continuous-integration

问题 I'm trying to figure out the best way to have my API documentation be the source of truth and use it to validate the actual Java REST code ideally through integration testing or something of that sort. We're using the contract first or consumer contract type of approach, so we don't want the documentation to be generated from annotated code necessarily and updating every time a developer makes a change. One thought has been to use Swagger, but I'm not sure how best to make it be used for

问题 I'm trying to figure out the best way to have my API documentation be the source of truth and use it to validate the actual Java REST code ideally through integration testing or something of that sort. We're using the contract first or consumer contract type of approach, so we don't want the documentation to be generated from annotated code necessarily and updating every time a developer makes a change. One thought has been to use Swagger, but I'm not sure how best to make it be used for

问题 I'm trying to figure out the best way to have my API documentation be the source of truth and use it to validate the actual Java REST code ideally through integration testing or something of that sort. We're using the contract first or consumer contract type of approach, so we don't want the documentation to be generated from annotated code necessarily and updating every time a developer makes a change. One thought has been to use Swagger, but I'm not sure how best to make it be used for

问题 Which CI server do you use for Flex? My main contenders are currently Hudson TeamCity Both seem to be popular on the general level but have you got some direct experience with using one or both with Flex development? Or would you recommend some other tool? 回答1: We use Hudson in combination with SVN and it works quite well. Both the Java code and the Flex code is built using Maven and the Flex Mojos plugin. The build also runs the unit tests for the Flex code and it integrates the results in

问题 Which CI server do you use for Flex? My main contenders are currently Hudson TeamCity Both seem to be popular on the general level but have you got some direct experience with using one or both with Flex development? Or would you recommend some other tool? 回答1: We use Hudson in combination with SVN and it works quite well. Both the Java code and the Flex code is built using Maven and the Flex Mojos plugin. The build also runs the unit tests for the Flex code and it integrates the results in

问题 I have a server where our test cases run for all API, which is on the compute engine of GCP. How can I connect it from cloud build CI/CD pipeline so that the CI/CD stage passes only on 200 response status code from the server? GCP says to create a custom build step (here). Docs are not very clear 回答1: You have 2 solutions. You can effectively create a custom step. Build a container, finish it by an ENTRYPOINT which will be invoked in the Cloud Build pipeline You can perform a curl call inside

问题 I am trying to use gitlab/fastlane to build my project. When i execute the scripts in the default project they work correctly, but when the runner checks out the project the build fails with the message "Task 'assembleRelease' not found in root project 'projectname'". When i open the CI project location in Android Studio I have to execute "sync project with gradle files" before build is available. After I have done this i can execute ./gradlew assembleRelease (and the fastlane script also

问题 I'm using the Gitlab release API in the gitlab-ci.yml to be able to automatically create a new release when deploying. Simply putting a curl request like here in the docs works just fine. For the description, the docs state that markdown is allowed, which is great. However, I can't seem to figure out or come up with an idea to load a description from a markdown file within the curl request. I've already tried storing the content of the markdown file in a variable in the gitlab-ci.yml prior to

问题 We are doing continuous integration at our company with TeamCity and we have unit tests running at every commit (1 min window). Lately, we are debating on how long a batch of unit tests should last but the shortest the better. However, I would like to know what is the best practice for the length of a batch of unit tests? 回答1: You could build priorities into your unit tests, and only use a subset as a check-in gate (Build Verification Test, or BVT). Run lower priority tests less often (e.g.

问题 I am using a local BitBucket server with Git. I am trying to get started with Continous Integration so I've set up a local Jenkins instance. The goal is to have Jenkins check out pull-requests and build the project, and then report back to BitBucket with the result. In BitBucket I am using Webhook to Jenkins for Stash which notifies my Jenkins instance each time a pull-request is created/updated. In Jenkins i am using Stash pullrequest builder plugin to have Jenkins checkout the pullrequest,