Last week New Relic launched a new suite of AI features for its New Relic One observability platform intended to reduce the background chatter of events allowing DevOps and SRE teams to detect and resolve incidents more quickly.
On-call engineers are bombarded with occasionally contradictory messages from an increasingly broad array of monitoring tools, and the new AIOps features are designed to cut through this noise and present the information in a more orderly fashion with added intelligence and context, the company says.
New Relic AI integrates with incident management tools such as Pager Duty, ServiceNow and VictorOps, feeding intelligence about incidents back into those tools so they can be grouped into contextual categories or sorted in order of urgency or some other metric. While many of those tools have their own AIOps features, director of product marketing, Michel Olson insists that the New Relic offering casts its net much wider.
"We give our customers the ability to easily ingest incidents, alerts and events from any source regardless of whether that's through New Relic or tools like Pager Duty or ServiceNow, or even other monitoring tools like Splunk or Prometheus or Grafana," he told Computing.
"We discover relationships across the alert and incident data that you've ingested into the system in order to produce a single more actionable alert that customers can then take action on."
Other capabilities that set the solution apart, according to Olson, are proactive anomaly detection that occurs earlier in the incident response process reducing the chance that small problems will grow into full-blown incidents, and there's also an emphasis on tunability and control.
"We're focused on providing full transparency into why and how correlations are performed, as well as enabling our customers to tune the system with their own decision logic by simply telling the system what data to compare and how they want to correlate that data," said Olson.
The AIOps suite is available as an optional extra for users of the New Relic One platform.
When the term first emerged around 2018, AIOps met with some scepticism. After all, the prefix AI was being attached to all and sundry, often with no real machine intelligence in evidence. However, industry insiders now see it as a genuine development, a set of features that can learn and respond in-real time in a way that was not possible before.
Roy Illsley, a distinguished analyst at Omdia, said the term AIOps has been used to cover a broad range of approaches by vendors in different parts of the operational management space.
"Some solutions are just using AI/ML to provide automation or deeper insights, but a few vendors are now beginning to use it as a connection between different teams so that operations are becoming more collaborative and efficient," he said.
Olson said the market is now using the term more consistently, and that AI-based ops systems really are starting to make on-call engineers' lives easier.
"AIOps uses AI and machine learning to analyse all the data that's generated by software systems in order to be able to predict possible problems, determine root causes and then drive automation to fix them. I think it's a good description and the term resonates with our customers."
Managing storage for large, complex distributed containerised applications is a whole different game; industry insiders discuss how the options stack up
Rancher has big - and small - plans for the container orchestration platform
Testers have to be destroyers, argues Mark Wilson. But too often testers aren't trying too hard to bust the software. Rather, they are trying to show it working
Design is all about applying the fundamental principles of engineering, believes Mark Wilson, in the latest in his ten-part series