Fixed the issue with the GlusterFS service host for Kubernetes-based clusters
referring to a wrong node, which caused Telegraf fail to gather metrics from
GlusterFS.
Fixed the issue with the proxy status metrics missing in the Grafana web UI
in case of Kubernetes deployments with OpenContrail.
Fixed the issue with alerts for dropped packets on PXE interfaces by removing
the Major alert for dropped packets and decreasing the default threshold
for the Warning alert for dropped packets from 100 to 60.
Fixed the issue with the CephPoolWriteOpsTooHigh alert firing on healthy
Ceph clusters by adding the capability to modify the alert threshold for
individual pools.