Debug OOMKilled
We currently have the problem, that some collector terminate after a short period with OOMKilled. We tried different things #219 (closed) #215 (closed) to get ride of the error without luck. There is aos a long standing issue #223 to keep the heap small.
Our collector (hallernet) is terminated after ~30s. This has nothing to do with #223, but is a different problem, maybe rooted in K8. We need to get a memory dump from production process in the cluster. In dev we cant reproduce the behavior. Here the memory is stable at 30mb. The prometheuse monitoring also show a memory consumption of about 25mb in production. Anyway we get a 137 error. Maybe the memory spikes
Edited by Tobinsk