Hugo Marcelo Del Negro

I don't see any logging information from Tile38. Now it keeps refusing my connectios for more than 12 seconds then starts working again for a few 30/40 minutes. This failed...

Yes it does not says too much so far than HOOKS failing intermittent: ``` 2018/09/10 03:30:03 [DEBU] Endpoint connect/send error: 1001: http://track-server/api/v2/fence/tile38_webhook/Kd1Tq5j5gx6PWw:a6b72d3c6caa1a715a6bd69db24500c97063e18db60d61756f4296b1ff4407ac: Post http://track-server/api/v2/fence/tile38_webhook/Kd1Tq5j5gx6PWw:a6b72d3c6caa1a715a6bd69db24500c97063e18db60d61756f4296b1ff4407ac: net/http: request canceled while waiting for...

@tidwall We are about ~14 fence hooks in the environment where I took the logs.

I am experiencing the same issue running this in Kubenetes: ``` admins-MacBook-Pro:scripts admin$ kubectl logs tile38-df59bd788-zbln2 _______ _______ | | | |____ | _ | Tile38 1.12.2 (d5af91c) 64 bit...

@tidwall my file is 470MB , could it be an issue ?

I have heapster running as explained by @Mahesh-MSFT but I am still experiencing de unknown message for current resources. Any other way to try that ?

@samlyhuang it should work if you enable heaspter in the cluster and apply resources/limits to the manifest for the container. did you try that ?