[BUG] Unintuitive output of install_cluster_tools/uninstall_cluster_tools within the context of spec tests.

[svteb]

Describe the bug
While running the spec tests with LOG_LEVEL=info the cluster_setup_spec.cr tests have quite the jarring output, printing out a caught exception, this output comes from the ./cnf-testsuite install_cluster_tools/uninstall_cluster tools command that is called as part of the tests. Without reading and understanding the code, one might assume that the test is somehow failing.

desc "Install CNF Test Suite Cluster Tools"
task "install_cluster_tools" do |_, args|
  Log.info { "install_cluster_tools" }
  begin
    ClusterTools.install
  rescue e : ClusterTools::NamespaceDoesNotExistException 
    raise "#{e.message}. please run cnf-testsuite setup!"          # <-- offending line
  end
end

To Reproduce

1. export LOG_LEVEL=info
2. crystal spec ./spec/cluster_setup_spec.cr
3. Observe the output.

Expected behavior
It is hard to say what the best code practice here is, but I believe that printing out the exception is quite unnecessary and misleading (especially in the context of spec tests), considering it is followed by the remediation: "please run cnf-testsuite setup!". I am unsure if removing the #{e.message}. part could be detrimental elsewhere, if anyone has better ideas (or thinks this is completely irrelevant 😄) please comment.

Jarring spec test output

[macaktom]

Good catch, I looked into it and think that we should remove #{e.message} from that line and if stack trace is needed, then we could just use Log.debug {"#{e.message}"} to print that for debugging purposes.