[omniORB] Missing "else" statement in omniORB 4.1.4 retry handling?

Peter Klotz peter.klotz at aon.at
Wed Jul 28 19:21:59 BST 2010


Hello Duncan

The attached patch (against SVN trunk) adds an "else" statement that in 
our opinion got lost in omniORB 4.1.4 in this checkin (file omniObjRef.cc):

   Revision 1.4.2.7  2008/10/28 15:54:51  dgrisby
   Internal CommFailure exception escapes after failed-on-forward call
   that is not retried.


There were two identical code blocks in omniORB 4.1.3 that were altered 
by the above checkin. In omniORB 4.1.4 a diff shows that the "else" 
statement misses in one of the blocks.

The erroneous behavior we are seeing is this:

* Client contacts Naming Service and establishes a connection
* Later the Naming Service closes the connection due to its 
inConScanPeriod setting
* Client contacts Naming Service again before its outConScanPeriod is over
* Client sees the dead connection but instead of performing a retry it 
immediately throws a COMM_FAILURE exception

omniORB 4.1.3 and patched 4.1.4 clients work as expected and open a new 
connection to the Naming Service.

Can you confirm the correctness of the patch?

Regards, Peter.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: omniORB-4.1.5-20100715-MissingElseStatement.patch
Type: text/x-patch
Size: 734 bytes
Desc: not available
Url : http://www.omniorb-support.com/pipermail/omniorb-list/attachments/20100728/7ce6c9e9/omniORB-4.1.5-20100715-MissingElseStatement.bin


More information about the omniORB-list mailing list