OPENSTACK-HELM : INGRESS CONTROLLER PODS STUCK IN PENDING
wilkers.steve at gmail.com
Wed May 29 21:04:42 UTC 2019
As Robson mentioned, please make sure you have the correct labels applied
to your nodes. Depending how you're deploying Kubernetes, you may have to
manually add these labels. The default node selector key/value labels for
the ingress controller can be found here:
>From what I see, you likely don't have these labels applied. Please apply
them and see if this fixes your issue.
On Wed, May 29, 2019 at 2:38 PM Robson Ramos Barreto <
robson.rbarreto at gmail.com> wrote:
> I'm trying with helm too and I have these node labels:
> kubectl label nodes --all openstack-control-plane=enabled
> kubectl label nodes --all openstack-compute-node=enabled
> kubectl label nodes --all openvswitch=enabled
> kubectl label nodes --all linuxbridge=enabled
> kubectl label nodes --all ceph-mon=enabled
> kubectl label nodes --all ceph-osd=enabled
> kubectl label nodes --all ceph-mds=enabled
> kubectl label nodes --all ceph-rgw=enabled
> kubectl label nodes --all ceph-mgr=enabled
> From kubectl describe pod you are able to see the needed node selector
> On Wed, May 29, 2019 at 3:57 PM Akki yadav <yadav.akshay58 at gmail.com>
>> Hello team,
>> I am running the following command in order to deploy the ingress
>> - ./tools/deployment/multinode/020-ingress.sh
>> It makes two pods ingress-error-pages-899888c7-pnxhv and
>> ingress-error-pages-899888c7-w5d5b which remains in pending state and the
>> script ends after 900 seconds stating in "kubectl describe" that :
>> - Warning FailedScheduling 60s (x3 over
>> 2m12s) default-scheduler 0/3 nodes are available: 3 node(s) didn't match
>> node selector.
>> As only default labels are associated with the cluster nodes, do i need
>> to add any specific one ? If yes, Please tell me where and how to add the
>> Please guide me how to resolve this issue.
>> Thanks in advance.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the openstack-discuss