[PKG-Openstack-devel] Using sda1 instead of vda1

[Alexander Birkner]

Hello,

we are using a OpenStack Cinder setup with a HPE 3PAR storage which supports fstrim via ISCSI.
By default OpenStack creates vda1 device for each instance. But in our case we’ve added the following flags to all Glance images:

glance image-update --property hw_scsi_model=virtio-scsi --property hw_disk_bus=scsi <image id>

Now OpenStack creates a /dev/sda1 device instead of /dev/vda1 to support fstrim.
But it seems the latest Debian 8 OpenStack image has hardcoded /dev/vda1 inside the booloader and the /etc/fstab file.

After booting the following error appears:

Loading, please wait...
Gave up waiting for root device.  Common problems:
 - Boot args (cat /proc/cmdline)
   - Check rootdelay= (did the system wait long enough?)
   - Check root= (did the system wait for the right device?)
 - Missing modules (cat /proc/modules; ls /dev)
ALERT!  /dev/vda1 does not exist.  Dropping to a shell!

The instance can only boot if the device is called vda1 but then no fstrim will work.

Regards,
Alexander Birkner