Renato Marroquín Mogrovejo wrote:
Thanks everybody,
I will definitely try to get in contact with OpenCirrus, and if not I can at
least test with some amazon hw so I can prove the application concept.
There's a lot to be said for bringing up 3 or 4 virtualbox or vmware VMs
and playing with that locally, as it does make sure that you have
distribution right. CPU performance on VMs is pretty good, it's only
disk IO that (currently) suffers.
>And
the Whirr project is really interesting, I will give it a look specially
because we really don't want to depend on a single service provider.
Thanks again for the suggestions.
I've been doing some work on GUIs/tooling too; this is something that
could really front end what whirr does, though it also needs a
cross-infrastructure HA story for managing the persistent data the
tooling needs too -user accounts, current set of machines, etc etc.
http://www.slideshare.net/steve_l/farming-hadoop-inthecloud
Renato M.
2010/6/30 Jeff Hammerbacher <[email protected]>
If you want to use an open API and be cloud-provider agnostic, check out
Whirr: http://incubator.apache.org/projects/whirr.html.
On Wed, Jun 30, 2010 at 10:52 AM, Yan, Weizhong (GE, Research)
<[email protected]>wrote:
Why don't you just simply try Amazon elastic MapReduce?
http://aws.amazon.com/elasticmapreduce/
-----Original Message-----
From: Renato Marroquín Mogrovejo [mailto:[email protected]]
Sent: Wednesday, June 30, 2010 11:47 AM
To: [email protected]
Subject: Hadoop Platform
Hi everyone,
I read a while ago that maybe IBM or Yahoo! were going to provide a
"cloud"
for academic purposes. Does anybody know if those intentions
materialised?
or if there any other ( free or not so expensive ) platform to test
Hadoop
jobs??? I mean, obviously there would be several limitations on it, but I
don't have access to a cluster and I have some Hadoop jobs I need to test
them, so I can prove to my people the power of it (:
Thanks in advance.
Renato M.
