There are storage limits in GAE that are very different than EMR or EC2. You don't just get complete access to a linux machine.
This is the key to the magical scaling provided by GAE, but it does impose a small porting burden. On Wed, Aug 24, 2011 at 4:06 PM, Sean Owen <[email protected]> wrote: > I don't know of anyone who has used GAE. I don't know of a reason it > wouldn't work, as it's just Java and doesn't (I think) use anything > that's unavailable on GAE. That is, in theory you can run Hadoop on > GAE, I think. I don't know how hard it is. In practice I am sure it's > much harder than just using Amazon EMR. > > 2011/8/24 Matthias Böhmer <[email protected]>: > > So, since that there was no answer I conclude that nobody is running > > Mahout on GAE -- or am I wrong? > > > > 2011/8/17 Matthias Böhmer <[email protected]>: > >> Hi there, > >> > >> I know there was some discussion about running Hadoop on Google App > Engine some month ago. But I am interested if anybody did investigate this > topic in the meanwhile. So, is anybody running Hadoop on GAE for > collaborative filtering? > >> > >> Cheers, > >> Matthias > >> > >> > > >
