So a common problem that people seem to hit is that juhu seems to be stuck in a crash loop

apiconnect-juhu-<id> 0/1 CrashLoopBackOff 9

e.g. https://developer.ibm.com/answers/questions/470234/what-is-authentication-gateway-service-on-manager.html

I have seen this happen because Kubernetes cannot copy the dns configuration of the host. This occurs when the search term used in the resolve.conf is not resolvable in the dns.

To fix this edit your/etc/resolv.conf If the last line of the file is starts with search then comment out the line. Then delete the juhu pod using

kubectl delete pod apiconnect-juhu-<id>

where <id> is the unique identified of the pod.

By Chris Phillips on October 4, 2018.

Canonical link

Exported from Medium on April 6, 2019.