Check if you have this patch deployed: https://review.opendev.org/#/q/I15ef91bbec20a8037d47902225eabb3082579705 On Fri, Nov 6, 2020 at 1:03 PM Ionut Biru <ionut@fleio.com> wrote:
Hi,
I wonder if it is because the image is from fedora:rawhide and something is too new in there and breaks the flow?
https://opendev.org/openstack/magnum/commit/1de7a6af5ee2b834360e3daba34adb6b...
On Fri, Nov 6, 2020 at 1:34 PM Ionut Biru <ionut@fleio.com> wrote:
Hi,
I'm using stable/victoria. It works fine with an older image.
On Fri, Nov 6, 2020 at 1:28 PM Spyros Trigazis <strigazi@gmail.com> wrote:
I see, this says kubectl misconfigured not missing. error: Missing or incomplete configuration info. Please point to an existing, complete config file:
I guess you miss some patches: https://review.opendev.org/#/q/I15ef91bbec20a8037d47902225eabb3082579705
Try using an older image of the agent or take the patch above.
Spyros
On Fri, Nov 6, 2020 at 12:01 PM Ionut Biru <ionut@fleio.com> wrote:
Hi,
Not sure if because kubectl is not found but here are the logs, I cannot figure out why it is not working with the new heat image.
journalctl https://paste.xinu.at/jsA/ heat-config master https://paste.xinu.at/pEz/ heat-config master cluster config https://paste.xinu.at/K85ZY5/
[root@cluster001-7dby23lm6ods-master-0 heat-config-script]# kubectl get pods --all-namespaces No resources found [root@cluster001-7dby23lm6ods-master-0 heat-config-script]# kubectl get all --all-namespaces NAMESPACE NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE default service/kubernetes ClusterIP 10.254.0.1 <none> 443/TCP 36m
It fails to deploy and I don't have any services configured.
On Fri, Nov 6, 2020 at 12:43 PM Spyros Trigazis <strigazi@gmail.com> wrote:
Am I missing something? It is there.
Spyros
ubuntu@str-focal-2xlarge-01:~$ docker run -it --entrypoint /usr/bin/kubectl openstackmagnum/heat-container-agent:victoria-dev version --client --short Client Version: v1.18.2 ubuntu@str-focal-2xlarge-01:~$ docker run -it --entrypoint /usr/bin/kubectl openstackmagnum/heat-container-agent@sha256:0f354341b1970b7dacd9825a1c7585bae5495842416123941965574374922b51 version --client --short Client Version: v1.18.2
On Fri, Nov 6, 2020 at 11:33 AM Ionut Biru <ionut@fleio.com> wrote:
I wanted to say, 20 hours ago instead of days.
On Fri, Nov 6, 2020 at 12:33 PM Ionut Biru <ionut@fleio.com> wrote:
> Hi, > > 20 days ago the image was updated and now all clusters fail to > deploy because it cannot find > /usr/bin/kubectl, i think. > > /usr/bin/kubectl apply -f > /srv/magnum/kubernetes/manifests/calico-deploy.yaml --namespace=kube-system > > Some services are installed using full path instead of simply > kubectl. > > Should the image be fixed or reverted the changes to old revision or > we should just fix the scripts? > > -- > Ionut Biru - https://fleio.com >
-- Ionut Biru - https://fleio.com
-- Ionut Biru - https://fleio.com
-- Ionut Biru - https://fleio.com
-- Ionut Biru - https://fleio.com