K9s: Stream read errors in the k9s log viewer

[NeonSludge]

Describe the bug
k9s displays a stream failed message after 5 seconds of viewing logs of any pod in a Kubernetes cluster and stops updating the logs.

To Reproduce
Steps to reproduce the behavior:

1. Open the pods view in k9s.
2. Select any running pod, press the l key.
3. Wait 5 seconds.
4. Observe the stream failed error

Expected behavior
k9s continues to stream the logs and doesn't display any errors in the log view.

Versions (please complete the following information):

• OS: CentOS 8.1
• K9s 0.20.1
• K8s 1.17.5

Additional context
This reproduces on a freshly created (via kubeadm) k8s cluster.
k9s 0.19.7 works fine and has no issues displaying logs from the same cluster.

There is also a stream read error in the k9s log that gets logged every time this issue is reproduced:

1:15PM WRN Stream READ error "flux/fluxcd-55fc7f5d57-686c7"::"flux" error="net/http: request canceled (Client.Timeout or context cancellation while reading body)"

[derailed]

@NeonSludge Thank you so much for reporting this! That's on me.
Should be happier on v0.20.2.
Tx!!

[billwright]

I'm still getting this error in 0.22.1. It looks like this:

my-pod-7f994d8fb6-zbqzc:stream failed: http2.StreamError{StreamID:0x625, Code:0x2, Cause:error(nil)}