Azure Pipeline Code Deploy over On-premise Kubernetes Cluster.

 


Pre-requisites: 

1. Azure DevOps account 
2. Kubernetes Cluster 

Agenda:

1. Create Azure Git repository
2. Create a git local repository
3. Create the Dockerfile 
4. Commit and merge to the master branch.
5. Create Build pipeline
6. Create a Deployment Group and PAT for On-Perm server access
7. Create a Service account for Cluster access from Azure
8. Create a Release pipeline.

Create Azure Git repository:

# navigate to repos and create a new empty repo and select generate Git credentials 


Create a git local repository: 
"since our motive to deploy in on-perm server, not going in detail on git commands"
"To Clone and use the demonstrated configuration files, try Git-repo  "
user@server: mkdir azurecicd && git init 
user@server: cat >> EOF > Dockerfile
   FROM centos:7
   MAINTAINER xploritec
   RUN yum -y install httpd
   COPY index.html /var/www/html
   CMD ["usr/sbin/httpd", "-D", "FOREGROUND"]
   EXPOSE 80
   EOF
## create the index.html or download one from 
$ user@server: git add . && git status && git commit -m "first commit"
# output  of git status 
On branch master
 No commits yet
 Changes to be committed:
      (use "git rm --cached <file>..." to unstage)

        new file:   Dockerfile
        new file:   index.html
user@server: git remote add origin https://user@dev.azure.com/user/project/_git/newrepo
$ user@server: git push -u origin --all 



Create an Azure Build Pipeline:

1. Navigate to azure Pipelines

Click on the new pipeline:


Click Azure Repos Git ( Since we choose Azure, git, as SCM):


Select the Repository:


Azure DevOps provides a great feature (Something like S2I) and provide us the appropriate templates to start editing: Here I select Docker.

Click Settings and select your "Docker hub" Service connection and name the repository something like {your docker repo}/demo-project. Save and run.





Azure Release Pipelines:
It's time to make a connection from Azure Pipelines to the On-prem server.
In my case, I have CentOS VM hosted on Oracle VirtualBox. 

1. Starting Virtual machines and check the server's status:



Log on to Azure DevOps:

1. We are going to create a Deployment group and execute the script to make a connection to Azure pipelines for continuous deployment.

2. Create a PAT to provide access for Release management.

3. Time to execute the snippet on the On-Perm server.                                                                                  



4. Check the Deployment group for Virtual machine status:
 

5. Create a service account for cluster access from Azure Pipelines.


6. Create a new release pipeline and add deployment group Job and Kubectl task as below.






Finally, we can see the successful deployment status:




Conclusion: Achieving continuous delivery is quite a process, but things will get in place once we set up for the first time. The same process should follow to add more clusters or deployment groups.









Comments

Post a Comment

Popular Posts