Modify the kind builder script to perform correct kind cleanup

[KMAnju-2021]

Signed-off-by: KMAnju-2021 km074btcse18@igdtuw.ac.in

Modify the kind builder script and destroy function in kind-setup.sh to perform correct kind cleanup.
If we create a kind cluster by using kind-setup.sh script then we should need to delete cluster using same script.

[KMAnju-2021]

/test-kind-conformance
/test-kind-networkpolicy

[antoninbas]

IMO it would make more sense to have this happen in destroy

We could change the destroy implementation to:

function destroy {
  if [[ $UNTIL_TIME_IN_MINS != "" ]]; then
      clean_kind
  else
      kind delete cluster --name $CLUSTER_NAME
  fi
  delete_networks
}
```

[KMAnju-2021]

done.

[antoninbas]

/test-kind-conformance
/test-kind-networkpolicy

[antoninbas]

/test-kind-networkpolicy

[antoninbas]

Kind / Validate metrics in Prometheus document match running deployment's is failing (I ran it twice), maybe it's somehow related to your PR.
Also, test-kind-networkpolicy keeps timing out, but I don't know if this is related.

[KMAnju-2021]

/test-kind-networkpolicy

[KMAnju-2021]

/test-kind-conformance
/test-kind-networkpolicy

[KMAnju-2021]

/test-kind-conformance
/test-kind-networkpolicy

[KMAnju-2021]

Kind / Validate metrics in Prometheus document match running deployment's is failing (I ran it twice), maybe it's somehow related to your PR.
Also, test-kind-networkpolicy keeps timing out, but I don't know if this is related.

Today i triggered it thrice but I didn't get any error and job finished with success. I think that was a temporary error?

[antoninbas]

Looks like the metrics job failing was a flake.

@KMAnju-2021 as for test-kind-conformance and test-kind-networkpolicy, we should try to investigate flakes and make sure these jobs are as stable as possible