YaCy Crashing Upon Remote Access

My current issue is that when I run YaCy from a remote computer and try to access it via another device it crashes fairly frequently on the hosting device. I don’t know what the conditions are for it to crash, but it seems to be accessing multiple pages at once. I can see from the pages taking a long time to load and the hosting devices’ CPU usage skyrockets to 70%+. This has been a problem since I’m currently trying to develop a web extension to improve YaCy’s functionality.

Is there a current fix to this issue? I’m running release: 1.924/10042 on win10.

I would ask for the get request information to restart YaCy in the affected computer, but it becomes completely unresponsive and must be killed via task manager.

Hi,
the best way to find out the cause of crash is examining the log file. See how.

Awesome, I’ll be sure to check next time this happens. Thank you

Is your hardware good?
Can you copy the Yacy folder to another pc and test?

The system crashed again but this time it was from trying to head to the admin panel. The error reported was:

W 2024/05/10 12:26:14 ConcurrentLog * java.lang.OutOfMemoryError: GC overhead limit exceeded
java.lang.OutOfMemoryError: GC overhead limit exceeded

The hardware shouldn’t be the issue. This had been running on worse hardware. The crashes are happening on this system:
i7 4790K
28gb DDR3
NVIDIA GTX 1070
But not on this system:
i7-4712MQ
8gb DDR3
Integrated graphics

Both are running in senior mode with crawling and the same version of 1.924/10042. I also noticed that the mouse pointer turns in to a black box when trying to type a search query if that’s any indication.

that seems to be a memory issue. one thing is RAM available on the machine, another is how much RAM is assigned to yacy. default value is pretty low. YaCy is RAM hungry, especially Kelondro crap.
See Performance Tuning. You can monitor RAM usage in First Steps > RAM/Disk usage. For me, usual value of MAX ram is something like 8 to 12 GB, with tens of millions pages in local index.

There was also a memory management improvement after 1.924 version, current github version is much better and stable. Unfortunately, @orbiter still didn’t make a release of 1.930, so a lot of people still download and use the old and buggy one, right away.