I am trying to deploy nginx on kubernetes, kubernetes version is v1.5.2,
I have deployed nginx with 3 replica, YAML file is below,
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
name: deployment-example
spec:
replicas: 3
revisionHistoryLimit: 2
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.10
ports:
- containerPort: 80
and now I want to expose its port 80 on port 30062 of node, for that I created a service below,
kind: Service
apiVersion: v1
metadata:
name: nginx-ils-service
spec:
ports:
- name: http
port: 80
nodePort: 30062
selector:
app: nginx
type: LoadBalancer
this service is working good as it should be, but it is showing as pending not only on kubernetes dashboard also on terminal.
It looks like you are using a custom Kubernetes Cluster (using minikube, kubeadm or the like). In this case, there is no LoadBalancer integrated (unlike AWS or Google Cloud). With this default setup, you can only use NodePort or an Ingress Controller.
With the Ingress Controller you can setup a domain name which maps to your pod; you don't need to give your Service the LoadBalancer type if you use an Ingress Controller.
If you are using Minikube, there is a magic command!
$ minikube tunnel
Hopefully someone can save a few minutes with this.
Reference link
https://minikube.sigs.k8s.io/docs/handbook/accessing/#using-minikube-tunnel
If you are not using GCE or EKS (you used kubeadm) you can add an externalIPs spec to your service YAML. You can use the IP associated with your node's primary interface such as eth0. You can then access the service externally, using the external IP of the node.
...
spec:
type: LoadBalancer
externalIPs:
- 192.168.0.10
I created a single node k8s cluster using kubeadm. When i tried PortForward and kubectl proxy, it showed external IP as pending.
$ kubectl get svc -n argocd argocd-server
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
argocd-server LoadBalancer 10.107.37.153 <pending> 80:30047/TCP,443:31307/TCP 110s
In my case I've patched the service like this:
kubectl patch svc <svc-name> -n <namespace> -p '{"spec": {"type": "LoadBalancer", "externalIPs":["172.31.71.218"]}}'
After this, it started serving over the public IP
$ kubectl get svc argo-ui -n argo
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
argo-ui LoadBalancer 10.103.219.8 172.31.71.218 80:30981/TCP 7m50s
To access a service on minikube, you need to run the following command:
minikube service [-n NAMESPACE] [--url] NAME
More information here : Minikube GitHub
When using Minikube, you can get the IP and port through which you
can access the service by running:
minikube service [service name]
E.g.:
minikube service kubia-http
If it is your private k8s cluster, MetalLB would be a better fit. Below are the steps.
Step 1: Install MetalLB in your cluster
kubectl apply -f https://raw.githubusercontent.com/metallb/metallb/v0.9.3/manifests/namespace.yaml
kubectl apply -f https://raw.githubusercontent.com/metallb/metallb/v0.9.3/manifests/metallb.yaml
# On first install only
kubectl create secret generic -n metallb-system memberlist --from-literal=secretkey="$(openssl rand -base64 128)"
Step 2: Configure it by using a configmap
apiVersion: v1
kind: ConfigMap
metadata:
namespace: metallb-system
name: config
data:
config: |
address-pools:
- name: default
protocol: layer2
addresses:
- 172.42.42.100-172.42.42.105 #Update this with your Nodes IP range
Step 3: Create your service to get an external IP (would be a private IP though).
FYR:
Before MetalLB installation:
After MetalLB installation:
If running on minikube, don't forget to mention namespace if you are not using default.
minikube service << service_name >> --url --namespace=<< namespace_name >>
Following #Javier's answer. I have decided to go with "patching up the external IP" for my load balancer.
$ kubectl patch service my-loadbalancer-service-name \
-n lb-service-namespace \
-p '{"spec": {"type": "LoadBalancer", "externalIPs":["192.168.39.25"]}}'
This will replace that 'pending' with a new patched up IP address you can use for your cluster.
For more on this. Please see karthik's post on LoadBalancer support with Minikube for Kubernetes
Not the cleanest way to do it. I needed a temporary solution. Hope this helps somebody.
If you are using minikube then run commands below from terminal,
$ minikube ip
$ 172.17.0.2 // then
$ curl http://172.17.0.2:31245
or simply
$ curl http://$(minikube ip):31245
In case someone is using MicroK8s: You need a network load balancer.
MicroK8s comes with metallb, you can enable it like this:
microk8s enable metallb
<pending> should turn into an actual IP address then.
A general way to expose an application running on a set of Pods as a network service is called service in Kubernetes. There are four types of service in Kubernetes.
