Basic install collectd operator for k8s
Test environment is currently using minikube v1.11.0 with the registry and OLM addons.
minikube start --cpus=4 --memory=49152 --driver=kvm2
for i in default-storageclass ingress olm registry storage-provisioner; do minikube addons enable $i; done
eval $(minikube podman-env)
source <(kubectl completion bash)
source <(minikube completion bash)
You can test that everything passes the Operator SDK scorecard. Currently we're
leveraging operator-sdk v0.15.2. You must run this when the collectd operator
is not already running. If the CRD is loaded, the scorecard will fail.
kubectl create namespace collectd
kubectl config set-context --current --namespace=collectd
operator-sdk scorecard
We're not currently making full use of the Operator Lifecycle Manager, so we'll be importing things manually.
First, create the OperatorGroup:
kubectl apply -f - <<EOF
apiVersion: operators.coreos.com/v1
kind: OperatorGroup
metadata:
name: collectd-og
namespace: collectd
spec:
targetNamespaces:
- collectd
EOF
Then create the ServiceAccount, ClusterRole, ClusterRoleBinding and ClusterServiceVersion:
kubectl create namespace collectd
kubectl config set-context --current --namespace=collectd
kubectl apply -f deploy/service_account.yaml -f deploy/role.yaml -f deploy/role_binding.yaml -f deploy/olm-catalog/collectd-operator/0.0.2/collectd.infra.watch_collectds_crd.yaml
sed -e "s#placeholder#collectd#g" deploy/olm-catalog/collectd-operator/0.0.2/collectd-operator.v0.0.2.clusterserviceversion.yaml | kubectl apply -f -
Then we can check that everything is working.
kubectl get all,csv
NAME READY STATUS RESTARTS AGE
pod/collectd-operator-65749b54b4-ngwn4 2/2 Running 0 7m50s
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/collectd-operator-metrics ClusterIP 10.103.102.203 <none> 8686/TCP,8383/TCP 7m44s
NAME READY UP-TO-DATE AVAILABLE AGE
deployment.apps/collectd-operator 1/1 1 1 7m50s
NAME DESIRED CURRENT READY AGE
replicaset.apps/collectd-operator-65749b54b4 1 1 1 7m50s
NAME DISPLAY VERSION REPLACES PHASE
clusterserviceversion.operators.coreos.com/collectd-operator.v0.0.2 Collectd Operator 0.0.2 collectd-operator.v0.0.1 Succeeded
Load additional components to setup a transport mechanism.
Load the ServiceAccount, Role, RoleBinding, ClusterRole, ClusterRoleBinding and CustomResourceDefinition:
git clone https://github.com/interconnectedcloud/qdr-operator.git
cd qdr-operator
kubectl apply -f deploy/service_account.yaml \
-f deploy/role.yaml \
-f deploy/role_binding.yaml \
-f deploy/cluster_role.yaml \
-f deploy/cluster_role_binding.yaml \
-f deploy/crds/interconnectedcloud_v1alpha1_interconnect_crd.yaml
Then load the ClusterServiceVersion:
| NOTE You may need to apply a small patch prior to running the next command. See interconnectedcloud/qdr-operator#75 |
|---|
sed -e "s#placeholder#collectd#g" deploy/olm-catalog/qdr-operator/0.4.0/qdr-operator.v0.4.0.clusterserviceversion.yaml | kubectl apply -f -
Then create an instance of QDR.
kubectl apply -f - <<EOF
apiVersion: interconnectedcloud.github.io/v1alpha1
kind: Interconnect
metadata:
name: 'qdr-interconnect'
namespace: 'collectd'
spec:
deploymentPlan:
size: 1
role: interior
livenessPort: 8888
placement: AntiAffinity
addresses:
- distribution: closest
prefix: closest
- distribution: multicast
prefix: multicast
- distribution: closest
prefix: unicast
- distribution: closest
prefix: exclusive
- distribution: multicast
prefix: broadcast
- distribution: multicast
prefix: collectd
- distribution: multicast
prefix: ceilometer
listeners:
- port: 5672
- expose: true
http: true
port: 8672
EOF
Instantiate the Smart Gateway Operator.
