For that “every time” part at least, we have something in the works. The launcher currently in testing, and soon active on testlive, has a continue functionality which should work also for direct connect.
For that “every time” part at least, we have something in the works. The launcher currently in testing, and soon active on testlive, has a continue functionality which should work also for direct connect.
I notice the UDP; have you done the port forwarding on both udp AND tcp?
Other than that, the common culprits are:
Not really any other idea, I would suggest joining the Admin discord and ask there if anyone can come up with a suggestion, https://discord.gg/UCuMS5d
(Btw, I assumed you were not trying to run anything else on the machine, like other games, or the client, etc… that you get the not successful test with just the server launcher running)
These are not the server logs, they are the Dedicated Server Launcher logs, you can access the actual game server logs by clicking the bottom left blue icon.
I had requests to add a “watchdog” to automatically kill the server if it gets stuck on shutdown/restart, but I guess technically this is more or less in the same category of server failure that makes the server unusable.
Does anyone know of a foolproof way to detect that the server is crashed or frozen?
I’ve actually never see the server crash with the “generic error dialog”, it’s supposed to have its own proper crash reporter, so that would mean that somewhat some parts of the code are not handled by the exception handler.
Now that you know where the logs are supposed to be, could you look and see what it was writing to it before it crashed and burn?
That’s easy to check:
So either start the server manually with the same parameters, or start the server with the launcher and quit it to let it crash by itself.
Regarding monitoring the process, I already have all the code to get the pid of the server, the window handle, etc… I’m just not sure of a 100% reliable way to detect the process is properly crashed or frozen, would be a shame to kill a server that just happen to be running on a potato machine and currently suffering from heavy memory swapping, for example.
That would just be what happened at the end of the log before the final crash report summary.
This would likely only be used by persons who have issues with crashing servers anyways right?
Which is basically everybody, at least at some point.
Generally speaking, for the launcher I try to implement things that work out of the box for everybody, because adding checkboxes and custom options just makes the tool harder to use/explain.
The bottom line is, I will try to find a way to detect a stuck/crash/unresponsive server.
Right now, the case of “the server crashed, sent a report and quit” is already handled, if you have the “start server if not running” it will automatically get restarted.
The two issues I have are a variant of the same case:
If I can find a reliable way to detect a non behaving server, that would easily handle both cases, the question is to find reliable metrics.
No idea, I just tried and it worked for me.
Which web browser are you using, and in which part of the world are you located?
Maybe for some reason the CDN (Content Delivery Network) is suffering an outage in your area?
Hmm…
Hmm…
Is it a popup thrown by the Dedicated Server Launcher, or by the actual game server?
Can you show the content of the Launcher log? (At the bottom, on the tabs on the right side)
Could you try to install the launcher in a folder without space, like instead of “C:\users\sweet potato” something like “C:\ConanServer” just to see?
Also, do you have a E: drive at all?
Only thing I can think of is to try to “validate” the install, but after that, I’m out of idea.
I’ve never, ever, seen this particular error, so I guess maybe join the server admin discord and ask there if anyone has seen that one.
I’m not sure, Exiles has a lot of log spam, and I’m not going to poke the colleagues on a Sunday evening (I’m not on the Exiles team, I’m only maintaining the tool)
Well, that was the good thing to do, just bad luck you managed to get one issue I’ve never seen before !
Actually, if you could somewhat make a video capture that shows the problem, like capturing the starting of the tool, launching, patching, up to the point where the e: popup appears, that would make it easier to show people to ask if they have any idea of what could cause that.
The function I’m calling is doing similar to what is shown on this old Code Project article:
https://www.codeproject.com/Articles/9113/Get-CPU-Usage-with-GetSystemTimes
Still a mysterious thing.
Ok, bonus question, on this machine, have you ever connected a mobile phone, or do you happen to have a memory card reader installed?
Just for a test, could you try to do something horrible like:
normally you should see a E:\ drive appear on the machine (basically pointing at the same content as your C: drive, but you could SUBST to some safe subfolder that contains nothing) see if that helps appeased the server?
Well, it’s not the launcher that stopped the server, it’s the server that stopped after crashing, which could be caused my many things (mods order or compatibility, game db content, etc…)
No matter what we try, make sure first that you have proper backups of your gamedb (I assume you have setup that) so you always have a version with all the content.
Now, regarding what causes the crash, I can’t help, so my suggestions would be to post that on the actual game crash threads, or possibly ask on the server admins discord.
Ok, I also get that when I try with an existing drive letter, so somewhat windows believe there is a E: drive, but it can’t be accessed.
I assume you tried to install all the windows updates, and reboot the machine?
If you can come up with a working strategy, I can investigate, but at first glance I don’t see an easy way of detecting/associating a startup error with the mods, if that was easy, I think the game would already implement that
You just jinxed it!