Merge pull request #2208 from vishh/monitoring

Use Heapster as an in-built monitoring solution for Kubernetes in GCE.

cluster/gce/config-default.sh

@@ -35,3 +35,5 @@ MINION_SCOPES="compute-rw"
 # Increase the sleep interval value if concerned about API rate limits. 3, in seconds, is the default.
 POLL_SLEEP_INTERVAL=3
 PORTAL_NET="10.0.0.0/16"
+# When set to true, heapster will be setup as part of the cluster bring up.
+MONITORING=true

cluster/gce/config-test.sh

@@ -35,3 +35,4 @@ MINION_SCOPES=""
 # Increase the sleep interval value if concerned about API rate limits. 3, in seconds, is the default.
 POLL_SLEEP_INTERVAL=3
 PORTAL_NET="10.0.0.0/16"
+MONITORING=false

cluster/gce/util.sh

@@ -569,3 +569,46 @@ function ssh-to-node {
 function restart-kube-proxy {
   ssh-to-node "$1" "sudo /etc/init.d/kube-proxy restart"
 }
+
+# Setup monitoring using heapster and InfluxDB
+function setup-monitoring {
+    if [ $MONITORING ]; then
+	teardown-monitoring
+	if ! gcutil getfirewall monitoring-heapster &> /dev/null; then
+	    gcutil addfirewall monitoring-heapster \
+		--project "${PROJECT}" \
+		--norespect_terminal_width \
+		--sleep_between_polls "${POLL_SLEEP_INTERVAL}" \
+		--target_tags="${MINION_TAG}" \
+		--allowed "tcp:80,tcp:8083,tcp:8086,tcp:9200";
+	    if [ $? -ne 0 ]; then
+		echo "Failed to Setup Firewall for Monitoring" && false
+	    fi
+	fi
+
+	kubectl.sh create -f "${KUBE_ROOT}/examples/monitoring/influx-grafana-pod.json" > /dev/null && 
+	kubectl.sh create -f "${KUBE_ROOT}/examples/monitoring/influx-grafana-service.json" > /dev/null &&
+	kubectl.sh create -f "${KUBE_ROOT}/examples/monitoring/heapster-pod.json" > /dev/null
+	if [ $? -ne 0 ]; then
+	    echo "Failed to Setup Monitoring"
+	    teardown-monitoring
+	else
+	    dashboardIP="http://admin:admin@`kubectl.sh get -o json pod influx-grafana | grep hostIP | awk '{print $2}' | sed 's/[,|\"]//g'`"
+	    echo "Grafana dashboard will be available at $dashboardIP. Wait for the monitoring dashboard to be online."
+	fi
+    fi
+}
+
+function teardown-monitoring {
+  if [ $MONITORING ]; then
+    kubectl.sh delete pods heapster &> /dev/null || true
+    kubectl.sh delete pods influx-grafana &> /dev/null || true
+    kubectl.sh delete services influx-master &> /dev/null || true
+    gcutil deletefirewall  \
+	--project "${PROJECT}" \
+	--norespect_terminal_width \
+	--sleep_between_polls "${POLL_SLEEP_INTERVAL}" \
+	--force \
+	monitoring-heapster || true > /dev/null
+  fi
+}

cluster/kube-up.sh

@@ -34,5 +34,6 @@ verify-prereqs
 kube-up
 
 "${KUBE_ROOT}/cluster/validate-cluster.sh"
+setup-monitoring
 
 echo "Done"

cluster/rackspace/util.sh

@@ -322,3 +322,7 @@ kube-up() {
   echo "    subject to \"Man in the middle\" type attacks."
   echo
 }
+
+function setup-monitoring {
+    echo "TODO"
+}

cluster/vagrant/util.sh

@@ -172,3 +172,7 @@ function ssh-to-node {
 function restart-kube-proxy {
   ssh-to-node "$1" "sudo systemctl restart kube-proxy"
 }
+
+function setup-monitoring {
+    echo "TODO"
+}

cluster/vsphere/util.sh

@@ -471,3 +471,7 @@ function test-setup {
 function test-teardown {
 	echo "TODO"
 }
+
+function setup-monitoring {
+    echo "TODO"
+}

examples/monitoring/README.md

@@ -0,0 +1,3 @@
+# Heapster
+
+Heapster enables monitoring of Kubernetes Clusters using [cAdvisor](https://github.com/google/cadvisor). Detailed information about heapster can be found [here](https://github.com/GoogleCloudPlatform/heapster).

examples/monitoring/heapster-pod.json

@@ -0,0 +1,18 @@
+{
+  "id": "heapster",
+  "kind": "Pod",
+  "apiVersion": "v1beta1",
+  "desiredState": {
+    "manifest": {
+      "version": "v1beta1",
+      "id": "heapster",
+      "containers": [{
+        "name": "heapster",
+        "image": "kubernetes/heapster",
+	}]
+    }
+  },
+  "labels": {
+    "name": "heapster",
+  }
+}

examples/monitoring/influx-grafana-pod.json

@@ -0,0 +1,34 @@
+{
+  "id": "influx-grafana",
+  "kind": "Pod",
+  "apiVersion": "v1beta1",
+  "desiredState": {
+    "manifest": {
+      "version": "v1beta1",
+      "id": "influx-grafana",
+      "containers": [{
+        "name": "influxdb",
+        "image": "kubernetes/heapster_influxdb",
+        "ports": [
+		 {"containerPort": 8083, "hostPort": 8083},
+		 {"containerPort": 8086, "hostPort": 8086},
+		 {"containerPort": 8090, "hostPort": 8090},
+		 {"containerPort": 8099, "hostPort": 8099}]
+      }, {
+	"name": "grafana",
+        "image": "kubernetes/heapster_grafana",
+        "ports": [{"containerPort": 80, "hostPort": 80}],
+	"env": [{"name": HTTP_USER, "value": admin},
+	        {"name": HTTP_PASS, "value": admin}],
+      }, {
+	"name": "elasticsearch",
+        "image": "dockerfile/elasticsearch",
+        "ports": [{"containerPort": 9200, "hostPort": 9200},
+		  {"containerPort": 9300}],
+      }]
+    },
+  },
+  "labels": {
+    "name": "influxdb",
+  }
+}

examples/monitoring/influx-grafana-service.json

@@ -0,0 +1,10 @@
+{
+  "id": "influx-master",
+  "kind": "Service",
+  "apiVersion": "v1beta1",
+  "port": 8085,
+  "containerPort": 8086,
+  "provider": "kubernetes-default",
+  "component": "influxdb",
+  "selector": { "name": "influxdb" }
+}