Ask Your Question
2

implement Openstack on multiple laptops and an hadoop cluster over it [closed]

asked 2015-02-09 01:04:16 -0600

Akash gravatar image

updated 2015-02-09 09:26:14 -0600

smaffulli gravatar image

I have a project to do, in which a demonstration has to be made of performing string manipulation algorithms on huge amount of data(in GBs) stored under hadoop. This hadoop framework should again be installed on the openstack platform and I must demonstrate this by using multiple laptops(i.e not by using some virtual machines on a single laptop). I am new to this Big Data and Cloud field. Can you please give me some guidelines on how to implement this or some docs having the steps to implement this?

edit retag flag offensive reopen merge delete

Closed for the following reason too subjective and argumentative by smaffulli
close date 2015-02-09 09:26:23.357503

add a comment

2 answers

Sort by ยป oldest newest most voted
0

answered 2015-02-09 09:27:27 -0600

smaffulli gravatar image

You had your answer so I'm closing the question that is way too wide to be useful. Please read https://ask.openstack.org/faq and if you want to have a conversation, consider using the OpenStack mailing list http://lists.openstack.org

edit flag offensive delete link more
add a comment
0

answered 2015-02-09 02:35:27 -0600

9lives gravatar image

updated 2015-02-09 06:11:00 -0600

you might take a look at OpenStack Sahara project(EDP as Service) based on Hadoop, you can benefit from Sahara with two things, No1. hadoop cluster provisioning; No2. EDP on Hadoop cluster.

The only thing you might need to consider is if hadoop jobs need to be running on vm or physical box, currently Sahara only support vm based cluster.

For more information refer to the Sahara wiki here

@Akash

Theoretically speaking, yes. you may implement your cloud with multiple nodes(laptops) via two options:

  • Single dc deployment with multi-nodes

    1 controller + 1 network node + many compute nodes

  • multiple region deployment with shared 1 keystone service:

    region 1: 1 controller + 1 network node + many compute nodes;

    region 2: 1 controller( point to the keystone service in region 1) + 1 network node + many compute nodes;

Hope that helps!

Vic

edit flag offensive delete link more

Comments

Thanx, but what if I want to use actual physical boxes??what should I use in that case??

Akash gravatar imageAkash ( 2015-02-09 03:13:34 -0600 )edit

if you really want to use physical box, IMO there are two options: 1. setup hadoop cluster manually 2. use ironic or other baremetal provisioning tools to install hadoop on physical box which might need more effort than options 1.

9lives gravatar image9lives ( 2015-02-09 03:38:12 -0600 )edit

By using Sahara, can we implement the cloud over multiple laptops?? And thanx again

Akash gravatar imageAkash ( 2015-02-09 04:51:54 -0600 )edit
add a comment

Get to know Ask OpenStack

Resources for moderators

Question Tools

2 followers

subscribe to rss feed

Stats

Asked: 2015-02-09 01:04:16 -0600

Seen: 567 times

Last updated: Feb 09 '15

Related questions

vm on multiple networks

devstack - swift and Heat

Neutron service not starting

No module named os_client_config

Some compute nodes are listed as down.

What happened if compute node machine restarts or stop working ??

How to see compute nodes from horizon?

How to configure two public network

Is this a kolla or docker-py issue?

Do we need to stop swift servers while adding new disk?