Expected behavior:

When using tsh kubectl exec to connect to a pod, the connection should be established, allowing the user to perform interactive commands in the pod's container.

Current behavior:

Upon executing tsh kubectl exec, the connection is established, but it results in an error message:

error: write tcp 127.0.0.1:54234->127.0.0.1:54230: write: broken pipe

Bug details:

• Teleport version: v15.2.2
• Recreation steps:
  1. Execute tsh kubectl exec -n <namespace> -it <pod-name> -- bash
  2. Receive error message: "error: write tcp 127.0.0.1:54234->127.0.0.1:54230: write: broken pipe"
• Debug logs:

  {"caller":"reverseproxy/reverse_proxy.go:223","component":"proxy:proxy:kube","level":"info","message":"Round trip: GET https://kube-teleport-proxy-alpn.teleport.cluster.local/api/v1/namespaces/mynamespace/pods/mypod, code: 200, duration: 44.911025ms tls:version: 304, tls:resume:false, tls:csuite:1301, tls:server:kube-teleport-proxy-alpn.teleport.cluster.local","pid":"7.1","timestamp":"2024-04-17T19:37:07Z"}
  {"caller":"proxy/remotecommand.go:164","component":null,"level":"info","message":"Negotiated protocol v4.channel.k8s.io.","timestamp":"2024-04-17T19:37:07Z"}
  {"timestamp":"2024-04-17T19:37:07Z","level":"info","caller":":0","message":"http: response.WriteHeader on hijacked connection from go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp.(*respWriterWrapper).WriteHeader (wrap.go:98)"}
  {"timestamp":"2024-04-17T19:37:07Z","level":"info","caller":":0","message":"http: response.Write on hijacked connection from go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp.(*respWriterWrapper).Write (wrap.go:80)"}
  {"caller":"proxy/forwarder.go:702","component":"proxy:proxy:kube","level":"warning","message":"Failed writing kube error response body: http: connection has been hijacked","pid":"7.1","timestamp":"2024-04-17T19:37:07Z"}
  

There is a similar issue described in #33020, but the fix for that is already included in v15.2.2.