[IPython-dev] [sage-notebook] Re: Using Google App Engine channels instead of zmq in new IPython
Mon Feb 13 17:22:07 CST 2012
On Mon, Feb 13, 2012 at 15:08, Fernando Perez <email@example.com> wrote:
> On Mon, Feb 13, 2012 at 3:01 PM, Jason Grout
> <firstname.lastname@example.org> wrote:
> > Those are some very nice diagrams!
> Indeed, all the credit goes to Min.
> > Thanks for the feedback. We already have a websocket<->zmq bridge,
> > I think now that GAE channels could basically replace the websocket
> > portion. I assume that would be much easier?
Yes, that would be much simpler, and the websocket:zmq layer is *extremely*
simple, as should a similar translation to any other messaging system.
> So the basic architecture
> > would be:
> > browser <-GAE channel-> Sage server on GAE <-GAE channel-> (bridge
> > IPython kernel)
> > where the part in parentheses would be the worker running on a compute
> > server, and the shim is the bridge between GAE channels and zmq. Does
> > seem more reasonable? What does the architecture of your websocket
> > look like?
> That's the part we use tornado for, so indeed if you wanted to go this
> route but run on GAE instead, you could replace our tornado-using code
> as indicated in your ascii-diagram above. The code that does this is
> mostly in the various modules here:
> That directory is the only part of IPython that uses tornado, which is
> effectively what you're talking about replacing.
True, but the architecture differs in that what is currently the
tornado-based NotebookApp covers everything from the browser to the zmq
side of the gae:zmq bridge. So there would have to be some work
duplication/proxying for the KernelManagers, which talk zmq to the kernels,
and would be separated from server code by an additional GAE layer in this
> IPython-dev mailing list
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the IPython-dev