Hi, Thanks for pointing out the story, It help me to fix it. Please review: https://review.opendev.org/#/c/761897/ On Mon, Nov 9, 2020 at 12:43 PM Spyros Trigazis <strigazi@gmail.com> wrote:
Please follow/comment https://storyboard.openstack.org/#!/story/2007591
Thanks, Spyros
On Mon, Nov 9, 2020 at 11:30 AM Ionut Biru <ionut@fleio.com> wrote:
Hi,
I tried to run manually from inside the heat container the same commands and they worked fine.
I think it doesn't inherit the values from /etc/bashrc
On Fri, Nov 6, 2020 at 2:40 PM Ionut Biru <ionut@fleio.com> wrote:
Hi,
yes, it's deployed: https://paste.xinu.at/pEz/#n855
On Fri, Nov 6, 2020 at 2:36 PM Spyros Trigazis <strigazi@gmail.com> wrote:
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
-- Ionut Biru - https://fleio.com
-- Ionut Biru - https://fleio.com
-- Ionut Biru - https://fleio.com