Thanks but I am not really looking to roll my own. I am looking for steps to make an image of existing ec2 instance in ami format and run it on openstack. There are tons of references on how to run if you have the correct ami,aki and ari images but none that says how to get them from a running ec2 instance. I think I had seen a way to basically dd the root partition out of the ec2 instance over sash and use the vmlinuz and initrd for aki and ari respectively but that doesn't seem to work because obviously aws aki and ari images will not be kvm compliant. On Feb 6, 2014, at 8:37 AM, Steve Gordon <sgordon at redhat.com> wrote: > ----- Original Message ----- >> From: "Abhishek Lahiri" <aviostack at gmail.com> >> To: openstack-operators at lists.openstack.org >> Sent: Thursday, February 6, 2014 10:32:53 AM >> Subject: [Openstack-operators] Using ami on openstack >> >> >> Hi, >> >> Does anyone have concrete examples on how to run RHEL ami's on openstack? >> Specifically where can I get the RH or centOS ari and ami? How to get the >> ami from a running instance on AWS? If it is not possible to get an AMI then >> want is the best image type to build? >> >> Thanks >> Al > > I can't help with ARI or AMI images specifically, but there is a RHEL guest image available for use on OpenStack clouds that use KVM hypervisors here (subscriber link only): > > https://rhn.redhat.com/rhn/software/channel/downloads/Download.do?cid=16952 > > There is work under way to make equivalents for CentOS as well but in the meantime there is a build of one here: > > http://repos.fedorapeople.org/repos/openstack/guest-images/ > > In both cases the image is in QCOW2 format. If you are looking to "roll your own" I would refer to the "OpenStack Virtual Machine Image Guide": > > http://docs.openstack.org/image-guide/content/ > > Or the following Installation and Configuration Guide: > > https://access.redhat.com/site/documentation/en-US/Red_Hat_Enterprise_Linux_OpenStack_Platform/4/html/Installation_and_Configuration_Guide/Building_a_Custom_Disk_Image.html > > Thanks, > > Steve