Practice Test - Application Failure

Take me to Practice Test of the Application Failure

Solution

A simple 2 tier application is deployed in the alpha namespace. It must display a green web page on success. Click on the App tab at the top of your terminal to view your application. It is currently failed. Troubleshoot and fix the issue.

Stick to the given architecture. Use the same names and port numbers as given in the below architecture diagram. Feel free to edit, delete or recreate objects as necessary.
1. Click on the App button at the top of the terminal. Observe the error message.
It is telling us
```
Name does not resolve
```
This is a DNS lookup issue. Coredns does not know any service called mysql-service, however the architecture diagram says that there should be this service.
1. Examine services in the alpha namespace
```
kubectl get service -n alpha
```
  We see there is a service mysql so the likelihood is that the service is deployed with incorrect name
2. Fix it
  
  Note that you cannot use kubectl edit to change a resource name.
```
kubectl get service -n alpha mysql -o yaml > service.yaml
vi service.yaml
```
```
apiVersion: v1
kind: Service
metadata:
  creationTimestamp: "2023-10-14T11:40:30Z"
  name: mysql           # <- Edit this to mysql-service
  namespace: alpha
  resourceVersion: "824"
  uid: d9a85021-547a-4a39-b254-0480830eab6a
spec:
```
3. Delete and recreate service
```
kubectl delete svc mysql -n alpha
kubectl create -f service.yaml
```

The same 2 tier application is deployed in the beta namespace. It must display a green web page on success. Click on the App tab at the top of your terminal to view your application. It is currently failed. Troubleshoot and fix the issue.

Stick to the given architecture. Use the same names and port numbers as given in the below architecture diagram. Feel free to edit, delete or recreate objects as necessary.

Click on the App button at the top of the terminal. Observe the error message.

It is telling us
```
Can't connect to MySQL server on 'mysql-service:3306'
```
Which suggests that the service exists but the port settings may be incorrect.

Edit the service

kubectl edit service mysql-service -n beta

apiVersion: v1
kind: Service
metadata:
  creationTimestamp: "2023-10-14T11:54:04Z"
  name: mysql-service
  namespace: beta
  resourceVersion: "1166"
  uid: d6b07c71-5c49-4118-849a-0b12dd382597
spec:
  clusterIP: 10.43.42.85
  clusterIPs:
  - 10.43.42.85
  internalTrafficPolicy: Cluster
  ipFamilies:
  - IPv4
  ipFamilyPolicy: SingleStack
  ports:
  - port: 3306        # <- Correct
    protocol: TCP
    targetPort: 8080  # <- Incorrect, should also be 3306
  selector:
    name: mysql
  sessionAffinity: None
  type: ClusterIP

The mysql database server listens on port 3306 meaning that is the port that its pod will expose, and where targetPort should be pointing. We can verify that like this

kubectl get po -n beta mysql -o jsonpath='{.spec.containers[*].ports[*].containerPort}'

or simply get the yaml for the pod.

The same 2 tier application is deployed in the gamma namespace. It must display a green web page on success. Click on the App tab at the top of your terminal to view your application. It is currently failed or unresponsive. Troubleshoot and fix the issue.

Stick to the given architecture. Use the same names and port numbers as given in the below architecture diagram. Feel free to edit, delete or recreate objects as necessary.
1. Click on the App button at the top of the terminal. Observe the error message.
  
  It is telling us
```
 Can't connect to MySQL server on 'mysql-service:3306' (111 Connection refused)
```
  Which suggests that the service exists but there is something wrong with it.
2. Edit the service
```
kubectl edit service mysql-service -n gamma
```
  This time the name is correct, as are both the ports. So perhaps the pod selector is the issue
  
  Open an additional terminal (+ button above terminal) so as not to have to quit vi now, and run
```
kubectl get pods -n gamma --show-labels
```
  Note the labels. Always ignore pod-template-hash label. It is used internally by kubernetes.
```
name=webapp-mysql,pod-template-hash=5456999f7b
```
  That doesn't match with sql00001 in the service selector. Switch back to your vi session in Terminal 1 and fix the selector to use the correct value for the name label.
```
apiVersion: v1
kind: Service
metadata:
  creationTimestamp: "2023-10-14T12:04:31Z"
  name: mysql-service
  namespace: gamma
  resourceVersion: "1441"
  uid: 872dfbcd-0b1d-4292-85f6-803d62d05b0a
spec:
  clusterIP: 10.43.21.189
  clusterIPs:
  - 10.43.21.189
  internalTrafficPolicy: Cluster
  ipFamilies:
  - IPv4
  ipFamilyPolicy: SingleStack
  ports:
  - port: 3306
    protocol: TCP
    targetPort: 3306
  selector:
    name: sql00001      # <- Fix this
  sessionAffinity: None
  type: ClusterIP
```
The same 2 tier application is deployed in the delta namespace. It must display a green web page on success. Click on the App tab at the top of your terminal to view your application. It is currently failed. Troubleshoot and fix the issue.

