无法通过AWS上的公共IP访问Prometheus

Question

Use kops install k8s cluster on AWS.

Use Helm installed Prometheus :

$ helm install stable/prometheus \
    --set server.persistentVolume.enabled=false \
    --set alertmanager.persistentVolume.enabled=false

Then followed this note to do port-forward :

Get the Prometheus server URL by running these commands in the same shell:
  export POD_NAME=$(kubectl get pods --namespace default -l "app=prometheus,component=server" -o jsonpath="{.items[0].metadata.name}")
  kubectl --namespace default port-forward $POD_NAME 9090

My EC2 instance public IP on AWS is 12.29.43.14 (not true). When I tried to access it from browser:

http://12.29.43.14:9090

Can't access the page. Why?

---

Another issue, after installed prometheus chart, the alertmanager pod didn't run:

ungaged-woodpecker-prometheus-alertmanager-6f9f8b98ff-qhhw4       1/2       CrashLoopBackOff   1          9s
ungaged-woodpecker-prometheus-kube-state-metrics-5fd97698cktsj5   1/1       Running            0          9s
ungaged-woodpecker-prometheus-node-exporter-45jtn                 1/1       Running            0          9s
ungaged-woodpecker-prometheus-node-exporter-ztj9w                 1/1       Running            0          9s
ungaged-woodpecker-prometheus-pushgateway-57b67c7575-c868b        0/1       Running            0          9s
ungaged-woodpecker-prometheus-server-7f858db57-w5h2j              1/2       Running            0          9s

Check pod details:

$ kubectl describe po ungaged-woodpecker-prometheus-alertmanager-6f9f8b98ff-qhhw4
Name:           ungaged-woodpecker-prometheus-alertmanager-6f9f8b98ff-qhhw4
Namespace:      default
Node:           ip-100.200.0.1.ap-northeast-1.compute.internal/100.200.0.1
Start Time:     Fri, 26 Jan 2018 02:45:10 +0000
Labels:         app=prometheus
                component=alertmanager
                pod-template-hash=2959465499
                release=ungaged-woodpecker
Annotations:    kubernetes.io/created-by={"kind":"SerializedReference","apiVersion":"v1","reference":{"kind":"ReplicaSet","namespace":"default","name":"ungaged-woodpecker-prometheus-alertmanager-6f9f8b98ff","uid":"ec...
                kubernetes.io/limit-ranger=LimitRanger plugin set: cpu request for container prometheus-alertmanager; cpu request for container prometheus-alertmanager-configmap-reload
Status:         Running
IP:             100.96.6.91
Created By:     ReplicaSet/ungaged-woodpecker-prometheus-alertmanager-6f9f8b98ff
Controlled By:  ReplicaSet/ungaged-woodpecker-prometheus-alertmanager-6f9f8b98ff
Containers:
  prometheus-alertmanager:
    Container ID:  docker://e9fe9d7bd4f78354f2c072d426fa935d955e0d6748c4ab67ebdb84b51b32d720
    Image:         prom/alertmanager:v0.9.1
    Image ID:      docker-pullable://prom/alertmanager@sha256:ed926b227327eecfa61a9703702c9b16fc7fe95b69e22baa656d93cfbe098320
    Port:          9093/TCP
    Args:
      --config.file=/etc/config/alertmanager.yml
      --storage.path=/data
    State:          Waiting
      Reason:       CrashLoopBackOff
    Last State:     Terminated
      Reason:       Error
      Exit Code:    1
      Started:      Fri, 26 Jan 2018 02:45:26 +0000
      Finished:     Fri, 26 Jan 2018 02:45:26 +0000
    Ready:          False
    Restart Count:  2
    Requests:
      cpu:        100m
    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-wppzm (ro)
  prometheus-alertmanager-configmap-reload:
    Container ID:  docker://9320a0f157aeee7c3947027667aa6a2e00728d7156520c19daec7f59c1bf6534
    Image:         jimmidyson/configmap-reload:v0.1
    Image ID:      docker-pullable://jimmidyson/configmap-reload@sha256:2d40c2eaa6f435b2511d0cfc5f6c0a681eeb2eaa455a5d5ac25f88ce5139986e
    Port:          <none>
    Args:
      --volume-dir=/etc/config
      --webhook-url=http://localhost:9093/-/reload
    State:          Running
      Started:      Fri, 26 Jan 2018 02:45:11 +0000
    Ready:          True
    Restart Count:  0
    Requests:
      cpu:        100m
    Environment:  <none>
    Mounts:
      /etc/config from config-volume (ro)
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-wppzm (ro)
Conditions:
  Type           Status
  Initialized    True
  Ready          False
  PodScheduled   True
Volumes:
  config-volume:
    Type:      ConfigMap (a volume populated by a ConfigMap)
    Name:      ungaged-woodpecker-prometheus-alertmanager
    Optional:  false
  storage-volume:
    Type:    EmptyDir (a temporary directory that shares a pod's lifetime)
    Medium:
  default-token-wppzm:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-wppzm
    Optional:    false
QoS Class:       Burstable
Node-Selectors:  <none>
Tolerations:     node.alpha.kubernetes.io/notReady:NoExecute for 300s
                 node.alpha.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason                 Age                From                                                      Message
  ----     ------                 ----               ----                                                      -------
  Normal   Scheduled              34s                default-scheduler                                         Successfully assigned ungaged-woodpecker-prometheus-alertmanager-6f9f8b98ff-qhhw4 to ip-100.200.0.1.ap-northeast-1.compute.internal
  Normal   SuccessfulMountVolume  34s                kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  MountVolume.SetUp succeeded for volume "storage-volume"
  Normal   SuccessfulMountVolume  34s                kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  MountVolume.SetUp succeeded for volume "config-volume"
  Normal   SuccessfulMountVolume  34s                kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  MountVolume.SetUp succeeded for volume "default-token-wppzm"
  Normal   Pulled                 33s                kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  Container image "jimmidyson/configmap-reload:v0.1" already present on machine
  Normal   Created                33s                kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  Created container
  Normal   Started                33s                kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  Started container
  Normal   Pulled                 18s (x3 over 34s)  kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  Container image "prom/alertmanager:v0.9.1" already present on machine
  Normal   Created                18s (x3 over 34s)  kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  Created container
  Normal   Started                18s (x3 over 33s)  kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  Started container
  Warning  BackOff                2s (x4 over 32s)   kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  Back-off restarting failed container
  Warning  FailedSync             2s (x4 over 32s)   kubelet, ip-100.200.0.1.ap-northeast-1.compute.internal  Error syncing pod

Not sure why it FailedSync .

Answer 1

When you do a kubectl port-forward with that command it makes the port available on your localhost. So run the command and then hit http://localhost:9090 .

You won't be able to directly hit the prometheus ports from the public IP, outside the cluster. In the longer run you may want expose prometheus at a nice domain name via ingress (which the chart supports), that's how I'd do it. To use the chart's support for ingress you will need to install an ingress controller in your cluster (like the nginx ingress controller for example), and then enable ingress by setting --set service.ingress.enabled=true and --set server.ingress.hosts[0]=prometheus.yourdomain.com . Ingress is a fairly large topic in itself, so I'll just refer you to the official docs for that one:

https://kubernetes.io/docs/concepts/services-networking/ingress/

And here's the nginx ingress controller:

https://github.com/kubernetes/ingress-nginx

As far as the pod that is showing FailedSync , take a look at the logs using kubectl logs ungaged-woodpecker-prometheus-alertmanager-6f9f8b98ff-qhhw4 to see if there's any additional information there.