are metrics – the data, the stats, the information. To Matthew, these three specific areas seemed to tie the entire DevOps world together.
Matthew went on to say, DevOps tends to change the way an organization thinks about software. It changes the approach. And ultimately results in the ability to deliver efficiently, learn quickly and proactively see problems.