Through debugging https://issues.redhat.com/browse/OCPBUGS-13810, we discovered that DNS records are not propagating to CI Cluster upstream DNS Servers (or even Google 8.8.8.8 for the matter) within our originally specified time of 5 minutes.

- Increasing the wait time for DNS to resolve to 15-20 minutes also fixed the issue, giving it enough time to resolve.
- The ELB Hostname as specified by the Service-type Loadbalancer, resolves in a reliable manner, the problem seems to be with Route53 DNS records
- Directly querying Google's 8.8.8.8 DNS Server also produced unreliable results (on & off resolving until 15 minutes in or so).
More details can be found in the analysis I wrote up here: https://issues.redhat.com/browse/OCPBUGS-13810?focusedId=22419881&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-22419881

4.11 - 4.17

50% ish

Option 1:
1. Run an old version of TestAWSELBConnectionIdleTimeout (without https://github.com/openshift/cluster-ingress-operator/pull/944 checked in) in CI
2. Try until it fails. This PR https://github.com/openshift/cluster-ingress-operator/pull/940 can help with debugging.

Option 2:
1. Get access to a CI cluster
2. Create a Route53 DNS record in another AWS account/cluster
3. See if DNS record resolves inside of the CI Cluster within 5 minutes

Route 53 DNS Records sometimes take 15-20 minutes to resolve 

Route 53 DNS Records should resolve in other AWS clusters/account within 5 minutes