We have an API scaled to 4, talking to a single database. In the environment there are a mix of centos and one photon host. When the database ended up on the photon host, api containers from the same stack that ended up on other hosts were unable to connect to the database. The api container running on the photon host was able to connect. They could still translate the short hostname, but timed out when trying to connect. Pings did not work either. moving off photon solved the problem. Any ideas what I would need to open up on the photon host?