[omniORB] problem with circularly federated OB Naming and OmniNames ...

Philip Lijnzaad lijnzaad@ebi.ac.uk
Tue, 7 Mar 2000 15:00:55 GMT


Hi all,

this may not be all that important, but it does seem that there are bugs to
be quashed here. I seem to run into trouble when I'm trying to set up a
federated naming service, with an OB nameserv referring to an Omninames
server and vice versa, as follows:

OB nameserv contains

  stuff
  omni/ # this is a context bound (as context) to the OmniNames root context

OmniNames contains
  stuff 
  ob/ # this is a context bound (as context) to the OB nameserv root


In principle, on the OB side, omni/ob/omni/ob/omni/ob etc. should point to
itself, and likewise on the Omni side, ob/omni/ob/omni/ob should.

Such a circular path can be used in e.g. a context.resolve() call, but when
the length becomes deeper than say 4 hops, my clients (Mico's nsadmin and
Omni's nameclt ... yes, I'm being very eclectic here) get stuck: it suddenly
takes 20 secs to resolve, and it seems to lasts exponentially longer with
increasing length of the futile this/that/this/that links.

The problem does not occur when I'm similary federating two separate OB name
servers, nor when I'm federating two separate Omni name servers.  (I have had
no time to see what mico's nsd makes of it ...)

This is using OB 3.3, OB Names 3.3 (compiled with g++ 2.95.2, running on
IRIX, but same problem on SunOS) and Omni 2.8.0, (compiled with native c++
compiler on IRIX 6.5)

I have no idea where the real problem is, but just thought that you might be
interested. And any pointers are of course appreciated. Cheers,

                                                                      Philip
-- 
Not getting what you want is sometimes a wonderful stroke of luck.
-----------------------------------------------------------------------------
Philip Lijnzaad, lijnzaad@ebi.ac.uk | European Bioinformatics Institute,rm A2-24
+44 (0)1223 49 4639                 | Wellcome Trust Genome Campus, Hinxton
+44 (0)1223 49 4468 (fax)           | Cambridgeshire CB10 1SD,  GREAT BRITAIN
PGP fingerprint: E1 03 BF 80 94 61 B6 FC  50 3D 1F 64 40 75 FB 53