478
478
Now we should be able to ssh the VM test-server1 from a server with the private
479
479
key we created above and see that vdb appears in /proc/partitions
481
##APPENDIX I - Economic Deployment
483
Although the above installation has indicated the ideal case for installing
484
OpenStack, there are often reasons for making more economic use of hardware at
485
the expense of robustness and failover capabilities. MAAS and Juju have several
486
capabilities to make this easier.
488
#### Deploying to specific nodes
490
It may be the case that your hardware is not identical. Some nodes may have more
491
compute power, some may have the additional NICs to run services which require
494
In this case it is possible to target services to be deployed on a certain type
495
of node. This requires that the nodes be [tagged in MAAS][MAAS_tags] beforehand.
496
The services can then be deployed specifically to that type of node using a Juju
497
feature called 'constraints'. For example, if you had a number of nodes which
498
were tagged 'compute', you could use constraints to tell Juju to only deploy
499
a service to a node with that tag:
502
juju deploy --constraints "tags=compute" nova-compute -n 2
505
#### Deploying multiple services to a single node
507
Juju can deploy multiple services to the same physical node by use of the '--to'
508
option. This can be useful for combining many services with light requirements
509
onto the same physical node.
510
This only requires that you know the machine number of the node you wish to
511
deploy to (as returned by the `juju status` command). For example, assume we
512
have already deployed `nova compute`:
515
juju deploy nova-cloud-controller
518
Now we run a `juju status` command and find amongst the output:
521
nova-cloud-controller:
522
charm: cs:trusty/nova-cloud-controller-36
526
- nova-cloud-controller
528
nova-cloud-controller/0:
534
...which indicates that machine "3" has been used for the deployment. We could
535
then deploy other services to the same node like this:
538
juju deploy --to 3 openstack-dashboard
539
juju deploy --to 3 glance
542
It is also possible to containerise these deployments with LXC, to provide a more robust separation of services:
545
juju deploy --to lxc:3 cinder
548
#### Example - an OpenStack control node
550
An example OpenStack control node could be a machine running the following
553
- Nova Cloud Controller
557
- OpenStack Dashboard
560
The recommended specs for the this node would be:
562
| Node Attribute | Specification |
563
|----------------------|-----------------------------------|
564
| Number of CPUs | 4 |
566
| Number of NIC ports | 2 (PCE Management and VM Network) |
483
571
[oog]: http://docs.openstack.org/ops/