On Wed, Jan 12, 2011 at 6:05 AM, Zooko O'Whielacronx <[email protected]> wrote: > On Tue, Jan 11, 2011 at 9:33 AM, Pandelis Theodosiou <[email protected]> > wrote: >> >> This is great, but I can't currently have a server that is constanly >> connected to the net. >> So, count me in for volunteergrid3. > > :-) > >> One the other hand, (and since point 5 is missing :), why should the lack of >> a reliable rever stop me - or anyone else? >> >> Isn't Lahoe designed for this kind of situations too? > > It is certainly good at maintaining confidentiality and integrity even > in this situation, because all of the data is encrypted and > integrity-checked. It is harder to maintain availability, because even > if the data *is* erasure-coded so that, say K=3 servers are required > out of N=10 servers total, you might still have too many of the > servers offline at once. > > Some of the problems may be work-aroundable by using it in a different > way. I don't know but I'd love to find out. > >> What I'm proposing is (another one), more experimental grid, where servers >> can be unreliable, like notebooks, your old pc you boot once in a week, etc? >> >> I have no idea what configuration it would need, but if there are other >> volunteers, I'm in for that! > > Go for it! > > I would be happy to learn the results of this experiment. People may > need to use higher expansion factors (larger N and/or smaller K), or > just to get used to their files being unavailable some of the time, > depending on the pattern of how many of their fellows's laptops are > open or closed. :-) > >> PS. Oh, and since this is my first post here, congrats for the nice tool you >> made. >
I'd volunteer for this new grid system as well. I've a server and space in a datacenter with some spare resources jimmy -- http://www.sgenomics.org/~jtang/ _______________________________________________ tahoe-dev mailing list [email protected] http://tahoe-lafs.org/cgi-bin/mailman/listinfo/tahoe-dev