ClusterIP
The Service is only reachable from within the cluster.
NodePort
You'll be able to communicate the Service from outside the cluster using NodeIP:NodePort.default node port range is 30000-32767, and this range can be changed by define --service-node-port-range in the time of cluster creation.
LoadBalancer
Exposes the Service externally using a cloud provider's load balancer.
ExternalName
Maps the Service to the contents of the externalName field (e.g. foo.bar.example.com), by returning a CNAME record with its value. No proxying of any kind is set up.
Only the LoadBalancer gives value for the External-IP Colum. and it only works if the Kubernetes cluster is able to assign an IP address for that particular service. you can use metalLB load balancer for provision IPs to your load balancer services.
I hope your doubt may go away.
You can patch the IP of Node where pods are hosted ( Private IP of Node ) , this is the easy workaround .
Taking reference with above posts , Following worked for me :
kubectl patch service my-loadbalancer-service-name \
-n lb-service-namespace \
-p '{"spec": {"type": "LoadBalancer", "externalIPs":["xxx.xxx.xxx.xxx Private IP of Physical Server - Node - where deployment is done "]}}'
Adding a solution for those who encountered this error while running on amazon-eks.
First of all run:
kubectl describe svc <service-name>
And then review the events field in the example output below:
Name: some-service
Namespace: default
Labels: <none>
Annotations: kubectl.kubernetes.io/last-applied-configuration:
{"apiVersion":"v1","kind":"Service","metadata":{"annotations":{},"name":"some-service","namespace":"default"},"spec":{"ports":[{"port":80,...
Selector: app=some
Type: LoadBalancer
IP: 10.100.91.19
Port: <unset> 80/TCP
TargetPort: 5000/TCP
NodePort: <unset> 31022/TCP
Endpoints: <none>
Session Affinity: None
External Traffic Policy: Cluster
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal EnsuringLoadBalancer 68s service-controller Ensuring load balancer
Warning SyncLoadBalancerFailed 67s service-controller Error syncing load balancer: failed to ensure load balancer: could not find any suitable subnets for creating the ELB
Review the error message:
Failed to ensure load balancer: could not find any suitable subnets for creating the ELB
In my case, the reason that no suitable subnets were provided for creating the ELB were:
1: The EKS cluster was deployed on the wrong subnets group - internal subnets instead of public facing.
(*) By default, services of type LoadBalancer create public-facing load balancers if no service.beta.kubernetes.io/aws-load-balancer-internal: "true" annotation was provided).
2: The Subnets weren't tagged according to the requirements mentioned here.
Tagging VPC with:
Key: kubernetes.io/cluster/yourEKSClusterName
Value: shared
Tagging public subnets with:
Key: kubernetes.io/role/elb
Value: 1
If you are using a bare metal you need the NodePort type
https://kubernetes.github.io/ingress-nginx/deploy/baremetal/
LoadBalancer works by default in other cloud providers like Digital Ocean, Aws, etc
k edit service ingress-nginx-controller
type: NodePort
spec:
externalIPs:
- xxx.xxx.xxx.xx
using the public IP
Use NodePort:
$ kubectl run user-login --replicas=2 --labels="run=user-login" --image=kingslayerr/teamproject:version2 --port=5000
$ kubectl expose deployment user-login --type=NodePort --name=user-login-service
$ kubectl describe services user-login-service
(Note down the port)
$ kubectl cluster-info
(IP-> Get The IP where master is running)
Your service is accessible at (IP):(port)
The LoadBalancer ServiceType will only work if the underlying infrastructure supports the automatic creation of Load Balancers and have the respective support in Kubernetes, as is the case with the Google Cloud Platform and AWS. If no such feature is configured, the LoadBalancer IP address field is not populated and still in pending status , and the Service will work the same way as a NodePort type Service
minikube tunnel
The below solution works in my case.
First of all, try this command:
minikube tunnel
If it's not working for you. follow the below:
I restart minikube container.
docker minikube stop
then
docker minikube start
After that re-run kubernetes
minikube dashboard
After finish execute :
minikube tunnel
I have the same problem.
Windows 10 Desktop + Docker Desktop 4.7.1 (77678) + Minikube v1.25.2
Following the official docs on my side, I resolve with:
PS C:\WINDOWS\system32> kubectl expose deployment sito-php --type=LoadBalancer --port=8080 --name=servizio-php
service/servizio-php exposed
PS C:\WINDOWS\system32> minikube tunnel
* Tunnel successfully started
* NOTE: Please do not close this terminal as this process must stay alive for the tunnel to be accessible ...
