I have a pipeline that contains test steps that produces artifacts like: html reports, logfiles, screenshots/videos etc. Tests are executed in the docker, I can save all the artifacts to a shared folder and then use “Upload Artifacts to S3” to upload all test artifacts to S3. The problem is that there is no link between build and those artifacts. And yes I could make sure that folder in S3 has a build id in its name, but it would still require to manually check build id in harness and then search for corresponding folder in S3. In Jenkins that I was using before archiving arifacts was linking that files with a build, and everyone was able to view reports even without pushing anything to cloud storage. Is it possible to achieve something similar in Harness?
Hey Mateusz
Thank you for your question .
Your request makes total sense, and the good news is that it’s already planned and coming soon.
In our upcoming quarter (Aug-Oct) we are planning to add the option to store artifacts on storage provided by harness, and to make these artifacts accessible (and if text-based - readable) through our UI
Regards,
Nofar Bluestein, CI product manager
Hi, Nofar thats great news. Regarding that readable text-based artifacts, are we talking only about plaintext logs or do you plan to support also hosting html files? To give you an example gatling.io a performance testing tool generates html report with css, images, js charts that need to be rendered to see detailed test results, without that it is just a pile of text, that makes little sense, so such report would need to be hosted just like every static html webpage (and Jenkins does that perfectly).
And if it is not possible to host those files, it would be just fine to have a link to files that were uploaded to S3, especially in case of some GUI tests that can generate large video files, that you may not want to store directly in harness.
Hey ,
the plan is to support different types of files - we want to allow viewing html, screen recordings (UI tests) , text files, images, etc.
Even after we deliver the support for storing files in Harness storage , some customers may still want to upload files to their own storage. and of course , it makes total sense for us to present links if possible, and/or other relevant info, on the upload.
we currently already do that for docker images that are built in the build using the “build and push” steps , but not for non-image artifact.
this is something we intend to work on in the upcoming quarter.