Add the md5 image hash check, if the image exists, check the md5
hash, if the hash is not the same, remove the old image and download
new one.
Change-Id: I2232129a814e89aaac9bf54ee16ede983fff5b48
Starting in 16.04 Ubuntu no longer installs python v2.7 by default.
Instead the default version of python is v3.5. If you still need v2.7
then add the `python` package to this list
Change-Id: Ieb47e319ef2124570cc72acf1219c5b9d9fc9814
SSHKEY is originally set in setup-host.sh, so this allows
setup-cobbler.sh to run independently of that script.
Change-Id: I4793e6a5db7359bec0eaa7b6a66efa7d58084126
the tilde symbol may cause problem when VMs don't have the same
folder path as the host, replace it with tmp will resolve this
issue.
Change-Id: I14114d018fa9dc9fa332f1e34077f88d75995f21
Starts from Newton release, Openstack needs ubuntu 16.0.4(xenial)
as the operation system to depoy openstack. add the xenial image
will help people to test or develop openstack projects.
Change-Id: Ida3a2102664abc4a69e8d94f6c8d6cc177a8c4c3
Add PRE_CONFIG_OSA to set up the user_variable and all the other
parameters for OSA powered cloud, and move the RUN_OSA to the playbook
part.
Change-Id: Ia962e42466dd2a123e30b1a9c37e60a371710842
Remove the opensuse expired key and opensuse cobbler repo after
cobbler being installed. OSA deployment doesn't need it.
And fix a typo and a link of README.rst
Change-Id: I24f5d823806f039d879dd69762b2eac3a6005628
Replace the service xinetd stop and start to service xinetd restart
, the stop will cause the unknow instance error and stop the
rebuild
Change-Id: I1a8a7a72f51b3d4dcdca3b0158f7771efd567032
Add the rm ~/.ssh/known_hosts for the Rerunning the build script
part in the instruction.
Change-Id: Iba7610e00f07bb61cd8b11892e208ea4c67a1d75
Closes-Bug: #1635037
Add an option $RUN_OSA to the deploy-osa.sh to let user have more
choices, if user want to deploy another OSA powered cloud on AIO-
multinode, he can switch it to false and then use the existing env
to deploy the cloud he wants to.
Change-Id: Ief8c701f596595607d3b86ac03f17d6690f5954f
Closes-Bug: #1634582
This change is needed because nova no longer provides flavors.
Change-Id: Icaf97d94007aef7fd54691174cab6ed2db89404b
Signed-off-by: Kevin Carter <kevin.carter@rackspace.com>
This change simple makes the system deploy a network stack and create
various images in a more production like fashion.
Change-Id: Ib316d5fcab08aa31f2965175a4fb31f1d3566228
Signed-off-by: Kevin Carter <kevin.carter@rackspace.com>
By default we generally deploy the network agents onto the infra hosts.
This patch simply forces the multi-node-aio to follow that process which
is being done to lower resource consumption.
Change-Id: I6612a85cb669596893dbd3ee0a858a8ee5476729
Signed-off-by: Kevin Carter <kevin.carter@rackspace.com>
The scp command will fail if the directory in the preceeding for
loop is empty. This adds a test to ensure that $i is acutally a
file name.
Change-Id: Ib2a3c929a7a15b290c2cf09c05d4079a1ec8c6e6
Closes-Bug: #1621160
In one interation a debug file was used on my test node which remained
as I posed this commit. Sadly the attmpted inclusion breaks the end of
the deploy VMs process. To resolve this, the bad file name has been
removed from the loop.
Change-Id: I44aafc8aa26bdad7efc43835719003637a608ee6
Signed-off-by: Kevin Carter <kevin.carter@rackspace.com>
The successerator function retries a failed playbook
that might of failed due to network saturation issues.
i.e. target host ssh unreachable errors.
Change-Id: I2164a22555474749228fcb278d34885cf154a743
This change imports the code from [ https://github.com/cloudnull/osa-multi-node-aio ]
so that others in the OSA community can benifit from and contribute to
the project.
Change-Id: Id9a2d928f7e8bbd185725349e8ce51344f19daa4
Signed-off-by: Kevin Carter <kevin.carter@rackspace.com>
