Make it easier for users to locate anomalies in pipelines
Insight
Based on the work done in UX Scorecard Part 2: Understand pipeline health... (gitlab-design#2555 - closed), users cannot quickly locate specific runs of pipelines or jobs that are “not normal” (duration or status was different than usual). This impacts their ability to optimize the pipeline because they must first pinpoint those anomalies themselves.
Supporting evidence
Whenever like we saw that we have very long pipelines. Usually what I would do if I were looking into it is just check which jobs are running the longest and try to figure out, you know, what should I do with these jobs? So a way to identify those jobs is would probably be helpful. But from there, it was usually a bit manual of like going into the logs figuring out why it's taking so long, how we can break it apart. I'm not sure where additional accessibility options would help here. I think the work with identifying things like anomalies in terms of run time is helpful. I definitely often like in, in past jobs ran into cases when a particular pipeline runs much longer in particular cases, but it's hard to identify because you're not keeping track of all the pipelines. I just, you know, happened to be looking at the pipeline's page and see that a particular pipeline ran for two hours instead of, you know, half an hour. And I was like, huh, this is strange. So any way to point out to anomalies would be helpful?
Action
Add some indication that a job or pipeline run is drastically different than it has been recently based on:
- Last X runs
- Failure rate difference
- Duration difference
Resources
Tasks
-
Assign this issue to the appropriate Product Manager, Product Designer, or UX Researcher. -
Add the appropriate Group
(such as~"group::source code"
) label to the issue. This helps identify and track actionable insights at the group level. -
Link this issue back to the original research issue in the GitLab UX Research project and the Dovetail project. -
Adjust confidentiality of this issue if applicable