I'm trying to figure out how to create a task (custom-command, not ant/rake, etc) to perform a deployment of a git-repo to some server/target (in this case Heroku). If I were to do this manually, it's just git push heroku master.
I've created a basic pipeline/stage/job/task (custom-command, in this case a Python script), and single agent. The pipeline has a material (git repo, with a name). Inside the script, I'm printing out os.environ.items() - it has several variables, including the SHA of the latest commit - but no URL for the actual repository.
So how is the agent (or task) supposed to know what repository to deploy?
The pipeline knows the material name, and I've tried passing in an Environment Variable such as ${materialName} (which didn't work). I could hard-code a URL in the task, but that's not a real solution.
Thoughtworks:GO's documentation is shiny, but a bit sparse in the details. I'd have thought something this basic would be well documented, but if so, I haven't found it so far.
When a task runs on an agent, it clones the repository specified in the material (config). The .git/config wouldn't have remote Heroku url and as such git push heroku master wouldn't work.
You would need to add Heroku remote url before you can perform a deployment of your git-repo to Heroku.
git remote add heroku git#heroku.com:project.git
where project is the name of your Heroku project. This is required to be done only once unless you perform a clean working directory every time (in Stage Settings which removes all files/directories in the working directory on the agent, you can see this option from the UI as well: Admin -> Piplelines -> Stages -> Stage Settings Tab) in which case you may have to add the remote url via a task before you run the task to deploy.
Once you've done so, you should be able to use the heroku xxxx commands (assuming you have the Heroku Toolbelt installed on the agent machine you are using for deploying), and should be able to push to Heroku as usual via git push heroku master.
Related
I would like to know if you have any ideas of what happens to the following pipeline I'm developing to build an xcode project.
I'm configuring a pipeline with the xcode build, copy files, and publish artifact tasks.
I have configured the build task, with my xcworkspace path, and set up the scheme
I have all of my repos, in bitbucket, but I have set it up a service connection between my Azure devops, and bitbucket, using an app password, which have all the permissions allowed, I have developed a pipeline to build an Android app, and I don't have problems with that one, just to mention that this one doesn't depends on other separated repos
The thing is that, the developer separated the api client in a repo, from the main repo were the iOS code is, so when Azure tries to clone this repo, it can't be done, due to a host key verification, he told me that this repo is dependency of the main iOS repo.
This is the error
I'm new to Azure devops, but I have searched for solutions, and nothing. I understand that I have to configure an SSH key, but, I'm using a pipeline with a hosted microsoft agent. Should I add a task to actually set up this SSH key to the agent? Or it is only permission problems?
The things I have done, are:
Creating the app password again, with all permissions set
Change my project path, and use the correct scheme, cause I was using the wrong one. But there's no difference
The only thing I have left is to add the user from my Azure devops to a group of admins, who actually have permissions to the repo Azure is trying to clone. But I need to open a ticket with my partner, cause I cannot logged into the mail account.
I am attempting to create a pipeline in our Azure DevOps org that will automatically 'pull' code down to a single remote server whenever a push request is sent to the master branch of my particular repo. I am having a difficult time understanding the entire process and what I actually need to accomplish this relatively simple pipeline.
Currently, my remote server has a folder on the C: drive with various .ps1 files. I would like to turn this into my repo and install the self-hosted agent on this same server so that way anytime I push something to the master branch on my local server it will automatically be pulled down to my remote server and the scheduled tasks I have running will pick be running the most up to date code.
I believe what I need to do first is install a self-hosted agent on my remote server. I am not completely sure though if this agent is suppose to be a deployment agent or a build agent.. or both? Since I am not technically building a project, but rather simply overwriting .ps1 files, I imagine it should only be given permissions for a deployment agent.
Something else I can't wrap my head around is how I specify the location of my repo on the remote server. Can I define this dynamically or do I need to specify in my path the target path of that specific repo?
According to your description, you could simplify your requirement to be: copy files from a source folder to a target folder on a remote machine over SSH using Copy Files Over SSH task, and then run related Git commands like the following.
cd repo_directory
git add .
git commit -m "upadte"
git push
Thus this remote repo is updated using SSH Deployment task.
In addition, you need to deploy a self-hosted build agent because it can be used in build pipeline.
Finally, configuring a schedule trigger for this build pipeline.
I have a unauthenticated (in vsts) external git repository that works in an local agent pool (who is auth) that is external too. Our build pipepline is the connection with our certificates that does this solution work.
The problem here is that there is no possibility to trigger automatically when someone push some changes on the master branch on this external unauthenticated (in vsts) git repository.
For this trigger option, there is only this configuration:
But after this, if someone push an commit, nothing happen.
Is there a limitation? Any configuration that I need to get this working? At now, it's always done manually to stat the build pipeline.
For the continuous integration with External Git, VSTS will connect to your external repository and check if there is new changes by the polling interval time you set. So you must configure the authentication in VSTS so that VSTS can access to your Git Repository to query new changes. When you configure the external git repository, it should ask you to provide user and password/token, enter the auth information there.
I need to manage a few servers that run code that is currently being deployed there as a couple of git repositories. I would like to be able to store in the project's repository the parts (if not all) of the playbook that is relevant for the repository. For example, the list of package dependencies, virtualenv requirements, configuration templates. This will also allow those to change in a per branch/commit way. Meaning I can make sure that if I need to deploy a specific branch/commit, playbook that is correct for that commit is being used, if, say, the configuration template being used changed.
It seems like the only solution is to checkout the git repository locally. Is it possible in ansible to tell it to run a remote play book (from the git repository that is being checked out on the server)? I was thinking of having ansible run a ansible using a local connection on the remote host, I haven't tried it to see if this will actually work out.
How do people manage to use ansible for continuous deployment based on git without some mechanisms for running a remote playbook?
Take a look at ansible-pull.
It pulls the repo and executes playbook.
I'm wondering if there are any convenient ways to automate deployment of code to a live server in GO, either standard built-in methods, or otherwise.
I want something google app engine like, I just run the command and it uploads to the server and triggers a restart.
(Ultimately I want a git commit to trigger a rebuild and redeploy, but thats for down the track in the future)
I recommend Travis CI + Heroku.
You can deploy to heroku directly with just a git push, but I like to use Travis to build and run the tests before that.
There are some guides online but I'll try to go directly to the point:
What you will need?
Github account
Travis account (linked with github, free if open source)
Empty Heroku app (Free dyno works great)
Setup
In your github repo, create the following files:
.travis.yml (more info on the Travis CI documentation)
Procfile
.go-dir
After that go to your Travis account, add your repository and enabled the build for it.
Here is a sample minimal config file content (based on my app that I deploy to heroku):
.travis.yml
language: go
go:
- tip
deploy:
provider: heroku
buildpack: https://github.com/kr/heroku-buildpack-go.git
api_key:
secure: <your heroku api key encripted with travis encrypt>
on: master
Procfile
worker: your-app-binary
.go-dir
your-app-binary
Procfile and .go-dir are heroku configs so it can vary if you are deploying a web app, you can read more at the heroku documentation
One important and easily missed point is the build pack, without it the deploy will not work.
Read the Travis docs to see how to encrypt the heroku key
How it works?
Basically, every push to your repository will trigger the Travis CI build, if it passes it will deploy the app to heroku, so you set this up once and build + deploy is just a push away ;)
Also Travis will build and updated the status of all Pull Requests to your repository automagically.
To see my config and build, please take a look at my Travis build and my repository with my working configs