I'm using minikube with traefik ingress to create a sticky sessions.
So i have done the deploy of traefik that user guide kubernetes provides me. https://docs.traefik.io/user-guide/kubernetes/
I deploy traefik using DaemonSet. Cause it's a small project and is my first time using kubernetes and docker.
This is my ingress yaml file
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
name: cp-pluggin
annotations:
kubernetes.io/ingress.class: traefik
spec:
rules:
- host: cppluggins.minikube
http:
paths:
- path: /
backend:
serviceName: cp-pluggin
servicePort: 80
My service yaml file
apiVersion: v1
kind: Service
metadata:
name: cp-pluggin
annotations:
traefik.ingress.kubernetes.io/affinity: "true"
traefik.ingress.kubernetes.io/session-cookie-name: "sticky"
spec:
type: NodePort
ports:
- port: 80
targetPort: 8080
protocol: TCP
name: http
selector:
app: cp-pluggin-app
Finally, my deployment yaml file
apiVersion: apps/v1
kind: Deployment
metadata:
name: cp-pluggin-app
labels:
app: cp-pluggin-app
spec:
replicas: 3
selector:
matchLabels:
app: cp-pluggin-app
template:
metadata:
labels:
app: cp-pluggin-app
spec:
containers:
- name: cp-pluggin-app
image: essoca/ubuntu-tornado
ports:
- containerPort: 8080
I expected
Hello world from: [ipserver]
But i get a
bad gateway
I assume you are using Traefik 2.0, the latest version as of now. There are quite some changes in this version, i.e. the annotations are not used anymore. Besides that, I think the code that you posted is missing a big part of the required changes.
Also, it's not very useful to use a DaemonSet because you are using minikube and that's always one node. Using a Deployment will at least allow you to play with the scale up/down functionality of Kubernetes.
I wrote this article that might be useful for you Traefik 2 as Ingress Controller
Related
Hi I don't know whether this is an issue that I am having since I don't have a lot of experience in Kubernetes.
I am trying a deployment in Kubernetes but the logic behind the container it is that it start initialising it self instantly and calling a backend endpoint to register a user.
From the screenshot below I don't know whether this is an misconfiguration about my deployment but I think that the POD starts multiple containers and this breaks my deployment since the flow goes like this:
Container starts
Calls backend endpoint and register user
Kubernetes starts another container
Calls backend endpoint and fails to register user because its already been registered
POD fails to deploy
Screenshot
As you can see in the screenshot the container count is 5
Adding as well the deployment yaml file
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: vulos-hyperledger-scanner
namespace: vuloss-scanner
spec:
replicas: 1
selector:
matchLabels:
app: vulos-hyperledger-scanner
template:
metadata:
labels:
app: vulos-hyperledger-scanner
spec:
containers:
- name: vulos-hyperledger-scanner
image: registry.digitalocean.com/notarised/vulos-hyperledger-scanner:1.6.5
ports:
- containerPort: 8080
env:
imagePullPolicy: Always
imagePullSecrets:
- name: do-registry
---
apiVersion: v1
kind: Service
metadata:
name: vulos-hyperledger-scanner
namespace: vuloss-scanner
spec:
selector:
app: vulos-hyperledger-scanner
ports:
- port: 80
targetPort: 8080
type: ClusterIP
---
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
annotations:
kubernetes.io/ingress.class: nginx
cert-manager.io/cluster-issuer: "letsencrypt-production"
name: vulos-hyperledger-scanner
namespace: vuloss-scanner
spec:
rules:
- host: explorer.vulos.io
http:
paths:
- backend:
serviceName: vulos-hyperledger-scanner
servicePort: 80
path: /
#This section is only required if TLS is to be enabled for the Ingress
tls:
- hosts:
- explorer.vulos.io
secretName: explorer-tls
I don't know whether this is an issue in my deployment and whether this statement of above makes sense in Kubernetes and whether I can make the POD to start only 1 container rather than multiple
Thank you
Stack:
Azure Kubernetes Service
NGINX Ingress Controller - https://github.com/kubernetes/ingress-nginx
AKS Loadbalancer
Docker containers
My goal is to create a K8s cluster that will allow me to use multiple pods, under a single IP, to create a microservice architecture. After working with tons of tutorials and documentation, I'm not having any luck with my endgoal. I got to the point of being able to access a single deployment using the Loadbalancer, but introducing the ingress has not been successful so far. The services are separated into their respective files for readability and ease of control.
