Keep in mind this will exit after first test failure, and the rest of the tests won't run.
It's easy to make it continue, but then the exit code of this script won't be reported should any of the tests fail:

  set +e
  run_test change_backend_service_test
  run_test failover_test
  run_test remove_neg_test
  run_test round_robin_test

If we want to report the code, we need to do something like

#!/usr/bin/env bash
set -eo pipefail

run_test() {
  echo $1
  if [[ $1 == "good_test" ]]; then
    python -c 'import sys; print("success"); sys.exit(0)'
  else
    python -c 'import sys; sys.exit("fail")'
  fi
}

main() {
  errors=0
  for test in good_test bad_test good_test bad_test
  do
    local exit_code=0
    run_test $test || (( errors++ ))
  done
  echo "Failed test suites: ${errors}"
  if (( errors > 0 )); then
    exit 1
  fi
}

main "$@"

But in this case python must be the last call in run_test, so set command on line 110 must be deleted.

Hah, I didn't actually see the edit here and went off the original email notification only. After some (way too much...) time spent on bash, I came up with (I believe) more or less the same thing, sans the for-loop. But the error message in your snippet here is clearer, so will update to what you provided :)