In the last few months, it seems like the CI tests are failing much more often. Re-running them usually fixes the problem, but perhaps there is something we are doing that exacerbates the failures?

See some comments here.