If you’re in DevOps you are well aware of the constant pressure to deliver fast – and then even faster. And while the ‘need for speed’ is a key driver of DevOps adoption, continually picking up the pace in the endless ocean of complexity can often seem impossible. This can be a major cause of job frustration, but in a nascent discipline such as DevOps, frustration can also be a powerful catalyst for progress.
When I first started out in the world of software and IT, one of the things that struck me most was the mounting impression that so many professionals across the industry are going to work every day feeling absolutely frustrated. And these were people with really good jobs, good professions that they chose themselves and mostly felt very passionate about. In time I learned that success as a developer often meant the facility to avoid failure; putting out the fires successfully was considered a good day. And then ‘one day’, DevOps was born and it seemed like there was light at the end of the tunnel for all these frustrated IT people; someone was finally taking control. But I was wrong!
Then came reality. Today it feels like although DevOps is undoubtedly “transforming the transformation“, I’m hearing lots of frustration from the DevOps world, now equally pressured – and challenged – to deliver on the promise of speed. Paralleling this is an endless proliferation of tools and tips being thrown at the DevOps world, while we all try to get our work done. So how do we accelerate the accelerator? In one word: Empowerment, in about 200 more, here are some things that will help:
1 ) Work with Smarter Automation
Ever heard of “work smarter not harder”? Well, you may not know that the phrase, despite having a sexy contemporary ring to it (start-up culture uses this to death), was actually coined in the 1930’s by Allan F. Mogensen, the creator of Work Simplification. Now, although simplification is a very simple answer to complexity, it holds endless value for the software world wherever achieved.
So while DevOps contributes to simplification there will always be a need to go further in the quest to be a company that moves fast. Because work simplification is the foundation of the systems concept, DevOps are embracing machine learning to simplify the process and shorten the time it takes to perform them. Also known as automation, giving the machine simple tasks to do is saving teams more and more time. While machines are becoming smarter and AI capabilities are increasingly added into the mix, the range of tasks that can be automated is growing. DevOps are successfully employing smart automation to reduce the time and resources spent on mundane tasks, but now that machines can learn to assimilate human thought, engineers can work with the system to teach it what to react to and how throughout the enterprises’ entire environment. Of course, once you reallocate manual tasks your team members are freed to use their skills more effectively and productivity gets another boost.
2 ) Introduce Uber-visibility
The only other term you hear more in IT than “complexity” is NOISE. Every industry has its noise, but in the IT industry, we call it DATA. The good news is that machines have learned to ingest data from numerous sources, the bad news is that there are still so many DevOps people, IT managers, operations rock stars and software programmers that are unable to view their data using one dashboard. Even more frustrating is the fact that many are still unequipped to view any data at all from various legacy, user facing or internal systems. This makes developing, testing, iterating, delivering, fixing -and almost anything – all the more complex, challenging and time-consuming.
More good news? An AI-powered system today knows how to ingest, process and display your data so that you can decide what to do with it. It will alert you to any activity or event throughout the entire environment. This is a huge disruptor for the software community, now able to fathom a proactive approach to any size environment. For DevOps being able to monitor during production, in real time and across the stack, is a time-saver to say the least.
3 ) Take Proactive a Step Further
Smart machines can perform data analysis on an unlimited number of logs – in an instant. So the 360-degree view of the environment also comes with a constant real-time analysis that is used to fix problems before they occur. AI machines enable a 360 view, they work with DevOps to establish a unique baseline and behavioral pattern over time so that anomalies are detected earlier and many failures can be altogether avoided. In the service of DevOps, smart machines use anomaly detection algorithms to enable a more fail-proof process and a better performing business.
Perhaps the most proactive capability DevOps are acquiring is the ability to instantly perform and act upon root cause analysis. The market is rapidly moving towards automated intelligence in the form of root cause analysis because of its ability to save resources and time traditionally used to manually conduct analysis – post event. Having the analysis performed constantly in the background is so much more than a time saver — but to use the understatement of the year – it saves time.
To sum up, I would definitely say that as a discipline that was created to enable the world to embrace technological progress, DevOps are most definitely setting an example and using progressive technology to reduce complexity, maximize resources and achieve superior results. I predict a huge alleviation of a great deal of professional frustration over the next few years as DevOps continues to embrace their new robotic partners in their quest to digitalize the world.
Disclaimer: I am definitely not pretending to be neutral here by not mentioning that the said ‘smart machine’ I am referring to is actually my company’s fantastic product – LOOM SYSTEMS – that I firmly believe in. I just don’t like anyone to think I am hard selling anything other than my beliefs.
I would love your input, questions, support on this or any of my posts and we are always here for questions.