* Starting tunnel for service servizio-php.
PS E:\docker\apache-php> kubectl get service
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 33h
servizio-php LoadBalancer 10.98.218.86 127.0.0.1 8080:30270/TCP 4m39s
The open browser on http://127.0.0.1:8080/
same issue:
os>kubectl get svc right-sabertooth-wordpress
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S)
right-sabertooth-wordpress LoadBalancer 10.97.130.7 "pending" 80:30454/TCP,443:30427/TCP
os>minikube service list
|-------------|----------------------------|--------------------------------|
| NAMESPACE | NAME | URL |
|-------------|----------------------------|--------------------------------|
| default | kubernetes | No node port |
| default | right-sabertooth-mariadb | No node port |
| default | right-sabertooth-wordpress | http://192.168.99.100:30454 |
| | | http://192.168.99.100:30427 |
| kube-system | kube-dns | No node port |
| kube-system | tiller-deploy | No node port |
|-------------|----------------------------|--------------------------------|
It is, however,accesible via that http://192.168.99.100:30454.
There are three types of exposing your service
Nodeport
ClusterIP
LoadBalancer
When we use a loadbalancer we basically ask our cloud provider to give us a dns which can be accessed online
Note not a domain name but a dns.
So loadbalancer type does not work in our local minikube env.
Those who are using minikube and trying to access the service of kind NodePort or LoadBalancer.
We don’t get the external IP to access the service on the local
system. So a good option is to use minikube IP
Use the below command to get the minikube IP once your service is exposed.
minikube service service-name --url
Now use that URL to serve your purpose.
Check kube-controller logs. I was able to solve this issue by setting the clusterID tags to the ec2 instance I deployed the cluster on.
If you are not on a supported cloud (aws, azure, gcloud etc..) you can't use LoadBalancer without MetalLB https://metallb.universe.tf/
but it's in beta yet..
Delete existing service and create a same new service solved my problems. My problems is that the loading balancing IP I defines is used so that external endpoint is pending. When I changed a new load balancing IP it still couldn't work.
Finally, delete existing service and create a new one solved my problem.
For your use case best option is to use NordPort service instead of loadbalancer type because loadbalancer is not available.
I was getting this error on the Docker-desktop. I just exit and turn it on again(Docker-desktop). It took few seconds, then It worked fine.
Deleting all older services and creating new resolved my issue. IP was bound to older service. just try "$kubectl get svc" and then delete all svc's one by one "$kubectl delete svc 'svc name' "
May be the subnet in which you are deploying your service, have not enough ip's
If you are trying to do this in your on-prem cloud, you need an L4LB service to create the LB instances.
Otherwise you end up with the endless "pending" message you described. It is visible in a video here: https://www.youtube.com/watch?v=p6FYtNpsT1M
You can use open source tools to solve this problem, the video provides some guidance on how the automation process should work.
I tried deploying on EKS, and my config.yaml follows this suggested format:
botfront:
app:
# The complete external host of the Botfront application (eg. botfront.yoursite.com). It must be set even if running on a private or local DNS (it populates the ROOT_URL).
host: botfront.yoursite.com
mongodb:
enabled: true # disable to use an external mongoDB host
# Username of the MongoDB user that will have read-write access to the Botfront database. This is not the root user
mongodbUsername: username
# Password of the MongoDB user that will have read-write access to the Botfront database. This is not the root user
mongodbPassword: password
# MongoDB root password
mongodbRootPassword: rootpassword
And I ran this command:
helm install -f config.yaml -n botfront --namespace botfront botfront/botfront
and the deployment appeared successful with all pods listed as running.
But botfront.yoursite.com goes nowhere. I checked the ingress and it matches, but there are no external ip addresses or anything. I don't know how to actually access my botfront site once deployed on kubernetes.
What am I missing?
