Dynatrace makes SRE Agent generally available for incident triage
Dynatrace has made its SRE Agent generally available, giving teams an automated way to investigate problems, assess their severity and route incidents as soon as an alert fires. The agent provides root cause context, evidence, impact analysis and remediation recommendations in the tools teams already use.
- Dynatrace’s SRE Agent is now generally available.
- It investigates problems, assesses impact and recommends remediation within minutes.
- Workflows can route findings to teams through Slack, Microsoft Teams, email, Jira and ServiceNow.
- Prepared fixes can remain subject to human approval.
- Teams can start with selected services and critical or major problems.
Investigation begins when a problem opens
When Dynatrace opens a problem, the SRE Agent starts investigating immediately. According to Dynatrace, its findings are ready within minutes and include an explanation of the cause, supporting evidence from logs, metrics, deployments or configuration changes, and an assessment of affected services, users and business impact.
The agent also recommends what to do next, such as rolling back a deployment or restarting a failing service. It shows what it checked and ruled out, allowing engineers to trace its findings back to the available evidence.
In Dynatrace’s example, a failed deployment to a payment service caused checkout to fail for 29 minutes. The SRE Agent traced the incident to a coding issue that had caused similar problems before, recommended a rollback, notified the payment team and prepared the revert for approval.
Workflows control routing and remediation
The SRE Agent is embedded as an action within a Dynatrace workflow. The agent investigates the unknown, while the workflow defines when it starts, what context and instructions it receives, and what happens with its result.
A workflow can start when a problem opens, an event fires or a schedule runs. Teams can set conditions based on problem type, impact or criticality, then have the agent examine related logs, events, deployments and metrics. It can also follow Smartscape dependencies, check earlier problems with the same cause and rule out explanations that do not fit.
Configured rules determine the urgency of an incident, while ownership data identifies the responsible team. The workflow can annotate a problem, open a ServiceNow, Jira or Jira Service Management ticket, send a summary through Slack, Microsoft Teams or email, create a postmortem, or prepare a pull request, configuration change or runbook step for approval.
Dynatrace says users retain control over the workflow steps. Operations can run autonomously when desired, while remediation actions can remain human approved.
Dynatrace ties the agent to its existing analysis
The SRE Agent does not begin with an unstructured alert stream. Dynatrace says Dynatrace Intelligence has already identified the root cause and affected services and users before the agent runs. It also groups events with the same cause, so engineers receive one problem instead of a flood of alerts.
That existing context gives the agent a starting point for its investigation. Dynatrace says the agent builds on the root cause found in the customer’s own data, adds evidence and impact context, and recommends an action that can be reviewed against the evidence chain.
Teams can start with services they own and limit the agent to critical or major problems. Dynatrace recommends checking the analysis against the team’s incident findings, adjusting prompts and filters, and then sending results to Jira, ServiceNow, Slack or Microsoft Teams. Once the analysis proves useful, teams can widen its scope and move ownership to a service account.
