Your understanding is correct, an artifact in the Jenkins sense is the result of a build - the intended output of the build process.

A common convention is to put the result of a build into a build, target or bin directory.

The Jenkins archiver can use globs (target/*.jar) to easily pick up the right file even if you have a unique name per build.


An artifact can be any result of your build process. The important thing is that it doesn't matter on which client it was built it will be tranfered from the workspace back to the master (server) and stored there with a link to the build. The advantage is that it is versionized this way, you only have to setup backup on your master and that all artifacts are accesible via the web interface even if all build clients are offline.

It is possible to define a regular expression as the artifact name. In my case I zipped all the files I wanted to store in one file with a constant name during the build.


Also, does Jenkins delete the artifacts after each build ? (not the archived artifacts, I know I can tell it to delete those)

No, Hudson/Jenkins does not, by itself, clear the workspace after a build. You might have actions in your build process that erase, overwrite, or move build artifacts from where you left them. There is an option in the job configuration, in Advanced Project Options (which must be expanded), called "Clean workspace before build" that will wipe the workspace at the beginning of a new build.