On Mon, 19 Dec 2005, Chris Wareham wrote:
I'm experiencing regular crashs in the MICO eventd daemon. The following
is a stacktrace from the core dump:
#0 0xfe59fc3c in _libc_kill () from /usr/lib/libc.so.1
(gdb) bt
#0 0xfe59fc3c in _libc_kill () from /usr/lib/libc.so.1
#1 0xfe53598c in abort () from /usr/lib/libc.so.1
#2 0xfe535c30 in _assert () from /usr/lib/libc.so.1
#3 0xfeb60c08 in CORBA::ORBInvokeRec::set_answer_invoke(CORBA::InvokeStatus,
CORBA::Object*, CORBA::ORBRequest*, short) (this=0x0, state=2949404,
o=0x2c8c20,
r=0xff00, ad=32556) at orb.cc:326
Line 326 of orb.cc is the first statement in the set_answer_invoke
method:
assert (_type == RequestInvoke);
The assertion appears to fail because the instance of ORBInvokeRec is
null.
This problem occurs with MICO 2.3.11 on Solaris 8, compiled with GCC
3.4.2 from the Sun Freeware archive. I have tried using MICO 2.3.12
instead, but my Orbacus Java clients are unable to connect to the MICO
Naming Service.
There is nothing which should make your OB java clients failing to
connect to 2.3.12RCx and working with 2.3.11. I would recommend to try
hard to test 2.3.12, since 2.3.11 is really quite out-dated now...
On MICO's nsd side you can use various debug option to better see what's
going on, I'm sure you can use also similar options on OB side. (on MICO's
side: -ORBDebug All switch on debugging of all subsystems...)
Cheers,
Karel
--
Karel Gardas [EMAIL PROTECTED]
ObjectSecurity Ltd. http://www.objectsecurity.com
---
Need experienced, fast, reliable technical MICO support?
---> http://www.objectsecurity.com/commsup.html <---
---
_______________________________________________
Mico-devel mailing list
[email protected]
http://www.mico.org/mailman/listinfo/mico-devel
