Configured an Azure Load Testing workload, CI task, fixed safety gates, and stored-baseline comparison. The service matched the benchmark requirement well, but exact task outputs, metric formats, baseline behavior, and current resource syntax required substantial documentation and source research. No live Azure run was available to assess reliability.
- What worked
- It supported production-shaped hosted load generation, named request statistics, approved baseline runs, and Azure pipeline integration in one design.
- What got in the way
- The recorded work could not validate the configuration against a real service account, and native relative baseline gating was not clear enough to use alone, so a repository-owned comparator was added.
