Error "missing content-type field"

Hi, we are currently testing out Dgraph in Kubernetes. When we were live loading some data, we sometimes see this error:

Connection lost with dgraph-dgraph-zero-1.dgraph-dgraph-zero-headless.dgraph.svc.cluster.local:5080. Error: rpc error: code = Internal desc = broken pipe                                         
Unable to sync memberships. Error: rpc error: code = Internal desc = broken pipe. State: <nil>                                                                                                  
Got address of a Zero leader: dgraph-dgraph-zero-1.dgraph-dgraph-zero-headless.dgraph.svc.cluster.local:5080                                                                                    
Starting a new membership stream receive from dgraph-dgraph-zero-1.dgraph-dgraph-zero-headless.dgraph.svc.cluster.local:5080.                                                                   
Unable to sync memberships. Error: rpc error: code = Unknown desc = OK: HTTP status code 200; transport: missing content-type field. State: <nil>                                               
Got address of a Zero leader: dgraph-dgraph-zero-1.dgraph-dgraph-zero-headless.dgraph.svc.cluster.local:5080                                                                                    
Starting a new membership stream receive from dgraph-dgraph-zero-1.dgraph-dgraph-zero-headless.dgraph.svc.cluster.local:5080.                                                                   
Error while retrieving timestamps: rpc error: code = Unknown desc = OK: HTTP status code 200; transport: missing content-type field with delay: 10ms. Will retry...

Can we know what’s wrong and how to prevent or fix it?

I would need more information before I can hazard a guess. Some questions:

  • What is the status of dgraph-zero-1 when you this happened?
    kubectl describe pod/dgraph-dgraph-zero-1
    kubectl logs dgraph-dgraph-zero-1
    
  • How are you running dgraph live, the full command line?
  • Where are you running this? For example, did you kube exec -ti dgraph-dgraph-alpha-1 into a particular pod to run this?

Lastly, this is what version of dgraph that was deployed? And out of curiosity, how was this deployed?