Re: Finding an initial name-servive using multicasting

[Magnus Bergman <magnus bergman observer net>]

> 1. I am a newbie with CORBA and Orbit.  There is an ORBit2 command
>    line option that triggers the generation of the shortened
>    "corbaloc" URL instead of IORs:
> 
>      orbit-name-server-2 -ORBIIOPIPv4=1 -ORBIIOPIPSock=2809 \
>                          -ORBIIOPUNIX=0 -ORBCorbaloc=1
> 
>    The last option makes orb.object_to_string() generate strings
>    in the "corbaloc" format, such as the one below.

Thanks, that leads me a step closer to a solution. But I'm slightly
confused, orbit doesn't seem to include orbit-name-server-2 any longer,
it just seems to be implemented in a library. That's the deal with that?
(I might be even more of a newbie.)


> 2. I've just applied the patch by Alejandro García Castro that
>    allows to refer to the ORBit2 implementation of the naming
>    service by a shortcut name ("key") such as NameService:
> 
>     
http://mail.gnome.org/archives/orbit-list/2004-November/msg00015.html
> 
>    The patch allowed to use the otherwise unimplemented --key
>    option of orbit-name-server-2:
> 
>      orbit-name-server-2 -ORBIIOPIPv4=1 -ORBIIOPIPSock=2809 \
>                          -ORBIIOPUNIX=0 -ORBCorbaloc=1 \
>                          --key=NameService

This I probably don't understand since I can't see the use for it. So I
just assume it's nothing that I need.


> 3. As for the discovery, I used the OpenSLP library and daemon to
>    register the name service's host name and port number under an
>    artificial service name such as "service:corbans.COMPANYNAME";.
>    I understand my solution is more complicated because it
>    involves 2 name servers (slpd and orbit-name-server-2).

That sounds very interesting. Perhaps exactly what I have in mind. I
will look closer at SLP and if that alone might be the solution to the
problem. There was really no strong reason for using howl (the only free
implementation of mdns that I know of) except that it seemed nice and
there was some talk about using it in gnome (which the later decided to
do). As long as it's a standard and not some random hack, then I'm
satisfied.


>    (I disabled the 6 second Directory Agent discovery in the
>    OpenSLP library against the authors' recommendations).
> 
>    Perhaps, one could reuse OpenSLP's service attributes to store
>    full IORs.  This would circumvent the standard CosNaming
>    service altogether.

Well, I don't want to circumvent the CosNaming service since that is the
corba standard. So I consider the double name service solution to be
right one (at least as long as orbit can't do multicast discovery by it
self). Registering the IOR (or other identifier) of the corba name
service was exactly what I had in mind. The only problem was that the
length of the IOR exceeded the maximum length of an mdns attribute. It
could be solved with a dirty hack or two, but I find it important to
follow the mdns standard too.


>    I haven't tried mdns, but I will be interested hearing from
>    others about it.

I just had a quick look at the SLP RFCs and it seems that mdns is a
little simpler, but otherwise the seem pretty much the same. Is there
anything specific you wonder about mdns?