EDIT:
With nginx lb installed kubectl get ingresses -n botfront now
returns:
NAME CLASS HOSTS ADDRESS PORTS AGE
botfront-app-ingress <none> botfront.cream.com a182b0b24e4fb4a0f8bd6300b440e5fa-423aebd224ce20ac.elb.us-east-2.amazonaws.com 80 4d1h
and
kubectl get svc -n botfront returns:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
botfront-api-service NodePort 10.100.207.27 <none> 80:31723/TCP 4d1h
botfront-app-service NodePort 10.100.26.173 <none> 80:30873/TCP 4d1h
botfront-duckling-service NodePort 10.100.75.248 <none> 80:31989/TCP 4d1h
botfront-mongodb-service NodePort 10.100.155.11 <none> 27017:30358/TCP 4d1h
If you run kubectl get svc -n botfront, it will show you all the Services that expose your botfront
$ kubectl get svc -n botfront
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
botfront-api-service NodePort 10.3.252.32 <none> 80:32077/TCP 63s
botfront-app-service NodePort 10.3.249.247 <none> 80:31201/TCP 63s
botfront-duckling-service NodePort 10.3.248.75 <none> 80:31209/TCP 63s
botfront-mongodb-service NodePort 10.3.252.26 <none> 27017:31939/TCP 64s
Each of them is of type NodePort, which means it exposes your app on the external IP address of each of your EKS cluster nodes on a specific port.
So if you your node1 ip happens to be 1.2.3.4 you can acess botfront-api-service on 1.2.3.4:32077. Don't forget to allow access to this port on firewall/security groups. If you have any registered domain e.g. yoursite.com you can configure for it a subdomain botfront.yoursite.com and point it to one of your EKS nodes. Then you'll be able to access it using your domain. This is the simplest way.
To be able to access it in a more effective way than by using specific node's IP and non-standard port, you may want to expose it via Ingress which will create an external load balancer, making your NodePort services available under one external IP adress and standard http port.
Update: I see that this chart already comes with ingress that exposes your app:
$ kubectl get ingresses -n botfront
NAME HOSTS ADDRESS PORTS AGE
botfront-app-ingress botfront.yoursite.com 80 70m
If you retrieve its yaml definition by:
$ kubectl get ingresses -n botfront -o yaml
you'll see that it uses the following annotation:
kubernetes.io/ingress.class: nginx
which means you need nginx-ingress controller installed on your EKS cluster. This might be one reason why it fails. As you can see in my example, this ingress doesn't get any external IP. That's because nginx-ingress wasn't installed on my GKE cluster. Not sure about EKS but as far as I know it doesn't come with nginx-ingress preinstalled.
One more thing: I assume that in your config.yaml you put some real domain name that you have registered instead of botfront.yoursite.com. Suppose your domain is yoursite.com and you successfully created subdomain botfront.yoursite.com, you should redirected it to the IP of your load balancer (the one used by your ingress).
If you run kubectl get ingresses -n botfront but the ADDRESS is empty, you probably don't have nginx-ingress installed and the underlying load balancer cannot be created. If you have here some external IP address, then redirect your registered domain to this address.
I Installed K8S with Helm Charts on EKS but the Loadbalancer EXTERNAL IP is in pending state , I see that EKS does support the service Type : LoadBalancer now.
Is it something I will have to check at the network outgoing traffic level ? Please share your experience if any.
Tx,
The Loadbalancer usually takes some seconds or a few minutes to provision you an IP.
If after 5 minutes the IP isn't provisioned:
- run kubectl get svc <SVC_NAME> -o yaml and if there is any different annotation set.
By default services with Type:LoadBalancer are provisioned with Classic Load Balancers automatically. Learn more here.
If you wish to use Network load Balancers you have to use the annotation:
service.beta.kubernetes.io/aws-load-balancer-type: nlb
The process is really automatic, you don't have to check for network traffic.
You can check if there is any issue with the Helm Chart you are deploying by manually creating a service with loadbalancer type and check if it gets provisioned:
$ kubectl run --generator=run-pod/v1 nginx --image=nginx --port=80
pod/nginx created
$ kubectl get pod nginx
NAME READY STATUS RESTARTS AGE
nginx 1/1 Running 0 34s
$ kubectl expose pod nginx --type=LoadBalancer
service/nginx exposed
$ kubectl get svc nginx -w
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
nginx LoadBalancer 10.1.63.178 <pending> 80:32522/TCP 7s
nginx LoadBalancer 10.1.63.178 35.238.146.136 80:32522/TCP 42s
In this example the LoadBalancer took 42s to be provisioned. This way you can verify if the issue is on the Helm Chart or something else.
If Kubernetes is running in an environment that doesn't support LoadBalancer services, the load balancer will not be provisioned, but the service will still behave like a NodePort service, your cloud/K8 engine should support LoadBalancer Service.
In that case, if you manage to add EIP or VIP to your node then you can attach to the EXTERNAL-IP of your TYPE=LoadBalancer in the K8 cluster, for example attaching the EIP/VIP address to the node 172.16.2.13.
kubectl patch svc ServiceName -p '{"spec": {"type": "LoadBalancer", "externalIPs":["172.16.2.13"]}}'
I got the following service defined:
apiVersion: v1
kind: Service
metadata:
name: customerservice
spec:
type: NodePort
selector:
app: customerapp
ports:
- protocol: TCP
port: 31004
nodePort: 31004
targetPort: 8080
Current situation: I am able to hit the pod via the service IP.
