Thread race in detecting the subnets when machines are added in a batch
There is a thread race in Stork when multiple Kea machines with the same subnets configured are added as a batch. It causes to duplicate the entries in the subnet table.
Steps to reproduce:
- Run demo
- Authorize both HA machines in a batch
- Observe duplicated entries in the
subnet
table.
When the subnets are duplicated, adding a host reservation to both HA servers at once is impossible.