Comment 6 for bug 1779304

Interesting point, I ran 10 instance single node scenario = "instance_metadata.yaml".

And checked that

 Lost agents: set(['shaker_frktis_agent_0', 'shaker_frktis_agent_1', 'shaker_frktis_agent_5', 'shaker_frktis_agent_6', 'shaker_frktis_agent_7', 'shaker_frktis_agent_8', 'shaker_frktis_agent_9'])

Means that shaker_frktis_agent_2,3,4 was not lost, I checked it on netstat -tapnl,

root@kolla-deploy:~# netstat -tapnl | grep 13337
tcp 0 0 0.0.0.0:13337 0.0.0.0:* LISTEN 17895/python2
tcp 0 0 192.168.1.10:13337 192.168.1.117:55058 ESTABLISHED 17895/python2
tcp 0 0 192.168.1.10:13337 192.168.1.111:60266 ESTABLISHED 17895/python2
tcp 0 0 192.168.1.10:13337 192.168.1.119:33362 ESTABLISHED 17895/python2
tcp 0 0 192.168.1.10:36314 192.168.1.10:13337 ESTABLISHED 17901/python2
tcp 0 0 192.168.1.10:13337 192.168.1.10:36314 ESTABLISHED 17895/python2

192.168.1.111
192.168.1.117
192.168.1.119
ips are for shaker_frktis_agent_2,3,4.
whereas others are lost.

Any idea?

I will try to boot them vm, next.