Store metadata on a configuration drive

Store metadata on a configuration drive

You can configure OpenStack to write metadata to a special configuration drive that attaches to the instance when it boots. The instance can mount this drive and read files from it to get information that is normally available through the metadata service. This metadata is different from the user data.

One use case for using the configuration drive is to pass a networking configuration when you do not use DHCP to assign IP addresses to instances. For example, you might pass the IP address configuration for the instance through the configuration drive, which the instance can mount and access before you configure the network settings for the instance.

Any modern guest operating system that is capable of mounting an ISO 9660 or VFAT file system can use the configuration drive.

Requirements and guidelines

To use the configuration drive, you must follow the following requirements for the compute host and image.

Compute host requirements

  • The following hypervisors support the configuration drive: libvirt, XenServer, Hyper-V, and VMware.

    Also, the Bare Metal service supports the configuration drive.

  • To use configuration drive with libvirt, XenServer, or VMware, you must first install the genisoimage package on each compute host. Otherwise, instances do not boot properly.

    Use the mkisofs_cmd flag to set the path where you install the genisoimage program. If genisoimage is in same path as the nova-compute service, you do not need to set this flag.

  • To use configuration drive with Hyper-V, you must set the mkisofs_cmd value to the full path to an mkisofs.exe installation. Additionally, you must set the qemu_img_cmd value in the hyperv configuration section to the full path to an qemu-img command installation.

  • To use configuration drive with the Bare Metal service, you do not need to prepare anything because the Bare Metal service treats the configuration drive properly.

Image requirements

  • An image built with a recent version of the cloud-init package can automatically access metadata passed through the configuration drive. The cloud-init package version 0.7.1 works with Ubuntu, Fedora based images (such as Red Hat Enterprise Linux) and openSUSE based images (such as SUSE Linux Enterprise Server).
  • If an image does not have the cloud-init package installed, you must customize the image to run a script that mounts the configuration drive on boot, reads the data from the drive, and takes appropriate action such as adding the public key to an account. You can read more details about how data is organized on the configuration drive.
  • If you use Xen with a configuration drive, use the xenapi_disable_agent configuration parameter to disable the agent.

Guidelines

  • Do not rely on the presence of the EC2 metadata in the configuration drive, because this content might be removed in a future release. For example, do not rely on files in the ec2 directory.
  • When you create images that access configuration drive data and multiple directories are under the openstack directory, always select the highest API version by date that your consumer supports. For example, if your guest image supports the 2012-03-05, 2012-08-05, and 2013-04-13 versions, try 2013-04-13 first and fall back to a previous version if 2013-04-13 is not present.

Enable and access the configuration drive

  1. To enable the configuration drive, pass the --config-drive true parameter to the openstack server create command.

    The following example enables the configuration drive and passes user data, two files, and two key/value metadata pairs, all of which are accessible from the configuration drive:

    $ openstack server create --config-drive true --image my-image-name \
      --flavor 1 --key-name mykey --user-data ./my-user-data.txt \
      --file /etc/network/interfaces=/home/myuser/instance-interfaces \
      --file known_hosts=/home/myuser/.ssh/known_hosts \
      --property role=webservers --property essential=false MYINSTANCE
    

    You can also configure the Compute service to always create a configuration drive by setting the following option in the /etc/nova/nova.conf file:

    force_config_drive = true
    

    Note

    If a user passes the --config-drive true flag to the nova boot command, an administrator cannot disable the configuration drive.

  2. If your guest operating system supports accessing disk by label, you can mount the configuration drive as the /dev/disk/by-label/configurationDriveVolumeLabel device. In the following example, the configuration drive has the config-2 volume label:

    # mkdir -p /mnt/config
    # mount /dev/disk/by-label/config-2 /mnt/config
    

Note

Ensure that you use at least version 0.3.1 of CirrOS for configuration drive support.

If your guest operating system does not use udev, the /dev/disk/by-label directory is not present.

You can use the blkid command to identify the block device that corresponds to the configuration drive. For example, when you boot the CirrOS image with the m1.tiny flavor, the device is /dev/vdb:

# blkid -t LABEL="config-2" -odevice
/dev/vdb

Once identified, you can mount the device:

# mkdir -p /mnt/config
# mount /dev/vdb /mnt/config

Configuration drive contents

In this example, the contents of the configuration drive are as follows:

ec2/2009-04-04/meta-data.json
ec2/2009-04-04/user-data
ec2/latest/meta-data.json
ec2/latest/user-data
openstack/2012-08-10/meta_data.json
openstack/2012-08-10/user_data
openstack/content
openstack/content/0000
openstack/content/0001
openstack/latest/meta_data.json
openstack/latest/user_data

The files that appear on the configuration drive depend on the arguments that you pass to the openstack server create command.

