I was trying to install Weave Cloud on a cluster remotely. The DNS check failed, so I thought I'd shell into the controller node directly and install Weave Cloud there (perhaps it is some firewall issue).
When I tried to do that, I got another error—strangely formatted, and with no suggested path to recovery.
When I tried again after waiting a while, I still got the DNS check failure
$ curl -Ls https://get.weave.works | sh -s -- --token=REDACTED
Downloading the Weave Cloud installer...
Preparing for Weave Cloud setup
Checking kubectl & kubernetes versions
Connecting cluster to "ATD preparation" (id: dapper-star-44) on Weave Cloud
Installing Weave Cloud agents on jml-atd-kubeadm at https://35.185.96.52:6443
Performing a check of the Kubernetes installation setup.
There was an error while performing a DNS check: timed out during DNS check
Please check that your cluster can download images and run pods.
jml@gallantry:~/go/src/github.com/weaveworks/atd-poc (master) (0)
$ gcloud compute ssh controller-0
Last login: Fri Apr 20 11:19:45 2018 from 82.21.187.81.in-addr.arpa
[jml@controller-0 ~]$ curl -Ls https://get.weave.works | sh -s -- --token=REDACTED
Downloading the Weave Cloud installer...
Preparing for Weave Cloud setup
Checking kubectl & kubernetes versions
Connecting cluster to "ATD preparation" (id: dapper-star-44) on Weave Cloud
Installing Weave Cloud agents on kubernetes at https://10.142.0.2:6443
Performing a check of the Kubernetes installation setup.
There was an error while performing a DNS check: Error from server (AlreadyExists): object is being deleted: pods "launcher-pre-flight" already exists
Full output:
Error from server (AlreadyExists): object is being deleted: pods "launcher-pre-flight" already exists
Please check that your cluster can download images and run pods.
The DNS check failure is a misconfiguration in my cluster that I'm still resolving. The '"launcher-pre-flight" already exists' error is probably a bug in launcher.
I was trying to install Weave Cloud on a cluster remotely. The DNS check failed, so I thought I'd shell into the controller node directly and install Weave Cloud there (perhaps it is some firewall issue).
When I tried to do that, I got another error—strangely formatted, and with no suggested path to recovery.
When I tried again after waiting a while, I still got the DNS check failure
The DNS check failure is a misconfiguration in my cluster that I'm still resolving. The '"launcher-pre-flight" already exists' error is probably a bug in launcher.