Opsview hangs after some time

[megabreit]

Opsview hangs after some unknown (longer) time displaying only the top menu (color and sort options) and no cluster anymore.

netstat in the kubeopsview container shows >2500 establisched connections to the redis pod:
netstat|grep kube-ops-view-6b6b5b6964-mhccj|wc -l
2554

[..]
tcp 2458306 0 kube-ops-view-6b6b5b6964-mhccj:57588 kube-ops-view-redis.ocp-ops-view.svc.cluster.local:6379 ESTABLISHED
tcp 1658433 0 kube-ops-view-6b6b5b6964-mhccj:52628 kube-ops-view-redis.ocp-ops-view.svc.cluster.local:6379 ESTABLISHED
Active UNIX domain sockets (w/o servers)
Proto RefCnt Flags Type State I-Node Path

Runtime is Openshit 3.11 (Kubernetes 1.11) with increased redis memory limits compared to the original deployment.

[megabreit]

Looking through the other open issues, this is probably similar (or duplicate) to #240

[megabreit]

Some more debugging:
ops-view is starting with 3 established connections to redis.
Every time the browser window is reloaded (either by reload button or the reload option) a new established connection appears. It seems that connections never terminate and consume resources until it crashes/hangs.