The patient is SLES11Sp4, recently (Sunday) patched up through what was available at the end of May. Has eDir 8.8.8.8 (HF2, IIRC) and IDM4027. Has three remote loaders configured. After the patch install, rdxml dies when the Jvm is loaded. Wasn't doing this before the patch, so the reasonable assumption is that the SLES patches somehow caused it. Not convinced on that, at least not so far.

The rdxml.binNNNN.log reports:

#
# If you would like to submit a bug report, please visit:
# http://bugreport.sun.com/bugreport/crash.jsp
#
#
# A fatal error has been detected by the Java Runtime Environment:
#
# SIGBUS (0x7) at pc=0x00007fdcb507049d, pid=774, tid=140586021922592
#
# JRE version: 7.0_21-b11
# Java VM: Java HotSpot(TM) 64-Bit Server VM (23.21-b01 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# j java.lang.Object.<clinit>()V+0
#
# Core dump written. Default location: /var/opt/novell/dirxml/rdxml/core or core.774 (max size 1 kB). To ensure a full core dump, try "ulimit -c unlimited" before starting Java again
#
# An error report file with more information is saved as:
# /var/opt/novell/dirxml/rdxml/hs_err_pid774.log

and /var/opt/novell/dirxml/rdxml/hs_err_pidNNN.log has pretty much the same, plus the usual Java barf of everything.

The only references I can find to Java + SIGBUS are "can't write to temp directory" and "disk full" errors. Neither is true here. So far, haven't found any other really good ideas. Running it with strace, I see the sigbus error, but I don't see a cause for it. It looks like it's actually a memory error, which also doesn't make sense, this being a VM.

Engine and drivers that run in the engine are fine. The eDir JRE seems to be ok, it compares (md5sum) ok with another box, with the exception of cacerts, which seems reasonable. Trace doesn't show anything interesting, because as soon as the JRE loads, it dies.

Anybody ever seen something like this before?