The only firewall requirement in this setup is that the Node must be able to contact cloud.crosser.io via HTTPs port 443.
In this case, you need to make sure to open up the required communication protocols from the DMZ to the private network. Beside that, make sure that you can reach the public network as well from the DMZ so your Node can connect to cloud.crosser.io and potentially other external services.
The downside of this approach is that you must allow communication into your ‘private network’ from the DMZ. In many industrial environments, this is a no-go and will not be allowed since the guidelines often accept outbound traffic from the private network. Fortunately, there is also a solution to this which is explained next.
If inbound connectivity to your data sources is restricted, you have to put a Crosser Node in the same network level in order to be able to connect to your required sources. Once you can connect to your data sources from that level, you might run into the limitation that you can not connect to your destination systems anymore from this network level. If that is the case, you can utilize the Node-to-Node communication concept to use a Node as some sort of Aggregator or Bridge to the outside world.
Keep in mind that in this concept the Node in the private network still needs access to cloud.crosser.io via HTTPs.
When you have to deal with a setup where your Node(s) run on Edge Gateways or similar in distributed environments/assets, your main challenge is usually to get outbound connectivity to your services.
On the southbound-side, you usually do not find any firewall, hence the connectivity to sources is most likely easy to set up without any network configuration.
In this scenario, we highly recommend using protocols like HTTPs / REST to send out the payload data to your destination. Especially if your network connection to the outside world is not that reliable (ie. Cellular), protocols like MQTT and AMQP are not recommended.