AWS CodePipeline is a fully managed continuous delivery service that helps users automate release pipelines. CodePipeline automates the build, test, and deploy phases of the release process every time there is a code change, based on the release model a user defines.
I think AWS CodePipeline is a great tool for anyone wanted automated deployments in a multi-server/container AWS environment. AWS also offers services like Elastic Beanstalk that provide a more managed hosting & deployment experience. CodePipeline is a good middle ground with solid, built-in automation with enough customizability to not lock people into one deployment or architecture philosophy.
GIT is good to be used for faster and high availability operations during code release cycle. Git provides a complete replica of the repository on the developer's local system which is why every developer will have complete repository available for quick access on his system and they can merge the specific branches that they have worked on back to the centralized repository. The limitations with GIT are seen when checking in large files.
Git has met all standards for a source control tool and even exceeded those standards. Git is so integrated with our work that I can't imagine a day without it.
Overall, I give AWS Codepipeline a 9 because it gets the job done and I can't complain much about the web interface as much of the action is taking place behind the scenes on the terminal locally or via Amazon's infrastructure anyway. It would be nicer to have a better flowing and visualizable web interface, however.
Our pipeline takes about 30 minutes to run through. Although this time depends on the applications you are using on either end, I feel that it is a reasonable time to make upgrades and updates to our system as it is not an every day push.
We didn't need a lot of support with AWS CodePipeline as it was pretty straightforward to configure and use, but where we ran into problems, the AWS community was able to help. AWS support agents were also helpful in resolving some of the minor issues we encountered, which we could not find a solution elsewhere.
I am not sure what the official Git support channels are like as I have never needed to use any official support. Because Git is so popular among all developers now, it is pretty easy to find the answer to almost any Git question with a quick Google search. I've never had trouble finding what I'm looking for.
CodeCommit and CodeDeploy can be used with CodePipeline so it’s not really fair to stack them against each other as they can be quite the compliment. The same goes for Beanstalk, which is often used as a deployment target in relation to CodePipeline.
CodePipeline fulfills the CI/CD duty, where the other services do not focus on that specific function. They are supplements, not replacements. CodePipeline will detect the updated code and handle deploying it to the actual instance via Beanstalk.
Jenkins is open source and not a native AWS service, that is its primary differentiator. Jenkins can also be used as a supplement to CodePipeline.
I've used both Apache Subversion & Git over the years and have maintained my allegiance to Git. Git is not objectively better than Subversion. It's different. The key difference is that it is decentralized. With Subversion, you have a problem here: The SVN Repository may be in a location you can't reach (behind a VPN, intranet - etc), you cannot commit. If you want to make a copy of your code, you have to literally copy/paste it. With Git, you do not have this problem. Your local copy is a repository, and you can commit to it and get all benefits of source control. When you regain connectivity to the main repository, you can commit against it. Another thing for consideration is that Git tracks content rather than files. Branches are lightweight and merging is easy, and I mean really easy. It's distributed, basically every repository is a branch. It's much easier to develop concurrently and collaboratively than with Subversion, in my opinion. It also makes offline development possible. It doesn't impose any workflow, as seen on the above linked website, there are many workflows possible with Git. A Subversion-style workflow is easily mimicked.
CodePipeline has reduced ongoing devops costs for my clients, especially around deployment & testing.
CodePipeline has sped up development workflow by making the deployment process automated off git pushes. Deployment takes very little coordination as the system will just trigger based on what is the latest commit in a branch.
CodePipeline offered a lot of out-of-the-box functionality that was much simpler to setup than a dedicated CI server. It allowed the deployment process to built and put into production with much less and effort and cost compared to rolling the functionality manually.
Git has saved our organization countless hours having to manually trace code to a breaking change or manage conflicting changes. It has no equal when it comes to scalability or manageability.
Git has allowed our engineering team to build code reviews into its workflow by preventing a developer from approving or merging in their own code; instead, all proposed changes are reviewed by another engineer to assess the impact of the code and whether or not it should be merged in first. This greatly reduces the likelihood of breaking changes getting into production.
Git has at times created some confusion among developers about what to do if they accidentally commit a change they decide later they want to roll back. There are multiple ways to address this problem and the best available option may not be obvious in all cases.
