I install kube1.10.3 in two virtualbox(centos 7.4) in my win10 machine. I use git clone to get prometheus yaml files.
git clone https://github.com/kubernetes/kubernetesThen I enter kubernetes/cluster/addons/prometheus annd follow this order to create pods:
alertmanager-configmap.yaml
alertmanager-pvc.yaml
alertmanager-deployment.yaml
alertmanager-service.yaml
kube-state-metrics-rbac.yaml
kube-state-metrics-deployment.yaml
kube-state-metrics-service.yaml
node-exporter-ds.yml
node-exporter-service.yaml
prometheus-configmap.yaml
prometheus-rbac.yaml
prometheus-statefulset.yaml
prometheus-service.yamlBut Prometheus and alertmanage are in pending state:
kube-system alertmanager-6bd9584b85-j4h5m 0/2 Pending 0 9m
kube-system calico-etcd-pnwtr 1/1 Running 0 16m
kube-system calico-kube-controllers-5d74847676-mjq4j 1/1 Running 0 16m
kube-system calico-node-59xfk 2/2 Running 1 16m
kube-system calico-node-rqsh5 2/2 Running 1 16m
kube-system coredns-7997f8864c-ckhsq 1/1 Running 0 16m
kube-system coredns-7997f8864c-jjtvq 1/1 Running 0 16m
kube-system etcd-master16g 1/1 Running 0 15m
kube-system heapster-589b7db6c9-mpmks 1/1 Running 0 16m
kube-system kube-apiserver-master16g 1/1 Running 0 15m
kube-system kube-controller-manager-master16g 1/1 Running 0 15m
kube-system kube-proxy-hqq49 1/1 Running 0 16m
kube-system kube-proxy-l8hmh 1/1 Running 0 16m
kube-system kube-scheduler-master16g 1/1 Running 0 16m
kube-system kube-state-metrics-8595f97c4-g6x5x 2/2 Running 0 8m
kube-system kubernetes-dashboard-7d5dcdb6d9-944xl 1/1 Running 0 16m
kube-system monitoring-grafana-7b767fb8dd-mg6dd 1/1 Running 0 16m
kube-system monitoring-influxdb-54bd58b4c9-z9tgd 1/1 Running 0 16m
kube-system node-exporter-f6pmw 1/1 Running 0 8m
kube-system node-exporter-zsd9b 1/1 Running 0 8m
kube-system prometheus-0 0/2 Pending 0 7mI checked prometheus pod by command shown below:
[root@master16g prometheus]# kubectl describe pod prometheus-0 -n kube-system
Name: prometheus-0
Namespace: kube-system
Node: <none>
Labels: controller-revision-hash=prometheus-8fc558cb5
k8s-app=prometheus
statefulset.kubernetes.io/pod-name=prometheus-0
Annotations: scheduler.alpha.kubernetes.io/critical-pod=
Status: Pending
IP:
Controlled By: StatefulSet/prometheus
Init Containers:
init-chown-data:
Image: busybox:latest
Port: <none>
Host Port: <none>
Command:
chown
-R
65534:65534
/data
Environment: <none>
Mounts:
/data from prometheus-data (rw)
/var/run/secrets/kubernetes.io/serviceaccount from prometheus-token-f6v42 (ro)
Containers:
prometheus-server-configmap-reload:
Image: jimmidyson/configmap-reload:v0.1
Port: <none>
Host Port: <none>
Args:
--volume-dir=/etc/config
--webhook-url=http://localhost:9090/-/reload
Limits:
cpu: 10m
memory: 10Mi
Requests:
cpu: 10m
memory: 10Mi
Environment: <none>
Mounts:
/etc/config from config-volume (ro)
/var/run/secrets/kubernetes.io/serviceaccount from prometheus-token-f6v42 (ro)
prometheus-server:
Image: prom/prometheus:v2.2.1
Port: 9090/TCP
Host Port: 0/TCP
Args:
--config.file=/etc/config/prometheus.yml
--storage.tsdb.path=/data
--web.console.libraries=/etc/prometheus/console_libraries
--web.console.templates=/etc/prometheus/consoles
--web.enable-lifecycle
Limits:
cpu: 200m
memory: 1000Mi
Requests:
cpu: 200m
memory: 1000Mi
Liveness: http-get http://:9090/-/healthy delay=30s timeout=30s period=10s #success=1 #failure=3
Readiness: http-get http://:9090/-/ready delay=30s timeout=30s period=10s #success=1 #failure=3
Environment: <none>
Mounts:
/data from prometheus-data (rw)
/etc/config from config-volume (rw)
/var/run/secrets/kubernetes.io/serviceaccount from prometheus-token-f6v42 (ro)
Conditions:
Type Status
PodScheduled False
Volumes:
prometheus-data:
Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
ClaimName: prometheus-data-prometheus-0
ReadOnly: false
config-volume:
Type: ConfigMap (a volume populated by a ConfigMap)
Name: prometheus-config
Optional: false
prometheus-token-f6v42:
Type: Secret (a volume populated by a Secret)
SecretName: prometheus-token-f6v42
Optional: false
QoS Class: Guaranteed
Node-Selectors: <none>
Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
node.kubernetes.io/unreachable:NoExecute for 300s
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Warning FailedScheduling 42s (x22 over 5m) default-scheduler pod has unbound PersistentVolumeClaims (repeated 2 times)In the last line, it shows warning message: pod has unbound PersistentVolumeClaims (repeated 2 times)
