Subject:
|
DISTRIBUTED/PARALLEL CLUSTER for legOS through n*RCX
|
Newsgroups:
|
lugnet.robotics.rcx.legos
|
Date:
|
Wed, 15 Jan 2003 15:41:59 GMT
|
Viewed:
|
2960 times
|
| |
|
|
Hi all. I'm taking a course on parallel/distributed operating systems. So
this post attempts to find out if someone has thought, or has implemented
some kind of distributed application over BrickOS/legOS (from now legOS, for
this post). I know there's support for IP/IR networking. But what I'd like
to do is to have an extra layer, that hides to the end user (the
programmer), the managment of the n*RCX, that conform this "cluster". This
would mean that to the programmer, this megaRCX would have 3*n
sensors/motors (where n is the # of RCXs that would conform the "cluster").
Of course every RCX attached to the cluster should process/take decisions
over his sensors/motors, and then just inform status to the main scheduler
(is just an idea of how to implement it, I'm not an expert on this).
Also, would be great to share the memory (which is so scarce on a single
RCX). In a nutshell: to have MPMD (multiple process over multiple data),
over this megaRCX. I know that there's a big problem with the speed (the IR
port), but for now, we don't want this "cluster" for a hard real time app,
do we? So, would be great to recieve comments about what's been done until
now, feasibility (is out of the planet to put this extra layer over what's
done until now of legOS?), pros/con, and of course, any kind of comment.
PD: also would be great if Markus (or any of his lieutenants), could give me
some light on any of the aspects I've mentioned above.
Best Regards,
CarlosLuis Rendon M.
|
|
Message has 3 Replies:
7 Messages in This Thread:
   
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
