Sumo Logic has developed “bidirectional” integration for Atlassian’s Opsgenie incident alert service and promised similar deals with other partners.
The machine data firm said the integration, which it launched at Atlassian’s Las Vegas customer event, meant that customers using both products can create automated alerts in the Aussie giant’s incident management platform based on machine data analytics from Sumo Logic. At the same time, they can use Sumo Logic to analyse Opsgenie alerts and incident data and their impact on their application stack and cloud infrastructure.
John Coyle, vp of corp development at Sumo Logic, said the firm already had an integration with Atlassian’s core product, Jira adding that “on the dev side, Jira is kind of the heartbeat. If Jira is down, dev comes to a stop.”
“What’s unique about this integration is it’s birectional,” he continued. “We’re monitoring a bunch of infrastructure applications for a customer based on criteria the customer sets out. Sumo triggers an alert, we can kick off how to address that with Opsgenie.”
“At the same time, we’re ingesting logs from Opsgenie to allow customers to get visibility of and manage and monitor all the incidents that Opsgenie creates, and analyse how well they’re responding to them.”
Given the trend towards open APIs, spurting data in both directions is not beyond either firms’ more technically adept customers, Coyle admitted.
“We think we all in the industry need to be more prescriptive to make it easier for those customers who may not be as sophisticated, or as is more commonly the case, just don’t have that manpower from a resource perspective.”
He added, “You’re going to see us do a lot more around those types of integrations.”
“The next step of what we’ll be doing is beginning to understand the workflow between Jira and Opsgenie, and provide some correlations where appropriate.”
Inevitably, this means applying machine learning to the data to head off incidents before they happen, Coyle said.
“Who wins is who can get to be predictive based on seeing all the data and all the things are happening, then gets to a point of not just saying something happened and alerting very quickly, but getting to the point where seeing stuff that develops before any incident happens.”