The Prometheus logs says:
[root@master16g prometheus]# kubectl logs prometheus-0 -n kube-system
Error from server (BadRequest): a container name must be specified for pod prometheus-0, choose one of: [prometheus-server-configmap-reload prometheus-server] or one of the init containers: [init-chown-data]The I describe alertmanager pod and its logs:
[root@master16g prometheus]# kubectl describe pod alertmanager-6bd9584b85-j4h5m -n kube-system
Name: alertmanager-6bd9584b85-j4h5m
Namespace: kube-system
Node: <none>
Labels: k8s-app=alertmanager
pod-template-hash=2685140641
version=v0.14.0
Annotations: scheduler.alpha.kubernetes.io/critical-pod=
Status: Pending
IP:
Controlled By: ReplicaSet/alertmanager-6bd9584b85
Containers:
prometheus-alertmanager:
Image: prom/alertmanager:v0.14.0
Port: 9093/TCP
Host Port: 0/TCP
Args:
--config.file=/etc/config/alertmanager.yml
--storage.path=/data
--web.external-url=/
Limits:
cpu: 10m
memory: 50Mi
Requests:
cpu: 10m
memory: 50Mi
Readiness: http-get http://:9093/%23/status delay=30s timeout=30s period=10s #success=1 #failure=3
Environment: <none>
Mounts:
/data from storage-volume (rw)
/etc/config from config-volume (rw)
/var/run/secrets/kubernetes.io/serviceaccount from default-token-snfrt (ro)
prometheus-alertmanager-configmap-reload:
Image: jimmidyson/configmap-reload:v0.1
Port: <none>
Host Port: <none>
Args:
--volume-dir=/etc/config
--webhook-url=http://localhost:9093/-/reload
Limits:
cpu: 10m
memory: 10Mi
Requests:
cpu: 10m
memory: 10Mi
Environment: <none>
Mounts:
/etc/config from config-volume (ro)
/var/run/secrets/kubernetes.io/serviceaccount from default-token-snfrt (ro)
Conditions:
Type Status
PodScheduled False
Volumes:
config-volume:
Type: ConfigMap (a volume populated by a ConfigMap)
Name: alertmanager-config
Optional: false
storage-volume:
Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
ClaimName: alertmanager
ReadOnly: false
default-token-snfrt:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-snfrt
Optional: false
QoS Class: Guaranteed
Node-Selectors: <none>
Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
node.kubernetes.io/unreachable:NoExecute for 300s
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Warning FailedScheduling 3m (x26 over 9m) default-scheduler pod has unbound PersistentVolumeClaims (repeated 2 times)And its log:
[root@master16g prometheus]# kubectl logs alertmanager-6bd9584b85-j4h5m -n kube-system
Error from server (BadRequest): a container name must be specified for pod alertmanager-6bd9584b85-j4h5m, choose one of: [prometheus-alertmanager prometheus-alertmanager-configmap-reload] It has same warning message as Prometheus:
pod has unbound PersistentVolumeClaims (repeated 2 times)Then I get pvc by issuing command as follows:
[root@master16g prometheus]# kubectl get pvc --all-namespaces
NAMESPACE NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
kube-system alertmanager Pending standard 20m
kube-system prometheus-data-prometheus-0 Pending standard 19mMy question is how to make bound persistentVolumnClaim? Why log says container name must be specified?
\===============================================================
Since pvc file defined storage class, so I need to define a storage class yaml. How to do it if I want Nfs or GlusterFs? In this way, I could avoid cloud vendor, like Google or AWS.
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: alertmanager
namespace: kube-system
labels:
kubernetes.io/cluster-service: "true"
addonmanager.kubernetes.io/mode: EnsureExists
spec:
storageClassName: standard
accessModes:
- ReadWriteOnce
resources:
requests:
storage: "2Gi"This log entry:
Error from server (BadRequest): a container name must be specified for pod alertmanager-6bd9584b85-j4h5m, choose one of: [prometheus-alertmanager prometheus-alertmanager-configmap-reload] means Pod alertmanager-6bd9584b85-j4h5m consists of two containers:
When you use kubectl logs for Pod which consists of more then one containers you must specify a name of the container to view its logs. Command template:
kubectl -n <namespace> logs <pod_name> <container_name>For example, if you want to view logs of the container prometheus-alertmanager which is a part of Pod alertmanager-6bd9584b85-j4h5m in the namespace kube-system you should use this command:
kubectl -n kube-system logs alertmanager-6bd9584b85-j4h5m prometheus-alertmanagerPending status of the PVCs could mean you have no corresponding PVs