Closed luenta closed 11 months ago
Same occurs to me also and when checking the log bundle also found:
Wondering if this project still alive or someone supporting this issues?!?
I hit this issue as well, and seems like 4.14 picks up the cluster-network-operator fix. I was able to deploy 4.14.0-0.okd-scos-2023-10-11-044258 (pre-release) without issues.
The next stable release will probably move to 4.14, which will include this fix as mentioned. We won't backport to 4.13 at this point.
https://github.com/openshift/cluster-network-operator/pull/2086
Am I allowed to try to fix like this or did I touch some holy magic that I shouldn't have?
Thanks for stepping up @luenta! PRs are always welcome, but unfortunately backports require additional QE scrutiny, so it won't be done at this point. OKD 4.14 will be out soon, so please stay tuned.
Tried to install new cluster bare-metal with the latest installer. Got many issues during master bootstrapping. For example sdn pods stuck on recreating, and the logs said: _/bin/bash: line 49: "RHELVERSION:" unbound variable
Found the issue solved here: https://github.com/openshift/cluster-network-operator/pull/2003/commits/9f83883307ce7b6046b5b2d76895e5ddcca24e3b#diff-ec483261857ecc5887c65ea0a5eddfb3f7768ac3cef206e99599bd82e140accb
When I edited the specified yaml s some processes went trough, but eventually they were set default ( for example the sdn by the network operator )
Any idea to workaround? Or will this be fixed soon? Thanks!