This post is from the XebiaLabs blog and has not been updated since the original publish date.
7 DevOps Monsters and How to Befriend Them
DevOps transformations can be scary – but sometimes, when we face our fears, we realize that these monsters aren’t as intimidating as they seem. Here are 7 DevOps monsters that want to hold you back from your DevOps transformation – AND how to befriend them!
Change is disturbing to dev
After investing years, or even decades, into using beloved tools and following certain processes, nothing makes a developer’s skin crawl more than being forced to change those things. They’ve put on their garlic necklaces, they’ve bathed in holy water, and they are letting in the sunlight to stave off the pesky vampire of change.
How do you befriend this DevOps monster? Management should clearly state the organization’s goals for the DevOps transformation. Based on these goals, instead of arming developers with weapons, arm them with guidance, tools, and education to make the transition as smooth as possible. Let them know the full value they bring to the transformation and how it will help them. And, consider a DevOps platform where developers can keep using the tools and processes they know and love!
Ops is spooked by change
Ops teams see change as risky – and that’s their job – to be creeped out by risk! If ops is offered some strange potion, lured to a gingerbread house in the woods with a bubbling cauldron, or almost convinced by a knobby-fingered, pointy hat-wearing witch to bite into a juicy apple, you can bet by the time you say BOO they will be long gone.
Is it possible to befriend the witch of change? OF COURSE IT IS! Again, make sure you involve the ops team in the transition to the DevOps transformation very early on. Open the doors of collaboration wider, and make sure that they feel empowered to share their ideas on how to improve the way the organization develops, delivers, and maintains applications. The witch has no chance to bamboozle a well-informed ops team!
Hungry zombie culture
The zombie apocalypse is always near, especially in a disembodied, pre-DevOps culture. “Blame zombies” roam freely, looking for people to pull apart and eat up. You’ve probably heard the old standby defense against zombies – simply sever the brainstem – but have you tried EMPATHY?
Imagine this – zombies are really just humans (with a hunger for other humans’ flesh) - have you ever just tried talking to one to learn about how they got to their state of being? That could be the magic key to understanding their perspective, needs, and fears.
In a DevOps culture, there is NO BLAME. It’s all about getting curious about team members from other departments so you can break down silos. Put yourself in their shoes and see where they're coming from. Instead of chainsaws or vaccines, let empathy and collaboration be your arsenal for solving the hungry zombie culture apocalypse.
Straight-jacketed CD pipeline
Occasionally, organizations are so enthusiastic to get their CD pipelines up and running that they can only support a small portion of their applications. Basically, their CD pipeline becomes a half-wrapped mummy – eager organizations can end up creating a siloed environment that only addresses 20 percent, or perhaps half, of their application development stack and only enables a small portion of developers to use the pipeline.
How to befriend the straight-jacketed mummy of CD: step back and take a comprehensive view of all of your development pipelines, build or deploy a platform that will support how all your development teams code software. Don’t build a pipeline that makes teams change how they develop. Atypical enterprise has different toolsets, languages, and processes among various development teams that feed into the primary continuous delivery pipeline.
Now that’s a tightly wrapped, happy mummy.
Stunted CD pipeline
There’s nothing more frightening than a CD pipeline that can’t scale, despite your best efforts. Don’t be the runt of the werewolf pack! Enterprises are more successful when they base their pipelines on a set of standard templates. This enables more effective reporting and the ability to compare various pipelines’ effectiveness, which in turn improves continuous feedback (in the form of happy howls at the moon).
Ghostly management teams
Sometimes, the transition to the light (DevOps) happens naturally from front-line developers. However, it’s important that executive management also crosses into the light. They need to see the actual dollar benefits and understand what’s happening on the ground in order to hang up their sheet once and for all and fully commit to the transformation.
Severed from metrics
Not measuring your progress (failures and successes) throughout the DevOps journey is like living as a headless horseman –carrying around your own pumpkin head certainly doesn’t help you make informed decisions. Be sure to come up with meaningful metrics –things like change failure rate across your delivery pipeline, change lead time, and deployment frequency –so you get a continuous feedback loop about the success of your transformation. Continually monitoring metrics also gives you a chance to improve the pain points in your process, celebrate your successes, and keep everyone’s head on straight.
- Check out a more in-depth explanation of these monsters in the whitepaper called The Seven All-too-common Barriers to Successful DevOps Transformation
- Take on DevOps like a mad scientist with our recorded webinar called The Formula for Enterprise DevOps
- Try the XebiaLabs DevOps Platform for free