Stick to the given architecture. Use the same names and port numbers as given in the below architecture diagram. Feel free to edit, delete or recreate objects as necessary.
1. Click on the App button at the top of the terminal. Observe the error message.
  
  It is telling us
```
Access denied for user 'sql-user'@'10.42.0.16'
```
  So this means the application is using an incorrect mysql user account. The architecture diagram in the question tells you what the correct credentials are.
  
  So this time the fix is going to be in the application deployment, rather than in the service.
2. Fix the deployment
```
kubectl edit deployment webapp-mysql -n delta
```
  Scroll down to the container's environment section and fix the user name
```
  template:
    metadata:
      creationTimestamp: null
      labels:
        name: webapp-mysql
      name: webapp-mysql
    spec:
      containers:
      - env:
        - name: DB_Host
          value: mysql-service
        - name: DB_User
          value: sql-user   # <- Fix this
        - name: DB_Password
          value: paswrd
        image: mmumshad/simple-webapp-mysql
        imagePullPolicy: Always
        name: webapp-mysql
```
The same 2 tier application is deployed in the epsilon namespace. It must display a green web page on success. Click on the App tab at the top of your terminal to view your application. It is currently failed. Troubleshoot and fix the issue.

The question indicates there are two issues that need fixing.
1. Click on the App button at the top of the terminal. Observe the error message.
  
  It is telling us
```
Access denied for user 'sql-user'@'10.42.0.16'
```
  So it is another authentication issue.
2. Check the deployment
  
  Vaildate the environment against the values provided in the architecture diagram. Looks like the same issue as previous question, so fix that the same way.
3. Wait! We fixed that, but it's still not working!
  
  Check the mysql pod, since mysql also needs some credential information on its end.
  
  We cannot use kubectl edit to change values of a standalone POD.
```
kubectl get pod -n epsilon mysql -o yaml > mysql.yaml
vi mysql.yaml
```
```
apiVersion: v1
kind: Pod
metadata:
  creationTimestamp: "2023-10-14T12:21:51Z"
  labels:
    name: mysql
  name: mysql
  namespace: epsilon
  resourceVersion: "2002"
  uid: f369770b-be24-43c1-b754-d7e1396d6952
spec:
  containers:
  - env:
    - name: MYSQL_ROOT_PASSWORD
      value: passwooooorrddd    # <- Fix this
    image: mysql:5.6
    imagePullPolicy: IfNotPresent
    name: mysql
    ports:
```
4. Recreate the pod
```
kubectl replace --force -f mysql.yaml
```
The same 2 tier application is deployed in the zeta namespace. It must display a green web page on success. Click on the App tab at the top of your terminal to view your application. It is currently failed. Troubleshoot and fix the issue.

The question indicates there are three issues to fix.
1. Click on the App button at the top of the terminal. Observe the error message.
  
  We get a 502 Bad Gateway error.
  
  This is indicative that the lab display infrastructure cannot connect to the service it's supposed to. Examine the URL in the browser address bar
```
30081-port-1795f98fde814933.labs.kodekloud.com/
```
  On KodeKloud labs, the 30081-port part indicates a node port it's trying to connect to. Note also that the infrastructure diagram states that 30081 should be the web-service nodeport.
2. Examine web-service since that's how we view the app.
```
kubectl edit service -n zeta web-service
```
```
apiVersion: v1
kind: Service
metadata:
  creationTimestamp: "2023-10-14T13:27:26Z"
  name: web-service
  namespace: zeta
  resourceVersion: "1530"
  uid: 9b655dda-675a-43ae-80e4-deadb3a38179
spec:
  clusterIP: 10.43.48.98
  clusterIPs:
  - 10.43.48.98
  externalTrafficPolicy: Cluster
  internalTrafficPolicy: Cluster
  ipFamilies:
  - IPv4
  ipFamilyPolicy: SingleStack
  ports:
  - nodePort: 30088  # <- Edit this
    port: 8080
    protocol: TCP
    targetPort: 8080
  selector:
    name: webapp-mysql
  sessionAffinity: None
  type: NodePort
```
3. Retry the app in the browser
  
  We have seen this message before! Fix it as per above, using namesapce zeta.
4. Retry the app in the browser
  
  We have also seen this message before! Fix it as per above, using namesapce zeta.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

03-Solution-Application-Failure.md

03-Solution-Application-Failure.md

Practice Test - Application Failure

Solution

Files

03-Solution-Application-Failure.md

Latest commit

History

03-Solution-Application-Failure.md

File metadata and controls

Practice Test - Application Failure

Solution