Sumo Logic today announced a range of new features in its cloud-based analytics platform that should improve its customers’ capability to glean operational and security insights from machine data emanating from modern applications.
Sumo Logic previously offered support for applications running in Docker containers and orchestrated using Kubernetes. Today at DockerCon 2018, Sumo Logic announced it’s bolstering support for those architectures with its machine data analytics platform.
Specifically, Sumo Logic announced that it’s offering native support for Amazon Elastic Container Service for Kubernetes (Amazon EKS), a new managed service offered by AWS. The company is also going to store machine data using Prometheus, an open standard used by the Kubernetes community.
The improved Kubernetes and Docker support helps to “close the gap” that exists in monitoring container-based workloads, says Ben Newton, a product marketing manager with the Redwood City, California company.
“As customers are moving to these new technologies in the cloud, they’re struggling to wrap their head around it,” Newton tells Datanami. “Kubernetes has really been a big thing for just a couple years. There’s a lot of open questions that customers have about the best way to monitor that and use it. That’s really where we can help customer and why we’re investing so much there.”
Sumo Logic is also letting customers convert existing Graphite-formatted performance metrics into a new format that’s richer and offers more flexibility going forward. That should improve their ability to take advantage of the latest machine data analytics capabilities without requiring invasive and expensive code changes, Newton says.
“When customers are moving to the cloud, it’s very hard to just go back and re-do everything to use the best thing possible that’s available,” he says. “Customers should not have to jump through hoops to use an analytics application. They should be able to send the data once, then we help them with the transformation so they get the most value out of it that’s possible.”
The third major announcement from Sumo Logic today revolves around how the company prepares unstructured machine data for analysis. Unstructured log data can be unwieldly to analyze, but thanks to a new optimization technique, Sumo Logic says it push log data through a new time series analytics engine and get answers into customers’ hands 10 to 100 times faster than before.
“We actually allow customers to convert these more unstructured log format that are not necessarily conducive to high performance analytics, and convert that to very high-performance time-series analytics,” Newton says.
Newton says this is about the move from big data to fast data. “You basically have to interact with customers on a constant basis,” he says. “You can potentially lose customer in minutes if you don’t take care of things very quickly. Customers want real time access to data and they want very fast analytics. You can’t afford to have these old big data infrastructure that take days or weeks to deliver the answer.”
Sumo Logic has emerged as a force to be reckoned with in the machine data analytics world. The eight-year-old company, which has raised $230 million in funding and is said to be on track for an IPO, ostensibly sells analytic solutions that help security and IT operations teams better track IT infrastructure and applications. But the company is finding that’s just the beginning.
“The two main use cases are operations and security. That’s where most of our customers use us,” Newton says. “What we find is, once customers they have started with a core use cases, we find they’re extending the value beyond core operations and security use cases.”
Other groups signing up for Sumo Logic subscriptions include customer success teams and “non-traditional” IT groups. It’s not uncommon for Sumo Logic customers to have several dashboards running continuously on big screens to help employees stay on top of what their customers are doing with their applications.
“It’s a natural progression because the reality is that machine data does not just fit into that one use case,” Newton explains. “Machine data has lots of information, both about how your application and infrastructure are working, but very often it has info about what users are doing, what customer are doing. You tend to be able to leverage that data across multiple use cases and multiple different teams.”