Now my goal is to reach the customerservice via the name of the service, which does not work right now. So I would simply type http://customerservice:31004 instead of http://<IP>:31004.
DNS resolution of services is ONLY available within the cluster, provided by CoreDNS/KubeDNS.
Should you wish to have access to this locally on your machine, you'd need to use another tool. One such tool is kubefwd:
https://github.com/txn2/kubefwd
A slightly simpler solution, is to use port-forward; which is a very simple way to access a single service locally.
kubectl port-forward --namespace=whatever svs/service-name port
EDIT:// I've made the assumption that you want to use the service DNS locally, as I'm assuming by saying:
I would simply type http://customerservice:31004
is in the context of your web browser.
Normal Services are assigned a DNS A record for a name of the form my-svc.my-namespace.svc.cluster.local. This resolves to the cluster IP of the Service. This DNS entry is present inside the kubernetes cluster only and hence you're able to access the service by name from inside the kubernetes pod.
Now, if you want to access your kubernetes service by name from one of the node you need to modify the /etc/resolve.conf of your node with <svc_name>.<namespace>.svc.cluster.local, please have a look at following /etc/resolve.conf
search ec2.internal default.svc.cluster.local svc.cluster.local cluster.local
nameserver 10.96.0.10
nameserver is clusterIP of kube-dns service, you can find it using kubectl get svc kube-dns -n kube-system
Now you will be able to curl your service as curl ui.default.svc.cluster.local:80
I've setup a NodePort service using the following config:
wordpress-service.yaml
apiVersion: v1
kind: Service
metadata:
labels:
app: wordpress
name: wordpress
spec:
type: NodePort
ports:
- port: 80
targetPort: 80
protocol: TCP
selector:
app: wordpress
Is this sufficient to access the service externally, if so how can I now access the service? What details do I need - and how do I determine them - for example node IP.
For Kubernetes on GCE:
We had the same question regarding services of type NodePort: How do we access node port services from our own host?
#ivan.sim 's answer (nodeIp:nodePort) is on mark however, you still wouldn't be able to access your service unless you add a firewall ingress (inbound to google cloud) traffic rule on the VPC network console to allow your host to be able to access your compute node
the above rule is dangerous and should be used only during development
You can find the node port using either the Google Cloud console or by running subsequent kubectl commands to find out the node running your pod which has your container. i.e kubectl get pods , kubectl describe pod your-pod-name, kubectl describe node node-that-runs-you-pod .status.addresses has your ExternalIP
It would be great if we could extract the node ip running our container in the pod using only a label/selector and a few line of commands, so here is what we did, in this case our selector is app: your-label:
$ nodename=$(kubectl get pods -o jsonpath='{.items[?(#.metadata.labels.app=="your-label")].spec.nodeName}')
$ nodeIp=$(kubectl get nodes -o jsonpath='{.items[?(#.metadata.name=="'$(echo $nodename)'")].status.addresses[?(#.type=="ExternalIP")].address}')
$ echo nodeIp
notice: we used json path to extract the information we desired, for more on json path see: json path
You could certainly turn this into a script that takes a label/selector as input and outputs an external ip of the node running your container !!!
To get the nodeport just type:
$ kubectl get services
under the PORT(S) columns you will see something like tagetPort:nodePort. this nodeport is what you want .
nodeIp:nodePort
When you define a service as type NodePort, every node in your cluster will proxy that port to your service. If you nodes are reachable from outside the Kubernetes cluster, you should be able to access the service at nodeIP:nodePort.
To determine nodeIP of a particular node, you can use either kubectl get no <node> -o yaml or kubectl describe no <node>. The status.Addresses field will be of interest. Generally, you will see fields like HostName, ExternalIP and InternalIP there.
To determine nodePort of your service, you can use either kubectl get svc wordpress -o yaml or kubectl describe svc wordpress. The spec.ports.nodePort is the port you need.
Service defined like this got assgned a high port number and is exposed on all your cluster nodes on that port (probably something like 3xxxx). Hard to tell the rest without proper knowledge of how your cluster is provisioned. kubectl get nodes should give you some knowledge about your nodes.
Although I assume you want to expose the service to the outside world. In the long run I suggest getting familiar with LoadBalancer type services and Ingress / IngressController