Why IT Operations Needs a Jason Bourne-Style Reboot
In the age of mobile, cloud and microservices, here's how IT pros maintain Bourne-style awareness.
There is a great scene in the first Bourne movie. The hero is trying to reconcile not knowing who he is with knowing lots of crazy stuff about his surroundings.
“I can tell you the license numbers of all six cars outside. I can tell you that our waitress is left-handed and the guy sitting up at the counter weighs two hundred fifteen pounds and knows how to handle himself…. Now, why would I know that?”
Well that’s simple, Jason―you have situational awareness. It’s a mandatory skill for covert-ops specialists, but it’s also a requirement for IT operations, albeit in a slightly less kick-ass context.
The times they are a-changing and so is IT operations.
Between the third Bourne movie―which was released in 2007―and this year’s reboot (simply titled Jason Bourne), much has changed for IT operations. Monolithic application designs are making way for microservices, cloud has become an accepted platform for deployment and mobile apps are the force de jour for customer engagement at scale.
So what’s new? Doesn’t managing technology change come with the IT operations territory?
Well, yes and no.
IT operations has always had a ton of new tech to manage―from big-iron to thin-clients. New tech, different day, so to speak. But what’s different now isn’t just the technology newness, it’s the fact that digital imperatives have become the impetus by which other teams can make operational calls.
So, if an agile development team needs the high availability of a cloud database like Cassandra for a critical data analytics project, that decision doesn’t necessarily need operational blessing. Similarly, if testing can’t wait for the manual provisioning of a lab or new microservices need a platform to support sudden changes in demand for scale, then acquiring public cloud services is a no-brainer.
Life is rough for IT operations. On the one hand they’ve lost control, while on the other they’re still expected to ensure stability for whatever comes over the wall. The added twist is that cloud delivery also removes the separation between function and operation. So, while customers expect a continuous stream of change, they also demand high-quality experience.
The applications supported by operations are also changing radically. Single platform architectures will make way for composite applications and hybrid clouds. The problem won’t be managing another technology, it’ll be determining how to maintain performance for applications composed of hundreds (even thousands) of always-changing services, operating on an always-changing platform.
How should IT operations respond? How must monitoring evolve? Well, again, that’s simple―IT operations must exhibit Jason Bourne–like qualities when managing the inevitable influx of cloud applications and infrastructure. That means increasing agility with situational awareness.
Situational Awareness―DevOps Style
Simply put, situational awareness means knowing what’s going around you. Now, while that’s second nature for government-trained assassins, it’s not so easily acquired in an IT Ops context. Of course, many would argue this just involves maintaining a watchful eye over the health and performance of applications, but is this enough?
In the context of today’s massively complex hybrid cloud-based systems, IT Ops professionals don’t always know what to look for with regard to performance anomalies. Also, worryingly, they’re not always sure they’re paying attention to the right signals or whether innocuous application behaviors are warning signs of bigger issues.
Old-style, rules-based application performance monitoring technologies can’t address these challenges. With mobile apps, microservices and containerized environments, the volume of alarms grows exponentially. Mix this with the fluid nature of hybrid cloud and finding root causes through traditional approaches becomes, well…mission impossible.
To address this, operations must adopt advanced monitoring methods to increase situational awareness. Like Bourne, these solutions will exhibit two essential behaviors:
Observing: In cloud environments, modern monitoring solutions will observe as much relevant information as is practical (like Bourne checking all the sightlines when he enters an unfamiliar building). But it’s one thing being able to collect events and alarms at scale, it’s another being able to process them all. In elastic environments, standard approaches are too time-consuming and inflexible, so the focus must shift toward automated analytical methods for ensuring service quality.
Orienting: Diagnostics only become useful when put into context. Solutions will move toward serving valuable information to a variety of stakeholders, including developers, architects and business users. To achieve this, best-guess baselining will be replaced by statistical techniques that detect anomalies. Then, by presenting results from the perspective of a role or required business outcome, information becomes immediately actionable. That’s incredibly helpful for addressing a range of problems, but also becomes the DevOps bedrock for feedback and continuous improvement.
A digital business can’t succeed with IT Ops suffering an identity crisis. Carry on that way and this critical function will become further isolated, even ignored. Smart practitioners recognize that managing new applications demands levels of situational savvy well beyond the scope of traditional monitoring tools and methods.
Kind of like in the trailer for Jason Bourne, they know the name: Agile Operations.