11
Here is my solution: (bitbucket-pipelines.yml)
pipelines:
default:
- step:
script:
- export DOCKER_COMPOSE_VERSION=1.18.0
- export DOCKER_COMPOSE_URL=https://github.com/docker/compose/releases/download/${DOCKER_COMPOSE_VERSION}/docker-compose-$(uname -s)-$(uname -m)
- curl -L $DOCKER_COMPOSE_URL > docker-...
9
--vault-password-file can instead point to an executable script that writes to stdout. This little-known feature should solve your problem.
First, write a simple, executable script that prints a system environment variable, and check that into your source control. Then, use Bitbucket's encrypted variable feature to set that environment variable to your ...
6
You don't make any mention of the scripting language you want to use, so I will talk specifically about the HTTP requests to the BitBucket API:
Assumptions
If you have a BitBucket Repository that has three commits in in it the first and the last are failing the build, the middle is passing:
4768815 ❌
49d7110 ✅
42d357f ❌
Get the list of commits
You can ...
6
We are less likely to see support of native docker-compose integration because pipelines API is a custom implementation of similar functionality, and also because possible security issues, but looks like more features are currently in development to implement more docker-compose-like features in this open issue.
UPDATE: added support for multi-container ...
6
What I really want is to add a comment to the accepted answer, but I don't have enough reputation to do so. So I'll note it here. I've closed BSERV-10344 because it's in the wrong tracker. This question is about Bitbucket Cloud (bitbucket.org), not Bitbucket Server (the behind-the-firewall version). The correct issue tracker for the feature request is at ...
5
We do this at work.
We have a small server, let's call it the receiver, it's the target of the GitHub webhook events. It runs a small application that parses the payload and incorporates logic around how to proceed e.g. create a new server on the infrastructure provider, update the load balancer, deploy to an existing server, destroy the server etc. This ...
5
I believe it is now possible to configure the webhook on per-repo basis:
https://mohamicorp.atlassian.net/wiki/spaces/DOC/pages/381419546/Configuring+Webhook+To+Jenkins+for+Bitbucket+Bitbucket+Branch+Source+Plugin
4
At the moment the docker login and docker push commands are only available to Pipelines Alpha members.
You can read about the features available to Alpha at "Pipelines Alpha".
In order to participate, the above link includes instructions on how you can opt-in to the added functionality.
Personally I received an e-mail from Atlassian a day after signing up ...
4
I am in a similar pair of shoes as you - I am currently, slowly, getting into the DevOps topics after having many, many years of conventional development (and close connections to conventional ops) behind me.
You seem to have read a few articles, or heard a few things. That is good. Let me suggest a practical approach how to go on:
Take your favourite ...
4
Apparently selecting the Close branch option in the very dialogue illustrated in your question will actually delete the branch as well. From
Why only "Close" a branch instead of "Delete" it?:
Ben Tatham Jun 15, 2014
It seems that I was mistaken. Marking a pull request to Close branch
when merged, does indeed Delete it. I'm not sure why they use
...
4
Try running with -d detached flag
docker run -td -p 3000:3000 -p 6379:6379 -p 8983:8983 my_dockerhub/image
# here ^
Containers started in detached mode exit when the root process used to run the container exits. This will prevent the step to freeze at docker run stage.
Docker docs -d flag
4
If you just need generic post-receive hooks, you can just do a regular web hook and hit the build API endpoint. For instance, if the name of your Jenkins server is jenkins.example.com and the name of the Pipeline job to trigger is my_awesome_job, then you can fire off a POST request to https://jenkins.example.com/job/my_awesome_job/build. Depending on your ...
3
Note: this answer comes from my background of building custom solutions, it's not a config-only one, which, if available, would obviously be preferable. But it's maybe something to consider otherwise.
You could teach the script executing the unit test job to persist a per-branch execution state.
When the script is launched it would need to obtain the ...
3
As you are already using Bitbucket, Bitbucket pipelines could be useful.
What languages are supported?
Node, Ruby, Python, PHP, and more. Anything you can run in a Docker
container is supported.
why not jenkins?
Jenkins could be an option as well. If Bitbucket could trigger Jenkins or Jenkins could check whether there are code changes.
3
To expand on the answer above - try using:
branch: "origin/pr/${pullRequestId}/from"
Because in groovy, string interpolation is not done for simple strings - i.e. strings within single quotes. See http://groovy-lang.org/syntax.html#_string_interpolation for details on that.