Additionally, the Ingress Controller was added to my cluster as described in the installation instructions using: kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v0.35.0/deploy/static/provider/cloud/deploy.yaml
LoadBalancer.yml:
apiVersion: v1
kind: Service
metadata:
name: backend
spec:
loadBalancerIP: x.x.x.x
selector:
app: ingress-service
tier: backend
ports:
- name: "default"
port: 80
targetPort: 80
type: LoadBalancer
IngressService.yml:
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadata:
name: ingress-service
annotations:
kubernetes.io/ingress.class: nginx
spec:
rules:
- http:
paths:
- path: /api
backend:
serviceName: api-service
servicePort: 80
api-deployment.yml
apiVersion: v1
kind: Service
metadata:
name: api-service
spec:
selector:
app: api
ports:
- port: 80
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: api-deployment
spec:
selector:
matchLabels:
app: api
tier: backend
track: stable
replicas: 1
template:
metadata:
labels:
app: api
tier: backend
track: stable
spec:
containers:
- name: api
image: image:tag
ports:
- containerPort: 80
imagePullPolicy: Always
imagePullSecrets:
- name: SECRET
The API in the image is exposed on port 80 correctly.
After applying each of the above yml services and deployments, I attempt a web request to one of the api resources via the LoadBalancer's IP and receive only a timeout on my requests.
Found my answer after hunting around enough. Basically, the problem was that the Ingress Controller has a Load Balancer built into the yaml, as mentioned in comments above. However, the selector for that LoadBalancer requires marking your Ingress service as part of the class. Then that Ingress service points to each of the services attached to your pods. I also had to make a small modification to allow using a static IP in the provided load balancer.
I am trying to redirect traffic to my Cluster Deployment object. My Deployment object is a node image which depends on a SQL server image.
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadata:
name: example-ingress
annotations:
nginx.ingress.kubernetes.io/rewrite-target: /
spec:
rules:
- http:
paths:
- path: /
backend:
serviceName: simptekapi-api-service
servicePort: 5000
MY Cluster IP service ====>
apiVersion: v1
kind: Service
metadata:
name: simptekapi-api-service
spec:
selector:
component: api
type: ClusterIP
ports:
- port: 5000
targetPort: 5000
My deployment Object ===>
apiVersion: apps/v1
kind: Deployment
metadata:
name: simptekapi-api-deployment
spec:
replicas: 3
selector:
matchLabels:
component: api
template:
metadata:
labels:
component: api
spec:
containers:
- name: simptekapi-api
image: nayan2/simptekapi
ports:
- containerPort: 5000
My node image is basically a Test API Project. I am sharing the SignIn url (http://Minikube_ip/api/v1/auth/SignIn)
Seems like everything ok to me. As am still literally a newbie with Kubernetes. It's hard for me to figure out what I am doing wrong.
While I have tried to reach my pods through Node-Port, everything was perfect. I don't know why I am unable to react it out through ingress.
kubectl get ingress example-ingress ==> Result
AND kubectl get pods -n ingress-nginx ==> Result
Currently processing on this tutorial,
https://github.com/argoproj/argocd-example-apps/tree/master/guestbook
https://argoproj.github.io/argo-cd/getting_started/#5-register-a-cluster-to-deploy-apps-to-optional
My short-term milestone is to render guest-book's UI on browser.
I'm trying to connect via Ingress, and it went wrong.
Error message's like this,
Status: 502
The server encountered a temporary error and could not complete your request.
I suppose something's wrong around service and pod.
guestbook-ui-svc.yaml
apiVersion: v1
kind: Service
metadata:
name: guestbook-ui-service
spec:
ports:
- port: 80
targetPort: 80
selector:
app: guestbook-ui
guestbook-ui-ingress.yaml
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
name: app-ingress
labels:
app: guestbook-ui
spec:
rules:
- http:
paths:
- path: /
backend:
serviceName: guestbook-ui-service
servicePort: 80
guestbook-ui-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: guestbook-ui
spec:
replicas: 1
revisionHistoryLimit: 3
selector:
matchLabels:
app: guestbook-ui
template:
metadata:
labels:
app: guestbook-ui
spec:
containers:
- image: gcr.io/heptio-images/ks-guestbook-demo:0.2
name: guestbook-ui
ports:
- containerPort: 80
I don't know which part I am missing, please lmk any ambiguous part or more detail.
