On 10/10/22 15:58, Jen-Ya Ku wrote:
We've deployed solr9 on OpenJDK 17 and it crashed after few hours with
following error:
# A fatal error has been detected by the Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x00007f834389b332, pid=8997, tid=9025
#
# JRE version: OpenJDK Runtime Environment Microsoft-40354
(17.0.4.1+1) (build 17.0.4.1+1-LTS)
# Java VM: OpenJDK 64-Bit Server VM Microsoft-40354 (17.0.4.1+1-LTS,
mixed mode, sharing, tiered, compressed oops, compressed class ptrs,
g1 gc, linux-amd64)
*# Problematic frame:
# V [libjvm.so+0xad1332]
PhaseIdealLoop::build_loop_late_post_work(Node*, bool)+0xf2*
#
I've attached the full log with this email. Is it a known issue?
Your attachment did not come through. It's generally a bad idea to rely
on attachments, because the mailing list eats almost all of them.
https://komodor.com/learn/sigsegv-segmentation-faults-signal-11-exit-code-139/
Way back in the dustbin of history, if somebody got signal 11 while
compiling the Linux kernel, the general advice was "This might be bad
hardware. It could be a software problem too."
The crash information looks to me like a problem with Java. I would
strongly recommend using Java 11 instead of Java 17 with current
versions of Solr. The latest 11.0.xx version you can get. Some variety
of OpenJDK is probably the way to go, unless you're willing to pay
Oracle for the right to run it in production. Everything seems to be
very stable for me running with OpenJDK 11.0.16 on Ubuntu Server. I did
give OpenJDK 16 a try. It was something older than the version you're
running.
It looks like you may have a JDK that's packaged by Microsoft for
Linux. Not sure I would trust Microsoft to get that right.
Thanks,
Shawn
