Question Randomly appearing crash | PAPER 1.17.1-408

[Zardexd]

We had weird symptoms before , both servers on the same machine just crash at the same time , we solved the problem, it turned out java used too much memory despite xmx xms. Now they crash independently. I recreated server from scratch , but it still crashes but with different problem now.

used too much memory​

Spoiler: OLD PROFILING

Spoiler: We used these java options before

java -Xms10G -Xmx10G -Xmn5G -XX:MaxMetaspaceSize=256M -XX:+HeapDumpOnOutOfMemoryError -XX:StartFlightRecording:filename=${JFR_FILE},maxsize=256MB -XX:+UseG1GC -XX:+ParallelRefProcEnabled -XX:MaxGCPauseMillis=200 -XX:+UnlockExperimentalVMOptions -XX:+DisableExplicitGC -XX:+AlwaysPreTouch -XX:G1NewSizePercent=30 -XX:G1MaxNewSizePercent=40 -XX:G1HeapRegionSize=16M -XX:G1ReservePercent=20 -XX:G1HeapWastePercent=5 -XX:G1MixedGCCountTarget=4 -XX:InitiatingHeapOccupancyPercent=15 -XX:G1MixedGCLiveThresholdPercent=90 -XX:G1RSetUpdatingPauseTimePercent=5 -XX:SurvivorRatio=32 -XX:+PerfDisableSharedMem -XX:MaxTenuringThreshold=1 -Dusing.aikars.flags=https://mcflags.emc.gs -Daikars.new.flags=true -jar paper-1.17.1-408.jar nogui

Spoiler: System info

Its dedicated hosting , data center is in St.Petersburg

Spoiler: Current Java options

java -Xms10G -Xmx10G -XX:+HeapDumpOnOutOfMemoryError -XX:StartFlightRecording:filename=${JFR_FILE},maxsize=256MB -jar paper-1.17.1-408.jar nogui

Spoiler: NEW PROFILING

Spoiler: PLUGINS LIST

Error log
Server log
FlightRecording output


We run server as root , and current java vesrion is
openjdk 16.0.2 2021-07-20 OpenJDK Runtime Environment Temurin-16.0.2+7 (build 16.0.2+7) OpenJDK 64-Bit Server VM Temurin-16.0.2+7 (build 16.0.2+7, mixed mode, sharing)

Server crashes randomly , there is 10-12 people at max , CPU and RAM capacities is 50% free most of the time.

Im sorry for my bad English , second language. Please help me , im a bit new in the field and i struggle a lot to understand whats even happening in those logs.

 

[electronicboy]

segmentation fault, aka, bad memory access

Generally caused by bad hardware or a bad Java install

 

[Zardexd]

segmentation fault, aka, bad memory access

Generally caused by bad hardware or a bad Java install

Looks like its true I saw couple of errors in the linux logs saying Bad Segment or smth, but after i googled there was said that its okay. So i should contact the hosting, and change the java. Do latest Oracle 17.0.2 LTS will fit well , i heard 1.17 Minecraft doesnt like 17.0 Java and its better to keep running on Java 16 , do i need to recreate the server after switching java?

 

[electronicboy]

if you have bad ram you're gonna wanna get the ram placed, not aware of linux logs generally logging ram issues, you sure that you weren't looking at stuff relating to networking?

java 17 is fine, 1.17 works with java 17 just fine too, this is likely a hardware/environment issue; You don't need to recreate the server if changing java versions

 

[Zardexd]

Most likely i messed up installation cause i use Linux like less than a week , and i installed JAVA from tar.gz manually. I could switch back to an actual Oracle java 17 that installed automatically. Thanks for help , we'll see how it goes.

I was looking at syslog.1 not sure what its even logging. (windows user moment)

 

[Zardexd]

https://pastebin.com/WtRA2W15
Crashed again on a new java. As far as i can google and undertand it says pretty much the same. But this time it points at JFR itself , should i try disable it?
upd. i did it anyways.
Or if its like because of faulty RAM it may point at anything?
JRE version: Java(TM) SE Runtime Environment (17.0.1+12) (build 17.0.1+12-LTS-39)
# Java VM: Java HotSpot(TM) 64-Bit Server VM (17.0.1+12-LTS-39

I requested hardware investigation from my hosting , it will be ready tomorrow.

 

[Zardexd]

Its 04:42AM for me , and hosting employees just finished their diagnostics. They replaced 2 boards of RAM , it was corrupted.

I cant believe the fact that i couldn’t just copy the “SIGSEGV” and google it. Im so unbelievable bad lol.

Thank you a lot , you have no clue how much time you saved for us! This bug terrorized us whole week. We were down the schedule like 5 days because of this problem. Thank you once again!