As far as I know you can create any bucket that doesn't already exist, and then you own it until you delete it. So actually, there's a bit of a domain-name-style land-grab opportunity here. I'd be pleased to hear that someone here had created a "mahout" bucket just to have on-hand.
The namespace issues are curious... I'd have expected all bucket namespaces to be prefixed with a user ID of some kind. On Sat, Feb 27, 2010 at 5:02 PM, Robin Anil <robin.a...@gmail.com> wrote: > yeah I am creating a split joiner class in utils might be useful for other > things as well. > > How do I claim mahout bucket. They have each bucket associated with each > secret and access key pair right > > Robin > > > On Sat, Feb 27, 2010 at 10:29 PM, Sean Owen <sro...@gmail.com> wrote: > >> Can't that be sharded? I'm accustomed to any input file being a series >> of files... *-00000, *-00001, etc. This also assists distributing >> input to mappers, anyhow, regardless of size caps. >> >> Can't you just claim the "mahout" bucket? >> >> On Sat, Feb 27, 2010 at 4:53 PM, Robin Anil <robin.a...@gmail.com> wrote: >> > Bah! Humbug! S3 has a 5Gb limit and wikipedia compressed seq file is >> 5.9GB >> > >> > On Sat, Feb 27, 2010 at 10:08 PM, Robin Anil <robin.a...@gmail.com> >> wrote: >> > >> >> Just curious. Was trying to put my wikipedia seqfiles for public >> >> consumption. I can put it on robinanil bucket. mahout would have been >> nicer. >> >> >> >> >> >> Robin >> >> >> > >> >
