Given the frequency with which IT discusses DevOps trends and developments, one could believe that the community has universally adopted its tenets and lessons. However, recent studies show that less than 5% of teams have adopted the precepts of DevOps exclusively and just over 50% said a quarter to half of their development groups have adopted DevOps.
Clearly, there is plenty of room at the DevOps Inn for more people to come aboard. There are however several questions to ask. Specifically, as companies look to begin adoption of DevOps, what principles and best practices should they look at and which should they avoid?
To that end, this blog will look at:
The goal of DevOps is to remove bottlenecks, increase automation, reduce mistakes and recover faster from system failures. Additional top drivers of adoption include the need to increase quality, improve the customer experience, reduce complexity, and reduce overall IT costs. Teams want to adopt DevOps because they see it as highly correlated with higher throughput and better stability.
These lessons can be applied to software releases and deployment as well as to writing code for infrastructure and creating new environments. Research has shown that whether you’re deploying commercial off-the-shelf software (COTS) or microservices in the cloud, you can attain high performance with DevOps practices. DevOps looks to increase staff collaboration and increase the sharing of responsibilities around application definition, development, deployment, and support.
Those organizations that have seen significant adoption of DevOps in their development and operations environment have seen more frequent deployments, faster lead time changes, faster time to recovery from failures. The following chart shows this in detail:
Today, organizations rely on DevOps because it produces better software and general IT outcomes with fewer errors and more reliability. Teams see these outcomes through the lens of throughput and stability of systems. Throughput measure how quickly teams can deploy code and stability measures how quickly teams recover from failure and downtime.
To recover effectively from downtime, DevOps teams need to know how to handle events when things go wrong. While a large percentage of surveyed respondents said they are first alerted to failures via monitoring and alerting systems already in place, a whopping 59% discover them via irate customers. Indeed, a significant test of the effectiveness of DevOps is when do the teams find out about downtime and failures. Do they find out about them or do the customers find the failures and report them? A successful DevOps practice should not only help eliminate failures, but enable IT to discover and remedy failures before users do.
To make sure that the proper incident management protocols are adopted, teams need to adopt a strong incident alert management platform. Proper incident alert management platforms will consolidate alerts onto one platform and provide responders and stakeholders with secure messaging, ticket updates and status reports. OnPage has an extremely strong and prominent application platform which meets these goals.
DevOps Trends to Avoid
With the goal of saving money in mind and hopes of increasing the speed of automations, some DevOps operations try to create in-house scripts, monitoring tools and alerting applications. DIY can also result from some developers simply wanting to ‘do it on their own.’ DIY can be seen as the result of a lack of maturity in automation tools or it may simply be the culture of IT staff overall. Regardless, most DIY efforts runs the risk of proprietary, error-prone solutions.
When DIY tools are not thoroughly vetted and don’t provide proper monitoring and alerting components, they create more problems than they solve. Proper monitoring and alerting cannot be an afterthought because both are critical to the proper functioning of any stack and any DevOps team.
There are no shortcuts to achieving an effective DevOps implementation. Instead, teams need to place equal importance on the implementation of testing, monitoring and incident management.
To learn more about the State of DevOps and best practices, download our whitepaper.