Re: [Vrs-development] IBM Token Ring Thingy...

[Chris Smith]

On Tuesday 07 May 2002 22:49, Bill Lance wrote:
> --- Chris Smith <address@hidden> wrote:
> > On Saturday 04 May 2002 00:45, Bill Lance wrote:
> > > --- address@hidden wrote:
> > > > I agree with the looseness of LDS within a
> >
> > cluster
> >
> > > > but a 2 phase commit will
> > > > provide better reliability and rollback support.
> > >
> > > How can we do it without locking up processes?
> >
> > Well, you're running within a middleware.
>
> Hehe ,, I keep trying to add more to this thread, but
> realized that GW will do some of what we are taking
> about, but other stuff will be application level, ...
> and I don't yet know where to draw the line.
>
> Guess we will have to wait to see what you are coming
> up with, Chris


If I do put transaction support in, then I'll have to do it properly, which 
will mean *buying* the XOpen XA Transaction Processing specification (about 
$200 I think).

This is the same spec that Oracle, Tuxedo etc comply with (oracle's 
Transaction Manager will link with Tuxedo's as they share the same API, which 
allows Tuxedo to control Oracle's transactions).  I'll have to do the same 
really, but it's not trivial!

In the short term we can create a stub API for the TP stuff, and some test 
harnesses to generate Rollback TP events so we can make sure that out TP 
complient application code behaves properly.

Incidently, Tuxedo has a concept of domains (DIFFERENT TO GOLDWATERS!!!!) and 
THEY DO NOT allow transactions to span domain boundaries.  It's very very 
complicated to do.  And I'll have to do it.  OH JOY! :o)

Chris (with an ever growing list of things to do... need to priorities!)

-- 
Chris Smith
  Technical Architect - netFluid Technology Limited.
  "Internet Technologies, Distributed Systems and Tuxedo Consultancy"
  E: address@hidden  W: http://www.nfluid.co.uk