Scheduling failed: nic subnet's matches host and vnet


I’ve been getting the following exception over and over every 2 seconds on my rancher server, and its eating cpu/IO of my server.

The exception is: scheduling failed: nic subnet’s matches host and vnet

Here is the pastebin of the full exception:

Anyone having a similar problem? I haven’t seen anything of the forums. I’m currently using Rancher 0.63.1, RancherOS 0.4.3 on three severs, and one server running CentOS 7.1.

I found that some condition happened when I removed a host from Rancher, It seems the host was marked as purged in the database, but on the Account table there were still two Agent Accounts trying to spawn on the removed host. After setting them as purged, and also setting the instance as purged manually on the database, it stopped trying to spawn the agent.