#Periodic 42.13.1 Server Error
15 messages · Page 1 of 1 (latest)
You're using Java 25?
try using 17
Can you sahre this log? steam/steamapps/common/Project Zomboid Dedicated Server/hs_err_pid6847.log
The game uses Java 25
The log seems to point to some sort of chunk load corruption, very likely to happen either due to mods or due to the map part that is loaded being corrupted.
Is it possible to reproduce it without mods and then share the logs? Otherwise I cannot help much, we do not recommend using mods at all in 42 due to the instability they will cause.
Thanks for looking! This log was without mods, but i understand what you're saying is that it could have corrupted due to the mods and cause the issue after removing. I had experienced this crash prior to adding in any mods, though (at least the error and behavior looked similar, but maybe it wasn't the same).
I've attached a crash file from when the server was completely vanilla, with no mods ever added to it. It happened three times in a row, and then started successfully. I unfortunately don't have the server console log anymore.
I did find an old thread and swapped "-XX:+UseZGC" to "-XX:+UseG1GC" if that's relevant. That was after this crash though. Swapping that vm arg allowed it to start, but now I'm thinking it was just a non-deterministic issue and i happened to not hit it on that start as opposed to that being the fix.
I added a couple mods after to fix things like dryingracks, composter lag but the crashes continued. They have definitely gotten worse, which could be attributable to these mods but idk why it was crashing prior. I'm wondering if I need to memtest my server or something.
Hm, check if swapping to UseG1GC helps in this case, it could be a problem with the GC
If it stops crashing, let me know, since we would need to investigate it as a bug
I'm going to try regenerating all of the chunks besides our base with no mods and see if it still happens using G1GC. I'll report back on the forum bug reports if so since I'll be sure it's not a PEBKAC error. Does that approach seem sound?
Sounds good to me
So the server was rock solid for two days after I wiped all map data besides our base, and then the same crash started again today. I can't get it to start at all now without crashing with the same error. I dont see anyone else reporting this issue, so I'm wondering if my machine just has a hardware issue causing it to save corrupt data over time.
Do you think it's worth trying another JRE or is that unsupported? If so, will unpacking a jre runtime directory to replace the current bundled jre dir sufficient?
edit: - my server just rebooted into initramfs saying there was corrupted data on my drive. I think the jdk crashes are a red herring for a hardware issue. Disregard this.