Hey people,
We couldn't have the sixth meet because, everybody had some other
engagement. But finally , we are back again.
This saturday, at 8 pm , same room, on BeFS. What do you say we make
it a 2-hour session and finish off BeFS ?
That leaves us with two more saturdays this semester ( Problems on
30th because of compres , anybody ??). I think we should do Hadoop on
one and FUSE on another. I keep coming back to FUSE because, the idea
seems pretty cool to me and is something practical to get our hands
dirty. If we're cool with FUSE, I can try it out and may be talk about
it in the last session.Somebody has to take up hadoop though.The paper
posted by Aastha from the other mail looked pretty good, so if anybody
is going to study and talk about it that will be really great. I can
also go over it so that we can have a discussion like always.
Also we are planning to document what we have studied this semester,
so we have something concrete and it'll also save others from going
over all the papers again.These are the modules involved that I can
picture now :
Filesystems and initial approach to file layout algorithms
File descriptors (process level, kernel level tables), vnodes
FFS Summary
Log structured file systems
BeFS(Pending)
Hadoop(Pending)
FUSE(undecided)
Do jump in with your ideas.....
Cheers
Ambarish