combined compute AND controller node

asked 2014-11-13 08:42:13 -0600

madejackson gravatar image

Hi there folks, I got a short question:

I got 2 physical nodes (Dual-CPU Xeon E5450, 32GB RAM, Dual-CPU Xeon E5420, 12GB RAM). Now I'd like to automate instance-deployment with Openstack (Ubuntu 14.04, KVM). Now my Question: Can I install the Controller on one node and use still both as compute nodes. Or is it possible to use both nodes as controller (high available) and still use both nodes as compute-nodes.

I think when I install the controller on one node and the compute on the other node, I do not user the full Potential of the Servers.

I know, Openstack seems to be overpowered for this, but I'd like to increase the amount of nodes in future, so it should be easy expendable.

edit retag flag offensive close merge delete

2 answers

Sort by ยป oldest newest most voted

answered 2014-11-13 11:58:45 -0600

dbaxps gravatar image

You can for sure ( with no problems) install the Controller on one node and use still both as compute nodes
To use both boxes as Controllers will require replication MySQL-Galera database. View for details Introduction to OpenStack High Availability. I believe that second configuration is not easy to implement.

edit flag offensive delete link more

answered 2014-11-14 04:09:43 -0600

You can do one of the following with your 2 nodes: a) 1. Controller + 2. Compute b) 1. (Controller + Compute) + 2. Compute c) 1. (Controller + Compute) + 2 (Controller + Compute) [Hight Availability mode]

You can do either of the above. Option "B" is optimal solution. If you really need HA then you can go for Option "C".

edit flag offensive delete link more

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2014-11-13 08:41:32 -0600

Seen: 227 times

Last updated: Nov 14 '14