Capturing a network trace in a pod

You will need the pod running as the root user in order to perform these steps. See Running a pod as the root user for more details.

Get the name of the pod.
```
 kubectl get pods 
```
Start a shell session in the pod. Issue the below command, and make the following substitutions:
```
kubectl exec -it <podname> --container=<container name> -- bash 
```
podname : The name of the pod.

container_name : is the name of the container. If the pod only has one container you can omit this parameter.

For example, if the pod name is jitsi-74d95d6d49-k5nts and the container name is jigasi, then the resulting command is:
```
kubectl exec -it jitsi-74d95d6d49-k5nts --container=jigasi --bash
```
Update the repositories by issuing the below command:
```
apt-get update 
```
Install tcpdump by issuing the below command:
```
apt-get install tcpdump 
```
Capture the network data, by issuing the command:
```
tcpdump -i any -w <filename>.pcap 
```
Reproduce the problem.
To stop and save the capture press Ctrl+C.

Parent Topic: Troubleshooting Sametime on Kubernetes