To the best of my understanding there are two parts to this, neither of which is fully where it ought to be:
1. It shouldn't be a required parameter. If you give it a volume with everything it needs and not an image it should try to boot from that without throwing an exception. Horizon only enforces the requirement because Nova does. 2. As an optional parameter, specifying an image should allow you to create the instance with that image loaded onto the volume, which is a very important part of the workflow for creating your own bootable volumes. This feature doesn't currently exist in Nova, however. I also don't know offhand if there are blueprints or bug reports for either of those... - Gabriel From: [email protected] [mailto:[email protected]] On Behalf Of Vishvananda Ishaya Sent: Saturday, May 26, 2012 8:35 PM To: Lorin Hochstein Cc: [email protected] ([email protected]) Subject: Re: [Openstack] Why is an image required when booting from volume If there is a separate kernel and ramdisk needed for the boot from volume, it is pulled from image properties. Otherwise it is basically useless. Vish On May 26, 2012, at 8:22 AM, Lorin Hochstein wrote: I'm trying to figure out boot from volume, both so I can use it and so I can add it to the docs. It seems that when calling "nova boot" or using Horizon, you need to specify an image. Why is that? I naively tried to create a volume image by creating a volume and then doing on my volume server: dd if=/tmp/precise-server-cloudimg-amd64-disk1.img of=/dev/nova-volumes/volume-0000000d Then I tried this: $ nova boot --flavor 2 --key_name lorin --block_device_mapping /dev/vda=13:::0 test Which generated an error: Invalid imageRef provided. (HTTP 400) If I try to specify an image, it at least attempts to boot: $ nova boot --flavor 2 --key_name lorin --block_device_mapping /dev/vda=13:::0 --image 7d6923d9-1c13-4405-ba0c-41c7487dd6bc test I noticed that the devstack example specifies an image: https://github.com/openstack-dev/devstack/blob/master/exercises/boot_from_volume.sh: VOL_VM_UUID=`nova boot --flavor $INSTANCE_TYPE --image $IMAGE --block_device_mapping vda=$VOLUME_ID:::0 --security_groups=$SECGROUP --key_name $KEY_NAME $VOL_INSTANCE_NAME | grep ' id ' | get_field 2` Looking at nova/api/openstack/compute/servers.py, it does look like _image_uuid_from_href() is called regardless of whether we are booting from volume or not. What is "--image" used for when booting from volume? Take care, Lorin -- Lorin Hochstein Lead Architect - Cloud Services Nimbis Services, Inc. www.nimbisservices.com<https://www.nimbisservices.com/> _______________________________________________ Mailing list: https://launchpad.net/~openstack Post to : [email protected]<mailto:[email protected]> Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
_______________________________________________ Mailing list: https://launchpad.net/~openstack Post to : [email protected] Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
