ECEProject

Execution

To execute the Task 3 code:

Clone this repository and cd into the directory.

Start the proxy server. Our metrics server depends on it.

kubectl proxy &

python3 main.py <job_simulation_file.txt>

Make sure to run cleanup.py before and after executing the above main file, to cleanup and make nodes active.

For monitoring of cluster, execute:

python3 start_monitor.py

Code Structure

Files

• main.py: This is start of the program. It starts the job scheduling and Global controller takes over for assignment of jobs.
• start_monitor.py: The driver code for monitoring. Should be started parallely.
• monitor.py: The functions in these return the utlization metrics to feed to the Controllers.
• middleware.py: The code that administrates pod creation.
• LocalController.py: The Controller class for individual nodes.
• GlobalController.py: The Controller class for the cluster.
• cleanup.py: To cleanup the pods and enable nodes.

Extra Credit - REST APIs

Do this for all nodes

For creating image

cd python-docker
docker build -t python-docker .
docker images

Metal LB (Network Load Balancer)

kubectl apply -f https://raw.githubusercontent.com/metallb/metallb/v0.13.12/config/manifests/metallb-native.yaml

Apply the following yaml file

apiVersion: metallb.io/v1beta1
kind: IPAddressPool
metadata: 
  name: first-pool
  namespace: metallb-system
spec:
  addresses: 
  - 128.110.217.70-128.110.217.75

Applying the following L2 advertisement

apiVersion: metallb.io/v1beta1
kind: L2Advertisement
metadata:
  name: group5
  namespace: metallb-system
spec:
  ipAddressPools:
  - first-pool

Applying nginx ingress

kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v1.8.2/deploy/static/provider/cloud/deploy.yaml

Apply the deployment yaml file

Add the External IP from the ingress-controller

$ kubectl get service ingress-nginx-controller --namespace=ingress-nginx
$ kubectl get ingress

sudo vi /etc/hosts 

128.110.217.71 	cloud-controller.com

Execution

kubectl apply -f deployment.yaml

python3 driver.py <job.txt>

Extra Credit - Logs from Fluentbit

To install Fluentbit and run it as containers in node1 and node2.

The installing command have beed added into Makefile:

sudo make

Create a configuration file for Fluent Bit, which is fluent-bit-config.yaml. Deploy Fluent Bit as a DaemonSet so that it runs on every node in your Kubernetes cluster by ceating fluent-bit-daemonset.yaml. Both files have beed added into repo.

Get necessary permission to list events at the luster scope, fix role-based access control (RBAC) issue: Create a ClusterRole that grants access to the events resources. Create a ClusterRoleBinding to grant the events-reader role to the default service account in the logging namespace.

kubectl apply -f events-clusterrole.yaml
kubectl apply -f events-clusterrolebinding.yaml

Apply configure files.

kubectl create namespace logging
kubectl apply -f fluent-bit-config.yaml
kubectl apply -f fluent-bit-daemonset.yaml

To check, run command. It should show two pods running in node1 and node2.

kubectl get pods -n logging

To get logs, run command

sudo kubectl logs -n logging [Pod name, like fluent-bit-xxx]