Java运行时环境已检测到致命错误：

Question

Can some help in ,

Our Production servers all of a sudden starts crashing, I researched a lot but didn't find a solution , other than people suggesting to upgrade the java, but we are already on latest patch (java 6_71)

#
# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0x00002b5b4dc9c2ba, pid=1278, tid=47671325686080
#
# JRE version: 6.0_71-b12
# Java VM: Java HotSpot(TM) 64-Bit Server VM (20.71-b01 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# C  [libcknfast.so+0x10f2ba]  long double+0x11a
#
# If you would like to submit a bug report, please visit:
#   http://java.sun.com/webapps/bugreport/crash.jsp
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.
#

---------------  T H R E A D  ---------------

Current thread (0x00000000590ea800):  JavaThread "http-0.0.0.0-8090-2" daemon [_thread_in_native, id=11189, stack(0x00002b5b585ef000,0x00002b5b586f0000)]

siginfo:si_signo=SIGSEGV: si_errno=0, si_code=1 (SEGV_MAPERR), si_addr=0x00000000000001a1

Answer 1

Short of knowing precisely what the issue is, consider the following in determining the problem's root cause and resolution.

The error message actually explains a lot:

# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.

The problematic frame is:

# Problematic frame:
# C  [libcknfast.so+0x10f2ba]  long double+0x11a

"[A]ll of a sudden starts crashing" combined with this error information suggests among other possibilities:

• A change to the Java code on the server that interacts with the PKCS11 library ( libcknfast.so ).
• A change in the PKCS11 library ( libcknfast.so ) itself.

Basically, consider whether your team deployed updated Java code, an updated PKCS11 library, JBoss updates, or other updates or configuration changes on which any of them depend prior to the problem manifesting itself.

Also, there are many SO questions & answers related to “The crash happened outside the Java Virtual Machine in native code.” - over 250 presently - that may help with specific hypotheses as you consider possible changes to the production server prior to the problem.

One that caught my attention in particular came down to simply passing an extra null to a native function (ie an example of a potential code change at the root of such an error); another that caught my attention came down to JVM memory parameters (ie an example of a possible configuration change at the heart of such an issue).