c85e5260e8
When resizing an instance's flavor, Trove asks the guest to start the datastore via an RPC call and then, to be safe, would check the service_status table in the database to make sure it was updated to RUNNING. The thing is, that second check was not only superfluous, it was causing resizes to fail in the post-conductor world, because while the guest would have sent a message to conductor describing the datastore's current state before it informed taskmanager that it had finished the call, conductor itself might not have updated the database. This commit changes things so taskmanager polls until the service_status is RUNNING after calling the guest. If there is an error the call to the guest will fail anyway. Change-Id: I4e1ca75a150ed58233c21372d21c6337596e43d0 Closes-Bug: 1278282
Trove
Trove is Database as a Service for Open Stack.
To run all tests and PEP8, run tox, like so: $ tox
To quickly run the tests for Python 2.7, run: $ tox -epy27
To quickly run PEP8, run: $ tox -epep8
To generate a coverage report,run: $ tox -ecover (note: on some boxes, the results may not be accurate unless you run it twice)
If you'd like to start up a fake Trove API daemon for integration testing with your own tool, run: $ ./bin/start_server
Stop the server with: $ ./bin/stop_server