A hands-on walkthrough of using Postman Agent Mode connected to Dynatrace via MCP (Model Context Protocol) to perform root cause analysis on a PostgreSQL table lock incident. The author built a lab with a .NET 10 API, injected a deliberate blocking transaction, and used a single plain-English prompt to identify the root cause in ~3 minutes — compared to 20–30 minutes of manual Dynatrace investigation. The post covers lab setup (OneAgent, ActiveGate, PostgreSQL logging config), the incident injection, what the AI agent found, honest limitations (Davis AI CoPilot required, free plan message limits, correlation vs comprehension), and a brief comparison of Dynatrace vs Datadog MCP server capabilities.
Table of contents
The Investigation That Eats Your MorningWhat MCP Is and Why It MattersBuilding the Lab: The Real Secret Is TelemetryActiveGate and PostgreSQL ExtensionPostgreSQL Logging ConfigurationOneAgent and Deep MonitoringThe Baseline:The Incident: Injecting the LockGet Gessiel Lima Jr ’s stories in your inboxThe Investigation: One Prompt in Plain EnglishWhat Agent Mode FoundWhat Worked, and the Honest LimitationsWhat worked wellThe honest limitationsDynatrace vs Datadog: A Brief NoteSort: