Is there a way to list multi-stage automated test run results, on a per stage basis, using Azure DevOps GET Api?
If you call /_apis/pipelines/{pl_def_id}/runs , you will get a list of builds from a defined pipeline.
If you get the latest build from that list, and if that build is multi-stage, the test runs do not seem to be hierarchically beneath the latest build, as you might expect.
To get the test runs, when you call /_apis/test/runs, ( https://learn.microsoft.com/en-us/rest/api/azure/devops/test/runs/list ) it only can give you the list of ALL runs within your org+project. There does not seem to be an id that references the multi-staged "test runs" tasks back to the "parent pipeline" having a "build id".
I looked at timeline queries, but they don't seem to have that info either.
HOW can this be achieved with ADO GET api? Am I overlooking the answer?
There is a parameter buildUri in the run list API as you mentioned the doc:Run List
You can use buildUri to get test runs from a specified build of pipeline.
However, if you want to get multi-stage automated test run results on a per stage, I’m afraid that there’s no supported API to achieve it.
You may submit a feature request at website below:
feature request
Related
We have a rather large deployment surface, say 10 apps that get deployed. For patch releases we sometimes deploy only one app and I'd like to have a stage run either after all 10 are deployed or if only one is deployed. A simplified graph looks like the following. The "Do Something" step will only run if all three app stages run and I don't want to have to duplicate it for each app so looking for a better way. I guess that I could live with it if it just ran one time on any successful dependent stage (doesn't need to wait for all of them).
So I think there are a couple options for this. Will need to look at YAML Multi stage release pipelines. Specifically, deployment jobs
First depending on complexity of the "Do something" stage, it could be a job template and loaded into each of the app stages. I realize you mentioned you don't want it to run every time so this is just an option.
The second option is if the pipeline is being ran adhoc you will have the ability to select which stages are ran. Thus you could manually select the app stages to run and select the "do something stage" It will look something similar to this. The hard part will be working on the dependencies:
I'd assumed you'd want the "do something" to be dependent on the success of one of the previous stages.
Run Azure Release Pipeline stage if one of many dependent stages runs
I am afraidthere is no such out of way to achieve this at this moment.
That is because there is no "OR" syntax for the depend on. And we could not add the condition for the depend on.
You could submit this request condition "OR" to our UserVoice site (https://developercommunity.visualstudio.com/content/idea/post.html?space=21 ), which is our main forum for product suggestions. Thank you for helping us build a better Azure DevOps.
As workaround:
The main idea of the solution is: You could try to set depend on for the stage Secure with [], then add a Inline powershell task before other tasks. This task will call the REST API Definitions - Get to monitor whether all the stages in the current release pipeline have inprocess and queue states. If so, wait for 30 seconds, and then loop again until all other stages in the current release pipeline have no inprocess and queue states. Then next execute other tasks will be executed.
You could check my previous ticket for detailed info:
This may really be more a QA question. I'm sort of hamstrung in my organization, I don't have API access for some of the Azure REST calls. I've got a number of test plans with various suites, and reporting the overall current status in terms of passed, failed, etc. is straightforward.
What I'm not able to do easily is get a historical view of all the runs of a given test case. So I can't seem to easily get a metric that says (as an example) we had to run this test case 12 times before we got a pass. Most of the help I'm seeing online references some great ways to get all test IDs if you have API access, I'm not sure if there's another way outside of it.
There is New Test Plans Page in azure devops Previous Features, which allows you to view the historical test result of a test case.
First you need go to "User setting-->Previous Features of your azure devops and Turn on New Test Plans Page
Then Go to your Test plan-->Go to Execute page-->Click 3dot-->Choose View execution History
Then you will see the historical test results in a new page. You can also filter the execution
history as below highlighted
I am trying to invoke the Rest API below for getting list of work items which are mapping to a specific iteration.
https://learn.microsoft.com/en-us/rest/api/azure/devops/work/iterations/get%20iteration%20work%20items?view=azure-devops-rest-5.1
Ultimately I want to invoke this as part of the Approval step in my Release pipeline so that the approver can verify if all work items for a given iteration are in Completed state.
I have two questions:
I don't know how to get the values for {team} and {iterationid}
How do I invoke this API as part of the approval gate. Should I use a Generic service connection? What user name and password do I need to provide?
Any working example here will be really helpful.
To get teams, you can use Teams - Get Teams on Team Project.
Then you can find all team iterations and their ids Iterations - List
You can use gates with azure function. Here is the example: Azure DevOps release gates with Azure Functions, PowerShell and VS Code
Additionally, you may consider to use quality gates with a work item query: Use approvals and gates to control your deployment
I have a bamboo plan that runs on every commit to a github pull request. In that bamboo plan there are a few custom variables on it such as Git Sha, Github Pull Request Number, etc.
I want to write a script that stops all previous builds (multiple concurrent builds) that have the same pull request number -- same custom variable value.
The reason for this is that if someone makes a quick change to their pull request (comments on the review, etc) that we don't have multiple builds running when only the last one is necessary.
I know it is possible to stop a build with a rest request, but I need a way to be able to get all running builds with custom variable value = 27 (pull request number). Once I know this, I can proceed.
At the time of writing, the REST API documentation doesn't list any method of querying the running builds for a particular build variable.
A solution would be to create your own plugin for Bamboo that exposes a REST service that does this query for you, but I don't know which of the Java APIs you would need to use in order to perform that query.
Here is how I solved this ...
You can call /rest/api/latest/result/<plankey>-latest?includeAllStates=true&expand=variables where plankey is the key for the specific Bamboo build plan.
You then loop through the results you get back, looking for a lifeCycleState value that is not Finished, and a custom variable with the desired name to see if it matches the PR number you have.
What information are we able to query on a workflow job? Anything regarding a particular build's stage status (succeeded, failed, hasn't reached yet, aborted, etc.)? I see we can interact with the input step using this method, but where can we find what metadata, if anything, can be obtained about our builds?
The REST exported API for builds (…/job/…/…/api/json?tree=…) is not very extensive yet. You can get some information about nodes in the flow graph (steps, and some associated block nodes—the stuff you see in Workflow Steps). It is possible to extract some information about stages from that, albeit not easily. Much more is available from the Java API.