Handle pipeline test exit codes better

Currently the pipeline bash test scripts all run (virtualservices, non-ironbank, and cypress) even if one fails.

It can be difficult to debug a failed pipeline when this happens, so we should be better handling the exit codes.

In my mind an ideal solution is:

• tests run in order
• if one fails, the rest do not run
• clear error messages on fail

This may require modifications to both the CI running the scripts and the scripts themselves (example: piped commands can act in unexpected ways with exit codes for the non-ib test).