Dynamic Plugins in degraded state since 4.14.0-0.okd-2024-01-06-084517 update

[globaldyne]

After upgrading to 4.14.0-0.okd-2024-01-06-084517 the dynamic plugins are in degraded state
mce
acm
monitoring-plugin

Disable/enable didn't help

Any ideas?

[melledouwsma]

Please see #1852 (comment).

[vrutkovs]

Also known issue - https://issues.redhat.com/browse/OCPBUGS-26983

[laralar]

Same issue with 4.17.0-0.okd-scos-2024-11-07-025119

I noticed

Failed to get a valid plugin manifest from /api/plugins/kubevirt-plugin/
r: failed to send GET request for "kubevirt-plugin" plugin: Get "https://kubevirt-console-plugin-service.kubevirt-hyperconverged.svc.cluster.local:9443/plugin-manifest.json": EOF

Failed to get a valid plugin manifest from /api/plugins/monitoring-plugin/
r: failed to send GET request for "monitoring-plugin" plugin: Get "https://monitoring-plugin.openshift-monitoring.svc.cluster.local:9443/plugin-manifest.json": EOF

Failed to get a valid plugin manifest from /api/plugins/networking-console-plugin/
r: failed to send GET request for "networking-console-plugin" plugin: Get "https://networking-console-plugin.openshift-network-console.svc.cluster.local:9443/plugin-manifest.json": EOF

In the following advisor I saw:
https://access.redhat.com/solutions/7049163

Starting with Red Hat OpenShift Container Platform 4.14, the monitoring pages in the Observe section of the Red Hat OpenShift Container Platform web console are deployed as a dynamic plugin. With this change, the Cluster Monitoring Operator (CMO) is now the component that deploys the Red Hat OpenShift Container Platform web console monitoring plugin resources in the openshift-monitoringnamespace.
Customers applying NetworkPolicy in namespace such as openshift-monitoring (which Red Hat does not recommend doing), are advised to adjust the NetworkPolicy in openshift-monitoring to allow traffic from openshift-console namespace for the service called monitoring-plugin.openshift-monitoring.svc.cluster.local on port 9443.

[glennodickson]

SImilar issue here with 4.18. and have noted here for any resolution or suggestions.

[laralar]

My particular case was a DNS issue. My dns was including search everything as search domains. That was taking priority over fqdn in kubernetes. The search domains should only be for *apps...

https://www.reddit.com/r/openshift/comments/1bdpion/comment/m1w2o31/?utm_source=share&utm_medium=mweb3x&utm_name=mweb3xcss&utm_term=1&utm_content=share_button

[glennodickson]

What corrective action did you apply ?

…

On Wed, 14 May 2025 at 18:34, laralar ***@***.***> wrote: My particular case was a DNS issue. My dns was including search everything as search domains. That was taking priority over fqdn in kubernetes. The search domains should only be for *apps... https://www.reddit.com/r/openshift/comments/1bdpion/comment/m1w2o31/?utm_source=share&utm_medium=mweb3x&utm_name=mweb3xcss&utm_term=1&utm_content=share_button — Reply to this email directly, view it on GitHub <#1854 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA5ROAI5JFFBIGBK5UUIL4D26ML2RAVCNFSM6AAAAABROYG42SVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMJUGMZDOOI> . You are receiving this because you commented.Message ID: ***@***.***>

[laralar]

As i mentioned. In my case i had too many search domains. I don't know if it's your case. It had to do with the DNS configuration. I was using another DNS service and there were too mmany search domains being assigned by the DHCP . It had nothing to do with the okd installation.

The DHCP should not give any search domains

Remove all search domains from your DNS server. There should be no *.domainname

For example of your domain name is

okd.example.com and DHCP provides IP addresses in that domain. Okd.example.com should not have any @ entry. My DNS was propagating okd.example.com as search domain. Therefore that was taking precedence over the plugin fqnd

Configure a DNS in a VM exactly as the documentation instructions. And use only that DNS server.

[glennodickson]

Hi Thanks heaps for your description. I've previously installed and run a cluster for sometime with version 4+. After upgrading to 4.17 I've had a plugin issue where the console would come up however, without the network and Observe menu items. Then upgraded to 4.18 and found the console unable to come up at all in the browser. All using the exact same configuration for DNS etc as I use a script to create for consistency. it certainly is due to the NDOT=5 and appending the domain name to make a FQDN when resolving the monitoring and network service. I'm not sure why it is having a problem with only those services and not the entire cluster? Since I use a script to create and configure DNS then it is the 4.14 dynamic plugin change of version 4.14+ which is the only thing which has changed. Any ideas would be appreciated. Regards Glenn

…

On Fri, 16 May 2025 at 00:55, laralar ***@***.***> wrote: As i mentioned. In my case i had too many search domains. I don't know if it's your case. It had to do with the DNS configuration. I was using another DNS service and there were too mmany search domains being assigned by the DHCP . It had nothing to do with the okd installation. The DHCP should not give any search domains Remove all search domains from your DNS server. There should be no *.domainname For example of your domain name is okd.example.com and DHCP provides IP addresses in that domain. Okd.example.com should not have any @ entry. My DNS was propagating okd.example.com as search domain. Therefore that was taking precedence over the plugin fqnd Configure a DNS in a VM exactly as the documentation instructions. And use only that DNS server. — Reply to this email directly, view it on GitHub <#1854 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA5ROAKUBK46IM4WSWNMSMD26TBJFAVCNFSM6AAAAABROYG42SVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMJWGE2DONI> . You are receiving this because you commented.Message ID: ***@***.***>