This 1.38 release fixes two major bugs in SQUAD when communicating
with LAVA backends:
1. When a testjob fails to fetch, and LAVA lab is out,
SQUAD should handle the outage an give up trying to
fetch such jobs.
2. When LAVA jobs fail due to "metadata is too large",
the results api won't return a valid metadata field,
thus causing SQUAD to break on that too
Also this release includes a health-check endpoint that
shows the availability of most of SQUAD's services.
Complete list of changes going in:
* Dockerfile: add suport to django-health-check
* Dockerfile: fix django-storage error on closing s3 file
* ci: backend: lava: handle null error_msg
* ci: backend: lava: raise TemporaryFetchIssue on fetch timeout
* frontend: build-nav: add test summary to build header
* frontend: project_settings: gather local options
* health_check: allow health-check endpoint to staff user only
* setup.cfg: ignore "C101 Coding magic comment not found" flake8 errors
* travis: disable testing on real LAVA