Plugin stuck creating exec stream

Using latest plugin, k8s `v1.19.1` and `v1.18.6`. Stuck at exec stream creation:
```
➜ ~ kubectl debug test -v 9
2020/10/21 00:37:32 Getting user name from default context 'test-4-admin@test-4'
2020/10/21 00:37:32 User name 'test-4-admin' received from context
kubectl-debug 2020/10/21 00:37:32 cmd.go:771: Getting id from containerStatus {Name:test State:{Waiting:nil Running:&ContainerStateRunning{StartedAt:2020-10-20 23:51:04 -0700 PDT,} Terminated:nil} LastTerminationState:
xxx
POST http://10.186.59.166:10027/api/v1/debug?authStr=&command=%5B%22bash%22%5D&container=containerd%3A%2F%2F4586b4d2e030426b0a3e8a753aac2cdae4f2dd8f68812bb69653c7b0ce81c6b5&hostname=hanlins-a01.vmware.com&image=docker.io%2Fnicolaka%2Fnetshoot%3Alatest&lxcfsEnabled=false®istrySkipTLS=false&username=test-4-admin&verbosity=9
kubectl-debug 2020/10/21 00:37:32 cmd.go:811: Creating exec Stream
```

On the agent side, got a timeout:
```
➜ kubectl-debug (master) ✔ kubectl logs debug-agent-w4bnh -f

2020/10/21 07:37:32 receive debug request
2020/10/21 07:37:32 http: superfluous response.WriteHeader call from k8s.io/apimachinery/pkg/util/httpstream/spdy.responseUpgrader.UpgradeResponse (upgrade.go:91)
E1021 07:37:47.473244 20211 httpstream.go:184] timed out waiting for client to create streams
```

From kubelet log, we can see the target pod (`test`) and agent being created but didn't see the reply:
```
Oct 21 06:50:43 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:50:43.657138 752 reconciler.go:224] operationExecutor.VerifyControllerAttachedVolume
started for volume "default-token-gnpv8" (UniqueName: "kubernetes.io/secret/7b7ffa74-9b85-4fa6-bbe8-b6eb3cf3ea93-default-token-gnpv8") pod "test" (UID: "7b
7ffa74-9b85-4fa6-bbe8-b6eb3cf3ea93")
Oct 21 06:55:19 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:55:19.715182 752 topology_manager.go:233] [topologymanager] Topology Admit Handler
Oct 21 06:55:19 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:55:19.855381 752 reconciler.go:224] operationExecutor.VerifyControllerAttachedVolume
started for volume "lxcfs" (UniqueName: "kubernetes.io/host-path/5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8-lxcfs") pod "debug-agent-w4bnh" (UID: "5f58ecd2-adac-
4c8d-8f9a-7d1896ec5cb8")
Oct 21 06:55:19 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:55:19.855872 752 reconciler.go:224] operationExecutor.VerifyControllerAttachedVolume
started for volume "runrunc" (UniqueName: "kubernetes.io/host-path/5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8-runrunc") pod "debug-agent-w4bnh" (UID: "5f58ecd2-a
dac-4c8d-8f9a-7d1896ec5cb8")
Oct 21 06:55:19 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:55:19.855971 752 reconciler.go:224] operationExecutor.VerifyControllerAttachedVolume
started for volume "default-token-gnpv8" (UniqueName: "kubernetes.io/secret/5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8-default-token-gnpv8") pod "debug-agent-w4b
nh" (UID: "5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8")
Oct 21 06:55:19 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:55:19.856100 752 reconciler.go:224] operationExecutor.VerifyControllerAttachedVolume
started for volume "docker" (UniqueName: "kubernetes.io/host-path/5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8-docker") pod "debug-agent-w4bnh" (UID: "5f58ecd2-ada
c-4c8d-8f9a-7d1896ec5cb8")
Oct 21 06:55:19 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:55:19.856283 752 reconciler.go:224] operationExecutor.VerifyControllerAttachedVolume
started for volume "runcontainerd" (UniqueName: "kubernetes.io/host-path/5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8-runcontainerd") pod "debug-agent-w4bnh" (UID:
"5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8")
Oct 21 06:55:19 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:55:19.856312 752 reconciler.go:224] operationExecutor.VerifyControllerAttachedVolume
started for volume "cgroup" (UniqueName: "kubernetes.io/host-path/5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8-cgroup") pod "debug-agent-w4bnh" (UID: "5f58ecd2-ada
c-4c8d-8f9a-7d1896ec5cb8")
Oct 21 06:55:19 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 06:55:19.856594 752 reconciler.go:224] operationExecutor.VerifyControllerAttachedVolume
started for volume "vardata" (UniqueName: "kubernetes.io/host-path/5f58ecd2-adac-4c8d-8f9a-7d1896ec5cb8-vardata") pod "debug-agent-w4bnh" (UID: "5f58ecd2-a
dac-4c8d-8f9a-7d1896ec5cb8")
Oct 21 07:43:10 test-4-md-0-74fb968bd8-cfqr9 kubelet[752]: I1021 07:43:10.888639 752 log.go:181] http: superfluous response.WriteHeader call from k8s.io
/kubernetes/vendor/k8s.io/apiserver/pkg/server/httplog.(*respLogger).WriteHeader (httplog.go:217)
```

It seems to be related to `kubelet` on attaching container, not sure if anyone experienced the same issue.

Reproduce the kubectl debug command against the reported Kubernetes versions, then trace the client’s “Creating exec Stream” message through the debug agent logs and kubelet attachment events. No files or tests are named, so first locate the exec-stream entry point and compare the client, agent, and kubelet handshake. Done means identifying the timeout cause and verifying a successful debug session.

Written by the indexing model from the issue text.