Changes are afoot in the world of monitoring. Companies will seek alternatives to APM and AI/machine learning will start to be used to deliver better insights.
Once upon a time, IT environments were far less complex. The chief challenge of monitoring tools was to gather data reliably and simply display it. Users were tech-savvy experts on their IT environments. Once they had the data, they knew exactly what to do with it.
It’s a good thing they did because monitoring systems offered little to no interpretation of the findings.
These days, a just-the-data monitoring system won’t fly. IT environments are incredibly complex–and growing more complicated all the time–and IT teams need a lot more help to make sense of them. There’s no way the average user can keep track of everything happening on multiple technology stacks, and gain a deeper understanding of the data on their own.
Monitoring solutions have since stepped up to the fill the void. What’s this mean for companies? And how will monitoring evolve over the next year? Here are my four predictions for what lies ahead this year.
Prediction #1 – Monitoring’s IQ keeps rising
Advanced algorithms and machine learning are welcome additions to the monitoring market. But to help Ops and DevOps engineers make smarter decisions based on the performance and delivery of services and applications, they require precise information. Some might even say better information.
To satisfy this requirement, customers will demand–and vendors will feel pressure to provide–more intelligent systems. Over the next two years, enthusiasm will grow for systems that move beyond simple signals, alerts and anomaly detection toward compound alerts that interpret findings and aid in decision-making.
Assisted by machine learning and Software-as-a-Service (SaaS)-driven tools, companies will gain intelligence of customer patterns. “Opinionated” systems will help companies deliver value directly to users by recognizing problem areas and helping companies to repair them.
Prediction #2 – Monitoring meets machine learning
Tremendous strides with artificial intelligence (AI) and machine learning have been made in signals connected to images, video, and speech. In contrast, similarly, killer AI hasn't cropped up in IT applications because companies aren't certain how to prepare the "right type" of signals and the related feedback that enables machine learning in the domain of IT.
However, this year things will start to change. And within three years, solution providers will figure out the right mix and aggregation of signals and feedback to enable machine learning and will create a breakthrough in monitoring strategies. Ultimately, these tools will increase team efficiencies. Instead of needing to hire experts, companies can operate through generalists, providing considerable customer value.
Prediction #3 – APM isn’t going anywhere but…
Application Performance Management (APM) tools traditionally have helped to monitor online purchasing site performance to make sure visitors have a good experience when making online transactions. These solutions tend to focus on the “front side” of the house, covering web-to-database paths.
APM isn't likely to disappear entirely from the monitoring landscape–5 to 10 percent of monitoring will be APM–but companies will continue to ease off their over-reliance on it. Many organizations have 80 to 90 percent of the code, and intellectual property focused on data-crunching and data-preparation activities that tend to be more back-end, asynchronous workloads and outside the APM footprint. That number will increase as companies engage more in analytics and drive them to seek different kinds of solutions to monitor the ever-expanding data crunching and data preparation side of the house.
Prediction #4 – Solutions emerge to next frontier problems
Many companies have been taking a lift-and-shift approach to cloud. Typical migration vendors move on-premises applications “as-is” to the cloud, hand off the new environment to the company team and then move along to their next customer. This approach has created problematic gaps because not all of the ecosystem–monitoring, for instance–moves with the applications.
I predict that this year, managed professional services and emerging products will address the problem of “the day after” a lift-and-shift problem, namely teams lacking visibility into what’s happening in the cloud and not knowing what to do after the migration is done. This is when they need visibility the most, and managed services and emerging products will address this significant business risk and ensure they get the visibility and the capabilities that they require to continue managing and smoothly take over the newly migrated cloud environment.
Some companies will leverage professional service solutions to replace security, monitoring, and change/cost management as they become hybrid or cloud-centered. Other companies will place a hybrid-supporting ecosystem before the lift-and-shift transition occurs and carry it over into the new environment. Either way, they’ll get around the problem of an ecosystem of on-premises solutions that can’t move to the cloud.
2019 promises to be an interesting year in the world of monitoring. Companies will seek alternatives to APM. Monitoring will take advantage of AI and machine learning to deliver better data insights. All of these predicted trends will mean greater visibility, flexibility, and agility to IT teams.