As part of QA pipeline(in Jenkins), goal is to automate provisioning and configuration of a VM to run the QA tests.
Jenkins pipeline can trigger Terraform code to automate provisioning of VM and ansible code for configuration of a VM, but, issues like rollback, error handling is not easy unless we use some vendor specific template like AzureResourceManager template.
So, with Jenkins pipeline,What should be the best approach to provision and configure a VM in Azure cloud? we write pipeline scripts for jenkins pipeline...
As the goal is to know the best approach to automate provisioning and configuration of a VM to run the QA tests so I would go with simple jenkins pipeline script by leveraging Azure CLI commands in it.
To be precise, I would just add an Azure service principal to Jenkins credential. And then write simple Jenkins pipeline script by having 'withCredentials([azureServicePrincipal('SERVICEPRINCIPALCREDENTIALID')])' and then by using 'sh' part to have Azure CLI command to provision and configure VM. For illustration related to this you may refer https://learn.microsoft.com/en-us/azure/jenkins/execute-cli-jenkins-pipeline#add-azure-service-principal-to-jenkins-credential.
Regarding the issues like rollback and error handling when going with the approach of having Jenkins pipeline that triggers Ansible code (with or without using ARM templates) that can automate provisioning and configuration of a VM to run the QA tests, (you might already be aware of this but wanted to let you know that) for certain types of tasks you may write custom modules that can leverage the error handling functionality and in few scenarios you may leverage 'failed_when' option. Also you may leverage 'blocks' functionality by which you can define a set of tasks to be executed in the rescue: section. This 'blocks' functionality specially should help in enabling us to get the things rolled back.
Hope this helps!! :)
Related
I have recently started to mess about with Jenkins and am unsure how to deploy my web app to a basic server. I've gotten into the Pipeline (https://jenkins.io/doc/book/pipeline/) and it seems like a fantastic way to work.
Where I'm a bit stuck is in two spots:
Once my repo is in my workspace within Jenkins, how do I prep it so I am only deploying the files necessary for the application? For example, I don't need my src/ directory or my Vagrantfile when I'm deploying things.
How do I deploy my app to the server? I see examples all over the place, but I am getting a bit lost since there seems to be so many ways to do this. I'm assuming scp or something like that...?
To build off of #2, is there a way to deploy web apps as transactions (in one shot) rather than file-by-file?
Please let me know if I can provide any information for potential answers!
I can't speak to your specific use case but a common way to do this is the build-and-deploy model, where you will have 2 Jenkins jobs. The "build" job will check out from source, run build commands such as maven or make, and lastly will "archive" the build artifacts. The latter is an option under the 'post-build actions' tab at the bottom.
In the "deploy" job, you will grab the artifacts of your choice. You can fetch a single file, all of them, and everything in between. This requires use of the 'Copy Artifact' plug-in and it allows you to copy files generated by other jobs. Now you can run your usual deploy script in the 'Execute Command' box. Most command line paradigms are supported out of the box such as setting environment variables.
The instructions above assume that you want to run your application off of a host that you've provisioned as a Jenkins slave.
Use artifacts as mentioned by Paul Back, or a 3rd party artifactory server as in video
This is always tricky and error-prone. Why not spin up a fresh server with new release (humanly verified once)
Jenkins & Ansible is the answer here. This is how I deploy to production, since I am in no need to use anything like Docker (too many issues with particular app) so have to run the app natively. Quick example would be
You monitor a specific branch in gitlab / github or whatever else and then call a webhook on push / merge etc on that branch, at this point you deal with anything you need to do by running a playbook on the jenkins job that monitors that branch (jenkins).
in my case jenkins and ansible run on the same server. Jenkins runs the ansible playbook that does whatever I need to do.
for example with ansible, I copy certain files that need to be there, run configs / change filenames etc. setup nginx, run composer,
you get the point.
My development setup is such that for every svn checkin code is built,unit tested, packaged and published in Artifactory. Now I want to automate my deployment process & run integration(Selenium) test as part of this process. I am thinking of using Puppet to managed the deployment
Is puppet the correct tool for this
What is the process I should use to trigger puppet master to initiate a fresh installation on agents, I couldn't find any Jenkins plugin that would actually trigger puppet. One option is to call
puppet apply ...
as a Jenkins post build task
Any suggestions welcome, thank you.
Have a look at this Selenium Jenkins article from Saucelabs, a service that automates cross-browser testing. Though they are a vendor with a service to sell, the article covers how to do Selenium testing yourself with Jenkins. It also exposes common pain points you are likely to run into with this approach.
A Puppet master doesn't serve the function of orchestrating client convergences. Take a look at Mcollective. This is a tool that will allow you to trigger puppet runs on target systems from a Jenkins agent via script commands.
