Lens does not show Workloads even after full uninstall

[ilelouis]

Describe the bug
After installing Lens, the workspace does not show anything. Any attempts to expand Workloads, Overview, Apps, etc do not show expected items.

I have attempted to uninstall Lens, remove the Lens directory and cache as well as completely remove the kubeconfig file but the issue persists. I installed from k8lens.dev and using brew, but the problem persisted. Several other users do not have this issue.

I am able to perform all actions via kubectl so I do not think this is a permissions issue within the cluster.

To Reproduce
Steps to reproduce the behavior:

1. Install Lens
2. use gcloud to populate the kubeconfig to connect to GKE cluster
3. Open Lens
4. Notice that nothing shows on Lens

Expected behavior
Expected behavior is to open Lens and be able to see all workloads, nodes and other objects.

Screenshots

Environment (please complete the following information):

• Lens Version: 5.0.2
• OS: Mac Catalina 10/15/7
• Installation method (e.g. snap or AppImage in Linux): Website or brew

Logs:
Only error in logs i see is:

error: Can't find working API for the Kubernetes resource ingresses

Additional context
I have tried multiple versions 4.x and 5.x without success.

[Nokel81]

Can you post a video of what you are seeing when you start lens?

[Nokel81]

Also I don't think "Workspace" is a kube object type.

[ilelouis]

I apologize for the delay here.

When i say "workspace" i really just mean the app UI. In the screenshot below, the user has logged into the GKE cluster and connected to it in Lens but cannot see any objects, nothing displays. The user is able to run kubectl commands to get pods, ingress, etc, but nothing shows in Lens

[Nokel81]

Sounds like the cluster that is being used in lens has a different user account. One that has some restrictive RBAC rules. Are there several clusters in the catalog?

[ilelouis]

We control access via Active Directory groups. The user in question is in the same group with the same RBAC assigned via Role as a dozen other users. I temporarily gave the user cluster-admin permissions thinking maybe there was something wrong, but issue persisted.

The user has added a second cluster to the catalog ad confirmed that they have the same issue in the second cluster.

Could there be some other location where Lens stores configuration files or objects that might need to be manually removed in order to attempt a full uninstall / re-install?

[Nokel81]

Yes that would be /Users/<user>/Library/Application Support/Lens

[ilelouis]

OK thanks. so we did have the user remove that folder many times while we tried to troubleshoot this issue. After every uninstall, we would remove that folder.

[Nokel81]

Do you see anything more with DEBUG=true running from the command line? Does the user have any local firewalls?

[ilelouis]

The only error at the time was error: Can't find working API for the Kubernetes resource ingresses as noted in the problem description. But i can have the user launch it again and send me the full log output and then scrub it, I will post that shortly

It is possible that one of the security tools installed on our laptops are blocking the connectivity and I have opened up a ticket internally to check on that.

[Nokel81]

That error occurs if all of the kube APIs URLs for the ingresses resource is rejected. A full log would be very helpful, thank you.

[yogageek]

able to run kubectl commands to get pods, ingress, etc, but nothing shows in Lens

I totally have the same issue. Workloads only show Overview, but inside are empty

Yes that would be /Users//Library/Application Support/Lens

Where can I find this folder on my windows?

[Nokel81]

able to run kubectl commands to get pods, ingress, etc, but nothing shows in Lens

I totally have the same issue. Workloads only show Overview, but inside are empty

Yes that would be /Users//Library/Application Support/Lens

Where can I find this folder on my windows?

Do you have RBAC? On windows that would be %AppData%\Lens

[yogageek]

I'm not sure what's mean by RBAC here. As I know I share one kubenetes with other developers, and my account role is ClusterOwner.

[Nokel81]

If you run the lens executable with the environment variable LOG_LEVEL set to silly. Do you see any log lines of the form [CLUSTER]: push-state { ... } and if so, what is in that logged object?

[mcm-ham]

I'm experiencing this bug too. In the logs I see this error:

info: Downloading kubectl 1.20.8 from https://storage.googleapis.com/kubernetes-release/release/v1.20.8/bin/darwin/arm64/kubectl to /Users/michael/Library/Application Support/Lens/binaries/kubectl/1.20.8/kubectl
error: Local kubectl failed to run properly (Command failed: "/Users/michael/Library/Application Support/Lens/binaries/kubectl/1.20.8/kubectl" version --client=true -o json
/Users/michael/Library/Application Support/Lens/binaries/kubectl/1.20.8/kubectl: line 1: syntax error near unexpected token `<'
/Users/michael/Library/Application Support/Lens/binaries/kubectl/1.20.8/kubectl: line 1: `<?xml version='1.0' encoding='UTF-8'?><Error><Code>NoSuchKey</Code><Message>The specified key does not exist.</Message><Details>No such object: kubernetes-release/release/v1.20.8/bin/darwin/arm64/kubectl</Details></Error>'
), unlinking

Seems it's trying to download kubectl from this URL which doesn't work https://storage.googleapis.com/kubernetes-release/release/v1.20.8/bin/darwin/arm64/kubectl if I try to download from this URL https://dl.k8s.io/release/v1.20.8/bin/darwin/amd64/kubectl and place in folder, it simply gets deleted the next time I try to run the app.

[Nokel81]

@mcm-ham You should be able to just download the version that you want to use and stick it somewhere else and then point Lens at it.

FYI your second URL is for an amd64 binary not a arm64 binary.