OpenStack metadata format

The following example shows the contents of the openstack/2012-08-10/meta_data.json and openstack/latest/meta_data.json files. These files are identical. The file contents are formatted for readability.

{
    "availability_zone": "nova",
    "files": [
        {
            "content_path": "/content/0000",
            "path": "/etc/network/interfaces"
        },
        {
            "content_path": "/content/0001",
            "path": "known_hosts"
        }
    ],
    "hostname": "test.novalocal",
    "launch_index": 0,
    "name": "test",
    "meta": {
        "role": "webservers",
        "essential": "false"
    },
    "public_keys": {
        "mykey": "ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAAAgQDBqUfVvCSez0/Wfpd8dLLgZXV9GtXQ7hnMN+Z0OWQUyebVEHey1CXuin0uY1cAJMhUq8j98SiW+cU0sU4J3x5l2+xi1bodDm1BtFWVeLIOQINpfV1n8fKjHB+ynPpe1F6tMDvrFGUlJs44t30BrujMXBe8Rq44cCk6wqyjATA3rQ== Generated by Nova\n"
    },
    "uuid": "83679162-1378-4288-a2d4-70e13ec132aa"
}

Note the effect of the --file /etc/network/interfaces=/home/myuser/instance-interfaces argument that was passed to the openstack server create command. The contents of this file are contained in the openstack/content/0000 file on the configuration drive, and the path is specified as /etc/network/interfaces in the meta_data.json file.

EC2 metadata format

The following example shows the contents of the ec2/2009-04-04/meta-data.json and the ec2/latest/meta-data.json files. These files are identical. The file contents are formatted to improve readability.

{
    "ami-id": "ami-00000001",
    "ami-launch-index": 0,
    "ami-manifest-path": "FIXME",
    "block-device-mapping": {
        "ami": "sda1",
        "ephemeral0": "sda2",
        "root": "/dev/sda1",
        "swap": "sda3"
    },
    "hostname": "test.novalocal",
    "instance-action": "none",
    "instance-id": "i-00000001",
    "instance-type": "m1.tiny",
    "kernel-id": "aki-00000002",
    "local-hostname": "test.novalocal",
    "local-ipv4": null,
    "placement": {
        "availability-zone": "nova"
    },
    "public-hostname": "test.novalocal",
    "public-ipv4": "",
    "public-keys": {
        "0": {
            "openssh-key": "ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAAAgQDBqUfVvCSez0/Wfpd8dLLgZXV9GtXQ7hnMN+Z0OWQUyebVEHey1CXuin0uY1cAJMhUq8j98SiW+cU0sU4J3x5l2+xi1bodDm1BtFWVeLIOQINpfV1n8fKjHB+ynPpe1F6tMDvrFGUlJs44t30BrujMXBe8Rq44cCk6wqyjATA3rQ== Generated by Nova\n"
        }
    },
    "ramdisk-id": "ari-00000003",
    "reservation-id": "r-7lfps8wj",
    "security-groups": [
        "default"
    ]
}

User data

The openstack/2012-08-10/user_data, openstack/latest/user_data, ec2/2009-04-04/user-data, and ec2/latest/user-data file are present only if the --user-data flag and the contents of the user data file are passed to the openstack server create command.

Configuration drive format

The default format of the configuration drive as an ISO 9660 file system. To explicitly specify the ISO 9660 format, add the following line to the /etc/nova/nova.conf file:

config_drive_format=iso9660

By default, you cannot attach the configuration drive image as a CD drive instead of as a disk drive. To attach a CD drive, add the following line to the /etc/nova/nova.conf file:

config_drive_cdrom=true

For legacy reasons, you can configure the configuration drive to use VFAT format instead of ISO 9660. It is unlikely that you would require VFAT format because ISO 9660 is widely supported across operating systems. However, to use the VFAT format, add the following line to the /etc/nova/nova.conf file:

config_drive_format=vfat

If you choose VFAT, the configuration drive is 64 MB.

Note

In current version (Liberty) of OpenStack Compute, live migration with config_drive on local disk is forbidden due to the bug in libvirt of copying a read-only disk. However, if we use VFAT as the format of config_drive, the function of live migration works well.

Creative Commons Attribution 3.0 License

Except where otherwise noted, this document is licensed under Creative Commons Attribution 3.0 License. See all OpenStack Legal Documents.