[HELP] Caves shard reset our characters

(I've only ever ran the dedicated server in local network so I never had this happen and I'm just guessing and poking...)

Can you describe the networking a bit more - are you actually running the (DST) servers on 2 different IP addresses which are assigned on the same host? That seems unusual.

If not, why use bind_ip*? This is only required if you want to run both shards, on the same host, on the same port - which you clearly aren't.

In any case, what's in your cluster_settings.ini? If you're running servers on different hosts, or bound to different addresses on the same host, you probably need to set master_ip and master_port in the shard section, setting them to the master shard's shard/bind_ip and network/server_port values, respectively.

EDIT:

* binding in this context refers to TCP/UDP socket binding. Normally, listening sockets are bound to "0.0.0.0:someport" (or [::]:someport if we ever make it to current-gen IP), meaning the process will receive requests addressed to any IP address of the host on the give port - this normally covers at least 2 different addresses, 127.0.0.1 and whatever address is used to access the wider internet. If you instead bind the TCP socket to a specific address, only requests addressed to that address are passed to the process. The typical use case is if you need to run multiple instances of some software that cannot change the port it listens on, on a single host - in the olden days, this would be your HTTP(S) server which has to run on port 80 or 443, because no one wants to force website visitors to use non-standard ports.

Edited Thursday at 08:39 AM by myxal