Thanks, in advance!
Use this service instead.
apiVersion: v1
kind: Service
metadata:
name: guestbook-ui-service
spec:
type: NodePort
ports:
- port: 80
targetPort: 80
selector:
app: guestbook-ui
It has type: NodePort added to it.
You can check really good example on how to deploy an app, expose it via a service and add an ingress to it. It's available in kubernetes docs Deploy a hello, world app.
Also if you are having problem understanding the difference between NodePort, ClusterIP and what Ingress is I recommend reading Kubernetes NodePort vs LoadBalancer vs Ingress? When should I use what?
I have an EKS cluster for which I want :
- 1 Load Balancer per cluster,
- Ingress rules to direct to the right namespace and the right service.
I have been following this guide : https://www.digitalocean.com/community/tutorials/how-to-set-up-an-nginx-ingress-with-cert-manager-on-digitalocean-kubernetes
My deployments:
apiVersion: apps/v1
kind: Deployment
metadata:
name: hello-world
namespace: default
spec:
replicas: 3
selector:
matchLabels:
app: hello-world
template:
metadata:
labels:
app: hello-world
spec:
containers:
- name: hello-world
image: IMAGENAME
ports:
- containerPort: 8000
name: hello-world
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: bleble
namespace: default
spec:
replicas: 3
selector:
matchLabels:
app: bleble
template:
metadata:
labels:
app: bleble
spec:
containers:
- name: bleble
image: IMAGENAME
ports:
- containerPort: 8000
name: bleble
the service of those deployments:
apiVersion: v1
kind: Service
metadata:
name: hello-world-svc
spec:
ports:
- port: 8080
protocol: TCP
targetPort: 8000
selector:
app: hello-world
type: NodePort
---
apiVersion: v1
kind: Service
metadata:
name: bleble-svc
spec:
ports:
- port: 8080
protocol: TCP
targetPort: 8000
selector:
app: bleble
type: NodePort
My Load balancer:
kind: Service
apiVersion: v1
metadata:
name: ingress-nginx
namespace: ingress-nginx
annotations:
service.beta.kubernetes.io/aws-load-balancer-internal: "true"
labels:
app.kubernetes.io/name: ingress-nginx
app.kubernetes.io/part-of: ingress-nginx
spec:
externalTrafficPolicy: Local
type: LoadBalancer
selector:
app.kubernetes.io/name: ingress-nginx
app.kubernetes.io/part-of: ingress-nginx
ports:
- name: http
port: 80
targetPort: http
My ingress:
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
name: simple-fanout-example
namespace : default
annotations:
kubernetes.io/ingress.class: nginx
nginx.ingress.kubernetes.io/rewrite-target: /
spec:
rules:
- host: internal-lb.aws.com
http:
paths:
- path: /bleble
backend:
serviceName: bleble-svc
servicePort: 80
- path: /hello-world
backend:
serviceName: hello-world-svc
servicePort: 80
I've set up the Nginx Ingress Controller with this : kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/nginx-0.24.1/deploy/mandatory.yaml
I am unsure why I get a 503 Service Temporarily Unavailable for one service and one 502 for another... I would guess it's a problem of ports or of namespace? In the guide, they don't define namespace for the deployment...
Every resources create correctly, and I think the ingress is actually working but is getting confused where to go.
Thanks for your help!
In general, use externalTrafficPolicy: Cluster instead of Local. You can gain some performance (latency) improvement by using Local but you need to configure those pod allocations with a lot efforts. You will hit 5xx errors with those misconfigurations. In addition, Cluster is the default option for externalTrafficPolicy.
In your ingress, you route /bleble to service bleble, but your service name is actually bleble-svc. please make them consistent. Also, you would need to set your servicePort to 8080 as you exposed 8080 in your service configuration.
For internal service like bleble-svc, Cluster IP is good enough in your case as it does not need external access.
Hope this helps.
Found it!
The containerPort in the Deployment were set to 8000, the targetport of the services as well, but the person who did the Dockerfile of the code exposed the port 80. Which was the reason it was getting the 502 Bad getaway!
Thanks a lot as well to #Fei who has been a fantastic helper!