I would say that this kind of thing was *intended* to be a feature, but in your use case, would be a bug :D
The reason this happens, is that a lot of data in actionHero is keyed to `
api.id`. This is important for a lot of things, but specifically the `task` and `stats` sub systems. With tasks, there are queues created for each server and worker combination so we can build up individual queues for "all" tasks. For stats, the ID is used to keep "local" stats, in addition to "global" stats.
In this commit
https://github.com/evantahler/actionHero/commit/4100a05968a1df09ff51b8306f17ba49b1bbf7d6 I allowed the optional config option of "
api.configData.general.id" to be defined which would overwrite the default id creation logic (which is based on IP address). Feel free to use this if you think your server IPs will be changing often. It's really importnat that each and every server running (simultaneously) has a unique ID. This includes running a node cluster with n workers. If you need it, in config.js, you can append the pid or other such things to solve this problem.