3
In groovy script, you need to reference environment variables in a different way than in bash.
So probably this line is causing trouble:
branch: ('origin/pr/${pullRequestId}/from')
Try using:
branch: ('origin/pr/' + env.pullRequestId + '/from')
3
I managed to use docker-compose to push images to a remote host by using the following image.
tmaier/docker-compose:latest
This is the basic docker image with docker-compose installed.
My bitbucket-pipelines.yml looks like this:
- step:
image: tmaier/docker-compose:latest
script:
- (umask 077 ; echo $DOCKER_PRIVATE_KEY | base64 -d > ./...
3
One could update the build status in bitbucket as follows:
Adding a build result to a commit
To associate a build result with a particular commit, you need to POST
a JSON object to the build status REST resource at:
https://<bitbucket-base-url>/rest/build-status/1.0/commits/<commit-hash>
The format of the JSON object that ...
3
In a typical Continuous Delivery/Deployment pipeline you would have the following happen:
Developer pushes one or more commits, or, a pull request is merged.
Jenkins automatically builds and executes tests.
If successful Jenkins publishes a deployment package to an Artefact Repository; if failure publishes nothing and notify developers.
Deployment ...
3
Since you have a single released version product one thing to consider would be keeping the Develop branch or not:
keeping it would have the advantage that at any moment you can bring in a hotfix in master even if Develop isn't stable enough for a merge into master. But this means you'd need to have 2 focus points instead of a single one - both branches ...
2
bitbucket-pipelines.yml
image: python:3.8.1
pipelines:
branches:
"**":
- step:
name: Build
services:
- docker
caches:
- docker
- pip
script:
- pip install docker-compose
- docker network create dockernet
- docker-compose build
...
2
On GitHub - you can't. Basically, their PRs are temporary branches, and there's no check if base commit is HEAD of branch to merge in.
This problem is solved with gating systems like Zuul, which puts all PRs into queue to run tests.
You are talking about systems for managing and storing code, while the issue is in CI scope (i.e. Gitlab flow has this issue ...
2
You could:
prune your branches. (guessing you'd have done this by now if feasible.)
implement Jenkinsfile, a code-driven pipeline,
and simply remove the jenkinsfile from the branches you don't want
built or triggered by SCM-polling.
2
You've chosen a fairly difficult stack to support. WAMP is largely used in dev environments, and the windows builds of mysql, php, and apache typically lag their Linux counterparts by a significant amount. You can likely reuse your code, barring any hard coded or absolute paths, on Linux. Once you move to Linux, there are lots of great deployment options, ...
2
Instead of using a symlink that will be stored as is in BitBucket one could use:
- name: Copy docker-compose.yml
copy:
src: ../../../../docker-compose.yml
dest: /path/to/docker-compose.yml
2
After doing some research and comparing various documents it seems that a bot account is basically a real user account and needs a separate email account in order to create such an account as using an email account to register a new account results in, e.g.: This email account is already in use.
Combined with quote retrieved from the link that is defined in ...
2
You can certainly show a code coverage badge in a bitbucket repository, but it would have to be from an external source like jenkins, codeclimate, etc.
According to your link, if you are using pipelines they now show a code coverage badge on the project overview page. If you want to add that to your README or display it somewhere external, that ...
2
I think you're taking the problem on the wrong side, you may start a whole new infrastructure aside and then switch loadbalancers from blue to green, loosing sessions, breaking ongoing transactions etc.
The proper way to achieve what you're doing is coding differently, when A relasease a breaking API ni version n, a should be able to continue to answer call ...
2
Temporary workaround
Instead of calculating changes between builds compare changes against specific branch.
Open job configuration
Go to Source Code Management -> Git -> Additional Behaviours
Add Calculate changelog against a specific branch
Enter name of repository (origin in my case), and branch (master in my case)
Now instead of showing only new ...
2
It is a convenience that Jenkins will pull from git for you. With such a complex setup as needing to pull from two repos you can simply use a pipeline job that runs "sh" to explicitly git checkout the code and pull the submodules:
sh "git checkout ${custom_env_var} ${custom_parameter}
The good news with this slightly manual approach is that you can debug ...
Only top voted, non community-wiki answers of a minimum length are eligible