Story #3021

Updated by Tom Clegg over 7 years ago

Places to look include:
* Performance of Workbench itself. Are some tests simply revealing performance problems in the application?
* API server setup/teardown during workbench tests. (This only happens once per suite since rails4, so perhaps it's not a big contributor.)
* API server setup/teardown during python tests.

Tools to fix this include:
* API server hook that resets test fixtures to a pristine state. Python/Workbench integration tests should use this instead of starting and stopping API server many times.
* Refactor "run_test_server" to run as a wrapper around an arbitrary test suite. It can set an environment variable to communicate to the test suite how to push the "reset fixtures" button (path to a trigger file, PID to send a USR1 signal to, or whatever).