With those changes, diff does show any differences between generated manifests
with and without the version override in a top level jsonnet file.

kube-prometheus: Add Prometheus Operator alerts

contrib/.../alerts: improve alert messages

Only trigger disk running full alerts when the disk capacity is over 85%
and use recording rule of the kubernetes-mixin to only consider real
physical filesystems.

This bumps the kubernetes-mixin dependency to add
https://github.com/kubernetes-monitoring/kubernetes-mixin/pull/76.

Signed-off-by: Sergiusz Urbaniak <sergiusz.urbaniak@gmail.com>

Signed-off-by: Sergiusz Urbaniak <sergiusz.urbaniak@gmail.com>

Currently, we are mounting `/root` from the host as `/host/root`.
Instead simply `/` from the host should be mounted.

This fixes it

Signed-off-by: Sergiusz Urbaniak <sergiusz.urbaniak@gmail.com>

kube-prometheus documentation

Trying to get travis-ci build to pass for "git diff --exit-code" command

Signed-off-by: Sergiusz Urbaniak <sergiusz.urbaniak@gmail.com>

Signed-off-by: Sergiusz Urbaniak <sergiusz.urbaniak@gmail.com>

Fixes https://github.com/prometheus/node_exporter#474

 in the operator
Fixes #569 (by making the mount explicit)

Signed-off-by: Sergiusz Urbaniak <sergiusz.urbaniak@gmail.com>

Update Prometheus to v2.3.2 and Alertmanager to 0.15.2

Adding the following accross the project:

```bash
/# exit immediately when a command fails
set -e
/# only exit with zero if all commands of the pipeline exit successfully
set -o pipefail
/# error on unset variables
+set -u

```

kube-prometheus: Allow skipping etcd TLS errors