I don't yet have specific steps to reproduce; however, I have a WebFlux based project that has become unresponsive after an extended period of time (~48-72 hours) w/ minimal traffic.
The behaviors I've observed is that javascript (compiled via the maven frontend plugin) and copied using the resources plugin seem to not serve. Eventually, all endpoints seem to hang minus the actuator endpoints (otherwise I'd be observing AWS ECS cycling the service).
From a grafana standpoint I didn't see anything anomalous w.r.t. spring-boot metrics being collected by prometheus.
The system does have a few @Scheduled jobs that perform things like database cleanup (R2DBC + MySQL). 3.1.6 seems stable since I reverted down.
Furthermore, nothing in logs really showed.
Are there any useful debugging steps anyone would recommend so I can provide something more detailed / useful for the maintainers here?
I don't yet have specific steps to reproduce; however, I have a WebFlux based project that has become unresponsive after an extended period of time (~48-72 hours) w/ minimal traffic.
The behaviors I've observed is that javascript (compiled via the maven frontend plugin) and copied using the resources plugin seem to not serve. Eventually, all endpoints seem to hang minus the actuator endpoints (otherwise I'd be observing AWS ECS cycling the service).
From a grafana standpoint I didn't see anything anomalous w.r.t. spring-boot metrics being collected by prometheus.
The system does have a few @Scheduled jobs that perform things like database cleanup (R2DBC + MySQL). 3.1.6 seems stable since I reverted down.
Furthermore, nothing in logs really showed.
Are there any useful debugging steps anyone would recommend so I can provide something more detailed / useful for the maintainers here?