I am using build 268 on a purpur 1.18.1 server. If I leave the profiler /spark profiler --thread * running for too long and then stop it, most of the times the upload will hang and never appear and If I try to run any profiler again it says. An active profiler is already running and if I try to stop it, it prints this and everything gets stuck.

https://nekobin.com/kagudireco

This is the error I managed to catch when a server crashes with /spark profiler --thread *
it doesnt always appear though
https://nekobin.com/yiyezemuli
@lucko

Stuck again..... new message though.
https://nekobin.com/yicanavara

Hmm, seems like the temporary file that spark writes to may be being deleted.

Could you give some more details about your hosting environment? OS versions, etc

Hmm, seems like the temporary file that spark writes to may be being deleted.

Could you give some more details about your hosting environment? OS versions, etc

OVH dedicated machine, I am using Ubuntu 20.04.3 LTS (Focal Fossa), I've also installed a ptero panel for my servers.
A bit unrelated perhaps but the command /spark profiler --thread * can also completely crash the server without any logs, warning or anything. Its like someone used a power kill switch.

Most likely fixed in newer versions by an async-profiler update. Please feel free to re-open if it is still happening.

Most likely fixed in newer versions by an async-profiler update. Please feel free to re-open if it is still happening.

I got the report but I also got this console error.
https://nekobin.com/zecavewoxe

Oops, should be fixed in 49b8f53

I am getting a different error now
https://nekobin.com/supakuweno
it doesn't always happen, only a few times, especially if i leave it for a long time.

Same issue here, after letting some time the profiler running without parameters added PurpurMC/Purpur#1058

I've replied on your Purpur issue - basically, don't run the profiler for so long.