Hi,

Our CI including clusterbot fails with an extremely high incidence with messages like:
"CI fails with Unable to connect to the server: dial tcp: lookup api-int..* no such host"

Sorry for not having researched this further so far, but it seems to be a systemic issue. From what I see in the search, it does happen a lot on AWS, it seems to be happening on GCP, too. I'm not sure about other platforms.

https://search.ci.openshift.org/?search=Unable+to+connect+to+the+server%3A+dial+tcp%3A+lookup&maxAge=6h&context=1&type=bug%2Bissue%2Bjunit&name=&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job

If you change the resolution from 6 hours to 12, to 1 or 2 days up to 7 days, you can see that the failures occur significantly more often within the last 6 hours.

Thanks,

Andreas