Some Mcollective getting started material:
http://www.slideshare.net/PuppetLabs/presentation-16281121
http://puppetlabs.com/mcollective
Recently, in our company, we decided to use Ansible for deployment and continuous integration. But when I started using Ansible I didn't find modules for building Java projects with Maven, or modules for running JUnit tests, or JMeter tests.
So, I'm in a doubtful state: it may be I'm using Ansible in a wrong way.
When I looked at Jenkins, it can do things like build, run tests, deploy. The missing thing in Hudson is creating/deleting an instance in cloud environments like AWS.
So, in general, for what purposes do we need to use Ansible/Jenkins? For CI do I need to use a combination of Ansible and Jenkins?
Please throw some light on correct usage of Ansible.
First, Jenkins and Hudson are basically the same project. I'll refer to it as Jenkins below. See How to choose between Hudson and Jenkins?, Hudson vs Jenkins in 2012, and What is the most notable difference between Jenkins and Hudson from a user perpective? for more.
Second, Ansible isn't meant to be a continuous integration engine. It (generally) doesn't poll git repos and run builds that fail in a sane way.
When can I simply use Jenkins?
If your machine environment and deployment process is very straightforward (such as Heroku or iron that is configured outside of your team), Jenkins may be enough. You can write a custom script that does a deploy as the final build step (or a chained step).
When can I simply use Ansible?
If you only need to "deploy" without needing to build/test, Ansible might be enough. For instance, you can run a deploy from the commandline or using Ansible Tower. This is great for small projects, static sites, etc.
How do they work together?
A good combination is to use Jenkins to build, test, and save artifacts. Add a step to call Ansible or Ansible Tower to handle the actual deployment process. That allows Ansible to handle machine configuration and lets Jenkins handle the CI process.
What are the alternatives to Jenkins?
I strongly recommend Thoughtworks Go (not to be confused with Go the language) instead of Jenkins. Others include CruiseControl, TravisCI, and Integrity.
Ansible is just a "glorified SSH loop".
CI is not only the software running, but the whole process of how success and failure is handled, who gets notification, and how the change is merged into the target version control.
If we only focus on the software, CI is a reactive scheduler triggered by code changes, and triggering typical build-validate-release-deploy sequence of "steps".
So in respect of software, Ansible without additional "sugaring" is just a toolkit to run things, which can be those very steps, but it is not CI.
The Ansible (without tower) totally lacks this reactive nature.
If you want to marry Ansible with CI, you can.
Ansible tower is a very Ansible oriented scheduler, but if you need CI software, I think you not necessarily need it. Any CI app capable of running shell script would be capable to launch Ansible playbooks.
Yet unlike Ansible tower - CI tools know to display test reports of all test frameworks, trigger notifications, etc.
Ansible tower can make sense in a complex environment with lots of groups touching Ansible code... The truth is I haven't seen a single real reason to pay for it. But if a manager liked the web interface nothing can stand "but others use it" logic.
I suspect the concept of Ansible tower was in response to puppet enterprise.
:)
I'm running Jenkins on one server and want to use chef and automatically install a snapshot (including runtime artifacts etc) on a separate server.
Currently Jenkins will use ssh to invoke chef on the seperate machine. Is there a better way?
Maven is also involved in this.
I've found that majority of "Deploy" type plugins are lacking in customization. We use "Execute" (bash or batch) build steps to trigger deployment scripts on remote machines (written in house, be they Puppet, Chef, or plain bash/batch).
The correlation between builds and deployments is achieved through "Promotions" and explained in detail here:
How to promote a specific build number from another job in Jenkins?
We are using Hudson as a CI tool. At present we are needed to use Jenkins, to deploy the build to Stage, Prod environment. What is the best aproach we should follow.
I know about promote buld plugin, but the issue is authentication. I want whevener we need to promote a build to deploy to Stage or Prod, it should ask for netqwork credential first. And then the promote job should execute the Batch command using the creadential supplied. At present, the promote plugin, runs using the credentials which the Tomcat server is configured to run.
Same issue with Build Pipeline plugin.
I want only dev or even hudson admin also should not be able to execute the promote build unless credential supplied. (We have windows 2008 r2 OS)
Can you please help me in resolving the issue. so that basically whenever a user click on Promote build to QA\Stage\Prod the plugin should ask for credential or should use the logged on users credential and execute the batch script using the logged users credential only and not by using the credentials of the account with which the tomcat server is configured.
Can you please help me?
Please suggests us the best aproach for making automated build on prod\stage.
For deployment I normally use SSH, Private/Public keys takes care of the authentication problems normally associated with running commands on other servers.
SSH is normally associated with unix based systems, but it does support windows.
Finally, I would recommend considering decoupling your build system (jenkins) from the system performing the deployment by using an intermediate repository. See the following answer for more details:
Jenkins : how to check out artifact from Nexus and Deploy on Tomcat-