Matus Mrekaj

Yes you're right, originally the issue was fixed when we switch to cilium with eBPF but that introduce other problems and we had to fallback to kube-proxy where this issue...

This issue has not been seen since https://github.com/berops/claudie/pull/1366 It possible that other bug fixes might have also fixed it. I'll close this as done for now, If it will resurface...

@bernardhalas, this is more of a user error, you have created a secret that has a whitespace (newline) at the end of the subscription ID (input manifest that you posted...

any updates @bernardhalas ?

This can be reproduced for examply by executing: ```bash echo "....secret...." | base64 ``` to correctly base64 the secret without the implicit newline from echo: ```bash echo -n "...secret..." |...

update kube-one to never versions to support hight kubernetes version in claudie manifest

> fire up provider CLI (from an image) and do the login This does not make much sense, would we spin up multiple images each for a unique provider ?...

encountered this once more, couldn't find out why. a simple `killall ansible-playbook` resolved this. ```bash bash-5.1$ ps PID USER TIME COMMAND 1 nobody 0:00 ./services/ansibler/server/server 26 nobody 0:00 [ssh] 27...

most of the test-sets got stuck on the `install-wireguard` playbook during the `TASK [./wireguard : Install Wireguard and dependencies]` task

encountered this bug again with `test-case2` same reason, the ansible playbook was stuck in install wireguard and a large number of ssh processes and ansible-playbooks were spawned, as described in...