AIOps, which is the practice of employing artificial intelligence to manage IT operations, has become quite the buzzword lately, which Gartner attributes to operations teams being snowed under with the volume, variety and velocity of data being generated by their systems.
It describes four phases of IT-operations-oriented machine learning: descriptive, diagnostic, proactive capabilities and root cause analysis.
While few vendors offer comprehensive, integrated AIOps platforms to date, it says, it predicts that within five years, wide-scope AIOps platforms will become the norm, rather than AIOps functionality embedded in a monitoring tool like application performance management, network performance management or IT infrastructure management tools.
Count OpsRamp among those taking a wider view of AIOps.
This week, the company unveiled enhancements to its OpsQ event management and intelligent correlation machine learning models, as well as new multicloud infrastructure monitoring capabilities for Amazon Web Services and Google Cloud Platform.
“You can look at it as a Big Data platform for operations management,” said Bhanu Singh, OpsRamp senior vice president of engineering and DevOps.
Some vendors are focused on one aspect of IT operations — a discovery problem or a monitoring problem — OpsRamp takes a more holistic view encompassing both cloud and on-prem infrastructure operations.
Singh told The New Stack previously: “AIOps is providing access to tools that can bring data, data analytics and machine intelligence together to make advanced decisions and perform automated actions by collecting and analyzing data.”
OpsRamp provides discoverability and observability of what resources you have — routers, switches, storage — or if you’re running things on the cloud — your EC2 instance, RDS database, IaaS on Azure, and makes sure you have one single view, he said.
Wading Through Noise
The San Jose-based company was founded in 2014 by Raju Chekuri and Varma Kunaparaju, whose backgrounds are in service management. In 2017, it raised 2017, $20 million in funding from Sapphire Ventures. Its customers include cloud services vendor GreenPages, advertising and marketing firm Epsilon, and managed service providers Carousel Industries and CloudBrix.
OpsRamp touts native discovery and native monitoring, though customers don’t have to use these SaaS capabilities if they’d prefer to stick with their existing tools for that, Singh said.
While ingesting, aggregating, analyzing data and providing recommendations to customers with on-prem and cloud operations, the company does the same itself, with operations on AWS and Azure as well as its own data centers.
It’s taking a different approach to AiOps, he said.
“[Other vendors say] ‘We collect your data and we put it in a magic box. We shake it up and give you an output on the other end. We have all the algorithms to figure it out.
“We’re not taking that approach,” he said. “We believe AiOps is not in isolation. …There’s always a need for more integrations.
“We do native monitoring. The algorithms of machine learning should be built-in and solving the problems… How do you do incident alerting? How do you do incident classification? Alert correlation? Anomaly detection? Root-cause analysis? … You have all the data.”
It provides a customizable dashboard, which customers can filter views by server, hypervisor, network, business unit, lines of business, location, cloud region, data center, branch office and more.
“We are keeping our eyes on a small group of vendors that came out of the gate with platforms that can deliver a range of monitoring and incident management functions. These vendors, including OpsRamp, FixStream and Centerity, represent a new breed of IT operations management (ITOM) vendor, serving both incident responders as well as senior-level executives,” 451 Research senior analyst Nancy Gohring wrote.
She sees the future as a hybrid for operations and also called Kubernetes monitoring a must-have for AiOps vendors since late 2018. OpsRamp added that capability in its summer release.
She recognized OpsRamp in two categories: infrastructure monitoring and event correlation.
Fellow 451 Research analyst William Fellows pointed out that OpsRamp faces competitors offering point products in monitoring, event analytics and incident management, as well as vendors offering some combination. They include Datadog and SignalFX, recently acquired by Splunk, ScienceLogic and more full-featured offerings from Moogsoft and Big Panda.
OpsRamp’s differentiator is its focus on hybrid environments, Singh said, and ability to cull meaning from the noise of alerts.
The New Stack recently featured a demo in which OnRamp’s AI processed almost 2 million raw events from hybrid infrastructure and boiled that down to 10,000 inferences that were issues that needed to be resolved.
Its survey of 200 IT decision-makers found the four most critical features for AiOps customers were inference models, incident visualization, data-agnostic ingestion, and an integrations ecosystem.
OpsRamp boasts more than 100 integrations with AWS, Azure and Google Cloud Platform. Its fall release adds integration with AWS IoT; AWS Developer tools like CodeCommit, CodeBuild, CodeDeploy, and CodePipeline; real-time discovery for GCP Platform Resources. It has simplified using webhook authentication to build customer integrations.
The new release also brings enhancements to OpsQ, its intelligent event management, alert correlation, and remediation solution. They include new inference models using similarity-based pattern recognition to discover related alerts faster for troubleshooting; fine-grained Observed Mode widgets; and improved resource context ingestion via alerts from other IT operations management tools.
And enhanced synthetic monitoring lets customers pinpoint website outages to specific network routes.