Frequent timeout when listing some resources but kubectl works

[maxsxu]

Describe the bug

The K9s unable to list the resources, but kubectl can do.

Here is the debug log for listing argo workflow templates:

�[2m5:17PM�[0m DBG No DAO registry entry. Using generics! �[2mgvr=�[0margoproj.io/v1alpha1/workflowtemplates
�[2m5:17PM�[0m �[91mERR�[0m Flashing error �[2merror=�[0m"unexpected error when reading response body. Please retry. Original error: net/http: request canceled (Client.Timeout or context cancellation while reading body)" �[2mmessage=�[0m"Watcher failed for argoproj.io/v1alpha1/workflowtemplates -- unexpected error when reading response body. Please retry. Original error: net/http: request canceled (Client.Timeout or context cancellation while reading body)"
�[2m5:17PM�[0m DBG History �[2mstack=�[0m"[context workflowtemplate]"

but it can list Workflow.

And sometimes it cannot list Secret, but kubectl can do.

To Reproduce
Steps to reproduce the behavior:

1. Open k9s
2. List argo WorkflowTemplate, or Secret
3. See error

Historical Documents
When applicable please include any supporting artifacts: k9s debug logs, configurations, resource manifests, ...

Expected behavior

K9s can list any resources as long as I have access and networking connection.

Screenshots

Versions (please complete the following information):

• OS: macOS
• K9s: v0.40.10
• K8s: v1.32.2

Additional context

At beginning I thought it's due to too many resource? but actually there're only ~130 WorkflowTemplates in my cluster,

[sunggun-yu]

similar issue I have. it stopped in context name listed. and freeze or timing out.
It started after upgrading to v0.40.10

[CCostaT]

I think I got the same issue. I've tried several versions and the last working one I found is 0.40.5. 0.40.6 onward is broken.

kubectl version output:

Client Version: v1.29.13
Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3
Server Version: v1.29.13

Logs (from k9s 0.40.6):

�[2m3:43PM�[0m �[92mINF�[0m 🐶 K9s starting up...
�[2m3:43PM�[0m �[92mINF�[0m ✅ Kubernetes connectivity OK
�[2m3:43PM�[0m �[92mINF�[0m [CONFIG] Saving K9s config to disk �[2mpath=�[0mC:\Users\[redacted]\AppData\Local\k9s\config.yaml
�[2m3:43PM�[0m �[93mWRN�[0m Fail to load global/context configuration �[2merror=�[0m"unable to locate associated cluster for context \"\": getcontext - invalid context specified: \"\""
�[2m3:43PM�[0m �[91mERR�[0m Unable to fetch server version �[2merror=�[0m"Get \"https://[redacted]/version?timeout=5s\": context deadline exceeded"
�[2m3:43PM�[0m �[91mERR�[0m Cluster updates failed. Giving up ;( �[2merror=�[0m"conn check failed (1/5)"
�[2m3:43PM�[0m �[91mERR�[0m Unable to fetch server version �[2merror=�[0m"Get \"https://[redacted]/version?timeout=5s\": net/http: request canceled (Client.Timeout exceeded while awaiting headers)"
�[2m3:43PM�[0m �[91mERR�[0m Cluster updates failed. Giving up ;( �[2merror=�[0m"conn check failed (2/5)"
�[2m3:43PM�[0m �[91mERR�[0m Unable to fetch server version �[2merror=�[0m"Get \"https://[redacted]/version?timeout=5s\": net/http: request canceled (Client.Timeout exceeded while awaiting headers)"
�[2m3:43PM�[0m �[91mERR�[0m Cluster updates failed. Giving up ;( �[2merror=�[0m"conn check failed (3/5)"
�[2m3:43PM�[0m �[91mERR�[0m Unable to fetch server version �[2merror=�[0m"Get \"https://[redacted]/version?timeout=5s\": net/http: request canceled (Client.Timeout exceeded while awaiting headers)"
�[2m3:43PM�[0m �[91mERR�[0m Cluster updates failed. Giving up ;( �[2merror=�[0m"conn check failed (4/5)"
�[2m3:43PM�[0m �[91mERR�[0m Unable to fetch server version �[2merror=�[0m"Get \"https://[redacted]/version?timeout=5s\": net/http: request canceled (Client.Timeout exceeded while awaiting headers)"
�[2m3:43PM�[0m �[91mERR�[0m Conn check failed. Bailing out! �[2mretry=�[0m5 �[2mmax-retries=�[0m5
�[2m3:43PM�[0m �[93mWRN�[0m Namespace validation failed �[2mns=�[0mmy-namespace �[2merror=�[0m"user not authorized to list all namespaces"
�[90m3:45PM�[0m �[32mINF�[0m �[1m🐶 K9s starting up...�[0m
�[90m3:45PM�[0m �[32mINF�[0m �[1m✅ Kubernetes connectivity�[0m
�[90m3:45PM�[0m �[31mERR�[0m �[1mFail to load global/context configuration�[0m �[36merror=�[0m�[31m�[1m"k9s config file \"C:\\\\Users\\\\[redacted]\\\\AppData\\\\Local\\\\k9s\\\\config.yaml\" load failed:\nAdditional property portForwardAddress is not allowed"�[0m�[0m

The logs mention "user not authorized to list all namespaces", however, this is not true because I can list the secrets just fine using kubectl get secrets -n my-namespace

edit: Forgot to add that I'm running it on Windows 11

[gottschd]

Same for me on Windows, wondering since 0.40.6 (and waited because of the frequent version updates lately)
k9s: 0.40.10

> kubectl version
Client Version: v1.32.3
Kustomize Version: v5.5.0
Server Version: v1.29.8+rke2r1

Had to downgrade to 0.40.5.

[MrVinkel]

I think some of the timeout errors mentioned here is due to the 'request-timeout' command line paramter was changed from "" to "5s" as default value from 0.40.5 to 0.40.6: https://github.com/derailed/k9s/pull/3185/files#diff-ab967ab1a2f3a1b769106eeb7bfe892ef0e81d1d27811fa15be08e6749feee1fL268

Running k9s --request-timeout 120s fixes the timeouts for me.

I think 5s is a very low default value

[maxsxu]

Running k9s --request-timeout 120s fixes the timeouts for me.

Thanks @MrVinkel , it works!

I think 5s is a very low default value

+1.

And according to kubectl docs, the default --request-timeout is 0 which means don't timeout requests.

[maxsxu]

I made a PR here for increasing the default request timeout value: #3253

[gottschd]

Running k9s --request-timeout 120s fixes the timeouts for me.

Works for me as well with 0.4.10.

@maxsxu Appreciate the PR, thx a lot

[sunggun-yu]

my issue fixed with 0.50.2

[sunggun-yu]

actually, 0.50.2, 0.50.3 didn't fix the issue. still experiencing timed out and working with --request-timeout

thanks for the PR!