Setting cpu_affinity breaks deploy-config

[dopheide-esnet]

As per the documentation, worker sections in the cluster config should support cpu_affinity. However, actually setting that results in failed deployment.

Example config lines:

[worker1]
instance=agent-zeek-test-cluster1.es.net
role = worker
interface=af_packet::enp175s0f1
cpu_affinity=0
env = ZEEK_PLUGIN_ACTIVATE=Zeek::AF_Packet

This results in a timeout:

zeek@zeek-test-cluster1:~$ zeek-client -vvv --controller zeek-test-cluster1.es.net:2149 deploy-config cluster.cfg
info: connecting to controller zeek-test-cluster1.es.net:2149
info: peered with controller zeek-test-cluster1.es.net:2149
error: no response received: websocket connection to zeek-test-cluster1.es.net:2149 timed out

For an older version of zeek (not using the websocket yet) we get a little cleaner error:

zeek@zeek-dev1:~/zeek_management$ zeek-client deploy-config cluster.cfg
{
  "errors": [
    "request timed out"
  ],
  "results": {}
}