git clone https://github.com/infrawatch/smart-gateway-operator
cd smart-gateway-operator
kubectl apply -f deploy/role_binding.yaml -f deploy/role.yaml -f deploy/service_account.yaml -f deploy/olm-catalog/smart-gateway-operator/1.0.2/smartgateway.infra.watch_smartgateways_crd.yaml
sed -e "s#placeholder#collectd#g" deploy/olm-catalog/smart-gateway-operator/1.0.2/smart-gateway-operator.v1.0.2.clusterserviceversion.yaml | kubectl apply -f -
Create a Smart Gateway instance.
kubectl apply -f - <<EOF
apiVersion: smartgateway.infra.watch/v2alpha1
kind: SmartGateway
metadata:
name: 'collectd-metrics-telemetry'
namespace: 'collectd'
spec:
amqpUrl: 'qdr-interconnect:5672/collectd/openshift-telemetry'
debug: false
serviceType: 'metrics'
size: 1
prefetch: 15000
useTimestamp: true
EOF
Subscribe to Prometheus Operator from the OperatorHubIO catalog source:
kubectl apply -f - <<EOF
apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
name: prometheus
namespace: collectd
spec:
channel: beta
installPlanApproval: Automatic
name: prometheus
source: operatorhubio-catalog
sourceNamespace: olm
startingCSV: prometheusoperator.0.37.0
EOF
Create a Prometheus instance:
kubectl apply -f - <<EOF
apiVersion: monitoring.coreos.com/v1
kind: Prometheus
metadata:
labels:
prometheus: 'prometheus'
name: prometheus
namespace: collectd
spec:
replicas: 1
ruleSelector: {}
securityContext: {}
serviceAccountName: prometheus-k8s
serviceMonitorSelector:
matchLabels:
app: smart-gateway
alerting:
alertmanagers:
- name: alertmanager-operated
namespace: collectd
port: web
EOF
Create a ServiceMonitor to result in Prometheus being configured to scrape our new Smart Gateway:
kubectl apply -f - <<EOF
apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
labels:
app: smart-gateway
name: 'collectd'
namespace: 'collectd'
spec:
endpoints:
- metricRelabelings:
- action: labeldrop
regex: pod
sourceLabels: []
- action: labeldrop
regex: namespace
sourceLabels: []
- action: labeldrop
regex: instance
sourceLabels: []
- action: labeldrop
regex: job
sourceLabels: []
port: prom-http
selector:
matchLabels:
app: smart-gateway
EOF
Create an Alertmanager instance:
kubectl apply -f - <<EOF
apiVersion: v1
kind: Secret
metadata:
name: 'alertmanager-collectd'
namespace: 'collectd'
type: Opaque
stringData:
alertmanager.yaml: |-
global:
resolve_timeout: 5m
route:
group_by: ['job']
group_wait: 30s
group_interval: 5m
repeat_interval: 12h
receiver: 'null'
receivers:
- name: 'null'
---
apiVersion: monitoring.coreos.com/v1
kind: Alertmanager
metadata:
labels:
alertmanager: 'alertmanager'
name: 'collectd'
namespace: 'collectd'
spec:
replicas: 1
serviceAccountName: prometheus-k8s
EOF
kubectl apply -f - <<EOF
apiVersion: collectd.infra.watch/v1alpha1
kind: Collectd
metadata:
name: 'collectd'
namespace: 'collectd'
spec:
collectdHost: qdr-interconnect
EOF
Accessing the Prometheus UI via minikube is done by exposing the
prometheus-operated service.
kubectl expose -n collectd service prometheus-operated --port=80 --target-port=9090 --name=prometheus-web --type=LoadBalancer
minikube service prometheus-web -n collectd