[omniORB] RE: Problem with IRIX 6.3

[Sai-Lai Lo]

--=-=-=




--=-=-=
Content-Type: message/rfc822
Content-Disposition: inline
Content-Transfer-Encoding: 8bit

Date: Thu, 29 Apr 1999 01:20:50 -0700
From: "Nilo Stolte" <nilo_stolte@my-dejanews.com>
Message-ID: <OBLOIJOKDINFBAAA@my-dejanews.com>
Subject: RE: Problem with IRIX 6.3
 
--

On 27 Apr 1999 11:51:35 +010   Sai-Lai Lo wrote:
>Nilo,
>
>As a general advice to all Irix users, it is best to
>apply all the patches SGI supply for the C++ compiler
>and anything related to pthread. Otherwise
>it is almost certain that omniORB programs would
>crash.
>
>Back to your question, your stack trace indicated 
>that eg2_clt completed the call successfully, so it 
>exits and the connection is closed. A thread
>inside eg2_impl sees the connection closed and throws
>a COMM_FAILURE exception. This is normal. The
>exception is caught in a try block before
>the thread exits orderly. The stack trace seems to
>indicate that the compiler runtime failed to locate
>the try block and hence decided that the
>exception was not caught and so called abort.
>
>Looks to me it is a compiler bug.


You are COMPLETLY right! We installed the patches
corresponding to the system and compiler and
now eg2_impl and eg2_clt work fine. While eg2_impl
is running I can call eg2_clt as many times I want
and it prints out the good message, so everything
is perfect from this side now.

However, I was not able to make eg3 to work yet.
I think that the ORB is not initializing well because
the server dies from an CORBA::NO_RESOURCES
exception in CORBA::ORB::resolve_initial_references,
file corbaORB.cc, line 338.

I did step by step and it entered a loop to look
for available resources (for (CORBA::ULong index=0;...)
but it doesn't find any. If I understood well, it should find the "NameService" there that I think it
should have been read from the .cfg file. If I am
correct this file is not being read correctly or not
being read at all. But I don't know yet if it is true.

Any clue?

>Regards,
>
>Sai-Lai

Regards,

Nilo




-----== Sent via Deja News, The Discussion Network ==-----
http://www.dejanews.com/  Easy access to 50,000+ discussion forums


--=-=-=


-- 
Sai-Lai Lo                                   S.Lo@uk.research.att.com
AT&T Laboratories Cambridge           WWW:   http://www.uk.research.att.com 
24a Trumpington Street                Tel:   +44 223 343000
Cambridge CB2 1QA                     Fax:   +44 223 313542
ENGLAND


--=-=-=--