Issue Description:
at times, especially on hi int/hi int shock server sometimes crashes and kicks me back to char select screen. No option to rejoin, just X amount of time wasted.
Steps to Reproduce:
n/a
Mission Name (If Applicable):
Excise Vault Spireside-13
Platform:
[Steam/]
Player ID:
Striges
Approx. Time of Issue & Timezone:
15.30-ish
Reproduction Rate:
Rare (<10%)
Upload Supporting Evidence:
[Screenshots, recordings, links to Twitch VODs, etc.]
basing on logs, the server was shut down because it was a spot instance of a VM. the spot instances this game uses (aws fleetiq service) can be decommisioned by cloud provider with 2 minute notice…
13:30:07.240 [Lua] INFO [PresenceManager] received push message={
type = “event_trigger”,
source = “instanceShutdownEvent”,
payload = “{“event_name”:“game_session_aborted”,“args”:[{“game_session_id”:“19374b13-78a9-4232-9fae-b1b97c5f3493”}]}”,
}
in general companies use spot instances because they are much cheaper to run, but there is always inherited chance they can be terminated.
idea is that application must make sure it handles handover to another server. it has 2 minutes to do that… seems either something went wrong while migrating game state to another server (you would then be able to connnect back to the game) or simply there is no such handling implemented.
