D-BUS for Network RPC

Havoc Pennington hp@redhat.com
Wed Jan 19 08:53:12 PST 2005


btw, the spec and tutorial were both made much more accurate in the last
couple days in CVS, vs. what's on the web site.

On Wed, 2005-01-19 at 10:42 -0500, Stephen J. Scheck wrote:
> - How does the design and architecture of D-BUS make it optimized for 
>   local use, and thus would make it unsuitable/inefficient for network 
>   RPC? Is there any reason why it wouldn't make a good candidate for
>   simple and light-weight network RPC applications?

It probably works OK, it just hasn't been designed or thought about in
that context.

> - How stable/tested is the TCP transport?

The code is virtually all the same as the local socket transport, so I
wouldn't expect problems except for one: there's no reasonable
authentication mechanism implemented for TCP.

> - How difficult would it be to make the TCP transport work with   
>   external security layers like IPsec or SSL/TLS? Could these
>   be configured easily and transparently to wrap a D-BUS TCP connection
>   or would modifications be necessary to D-BUS? And how would they 
>   interact, if at all, with the SASL authentication used by D-BUS?

Unknown. You probably have to be prepared to hack on D-BUS a bit to get
some auth/encryption in place.

> - Is there any other documentation or examples for the TCP transport? 
>   The docs don't seem to specify the address format for TCP transport.

Hmm, it's something pretty simple IIRC like port=whatever,host=whatever

If you're going to get this working you probably have to dive into the
dbus code a bit, and the address format should be simple to find in
there. (And add to the docs ;-))


More information about the dbus mailing list