Closed obenkenobi closed 4 months ago
This occurs in a scenario where more transactions are being created than completed. This is expected when more reactive operations are being created than finished. There doesn't seem to be anything wrong with the agent. So I am closing the issue.
Description
When I ran a spring boot app that did scheduled jobs that made reactive calls to r2dbc-postgres, I encountered connection failures when connecting to postgres with the agent.
This resulted in a memory leak in my application.
Even disabling the r2dbc-postgres instrumentation did not fix the memory leak though this is most likely related to our reactor instrumentation.
Test App Link
Expected Behavior
No memory leak should be created even with continuous connection failures to postgres.
Steps to Reproduce
Run this test app Test App Link connecting to a local postgres instance. Then stop your postgres instance from running so connection failures begin.