[Spread-users] spread as a shared library?
Neil Conway
neilc at samurai.com
Thu Jun 30 00:39:41 EDT 2005
Theo Schlossnagle wrote:
> going through the effort of bulding a library that imposes on users that
> they can only have one "connection to spread" would be disappointing.
> There are many legimate situations in which it can be advantageous to
> have multiple sessions from within a single app
Yeah, I agree. I meant that our motivation is that we don't need to
multiplex multiple clients through a single Spread daemon, but of course
that's no reason to only allow a single Spread connection per process
using the library.
> There are many simple configuration file "mistakes" (like leaving one
> server out of the list on a single machine) that will cause Spread to
> not work with a non-obious error message.
Yes, Spread error reporting and recovery in general could do with a lot
of work.
> Also, many common network configuration errors that don't cause
> problems for other protocols can reek havoc on a Spread cluster.
Can you elaborate on what you mean?
> If you do surgery that dramatic, be prepared for a Frankenstein. I'd
> recommend a rewrite based on the protocols and concepts.
Yes, I've been considering that as well. It would let us get out from
under the BSD-with-advertising license (I'd prefer 3-clause BSD), but
I'm not keen to reinvent the wheel unless there's a compelling reason we
can't refactor Spread to do what we need.
> It would allow a different messaging API to be used and allow much
> higher performance as you could effectively do zero-copy messaging in
> many cases.
[...]
> This is on our todo list here at OmniTI (along with an epic laundry list
> of enterprise enhancements).
If you have any TODO items beyond Spread's TODO, feel free to share
them. We're looking at making a bunch of improvements to Spread anyway,
but if you've already run into areas where it could be improved that
would be helpful.
-Neil
More information about the Spread-users
mailing list