My use case is to use this for project to deploy to multiple servers in a dynamic environment. Because of the the two way adding of device ids, it’s pretty hard to propogate the configuration changes for every server. If you if N is the number of servers I already have setup, each new server I add, I have to find a way to update 1+N configuration files.
Is there an easier way to handle this? Sorry if this is a common question, I tried to search first but did not find anything.
I would suggest that everytime you add a new node, you add it as “Introducer”. When you do so, any new node will be automatically added to all the nodes. Here is a simple procedure I applied on 13 machines:
Normally, every node should be added to each other, and all nodes added to the shares. Everytime you add a new node, add the master node as an introducer, and all other nodes will be added automatically. A restart of each node might be necessary if you want the node to be added right away. If your nodes are desktop/laptops, that should not be an issue since these are rebooted usually on a daily basis. If your nodes are servers, you can wait for an update of ST or program a crontab to restart ST on a regular basis.
One remark: I think it is important to always add the same master node, since I am not sure how the “Introducer” characteristics is spread amongst the nodes. If you want more flexibility, ensure that all nodes are configured as “Introducer” on every node. But that requires some extra work…
Actually it should be enough to restart the master node after you accepted the new node. Also, I think the introducer flag propagates, so master might become irrelevant.