Ask Your Question

Block Device Mapping is Invalid: failed to get image

asked 2018-11-03 09:11:42 -0600

Hobnob gravatar image

updated 2018-11-03 14:35:54 -0600

My network setup is as follows: Controller node @, (ip tables NAT set up on controller to give management network internet) Compute node @ (connected to the provider network, but only access for vm's)

Followed the default install guide, creating a tier 2 network, set up cinder as follows:

  • Controller node running the standard controller setup, as per default instructions on installation guide
  • Compute node running a cinder node setup with both LVM and NFS, Using the command line I can create images just fine.

When I use the Horizion WebUI to make a basic VM, I get the following error:

Block Device Mapping is Invalid: failed to get image 9d1e61c7-c490-403a-ace1-019c46da2cf4. (HTTP 400) (Request-ID: req-b118e8b2-54ef-4f98-8b44-deb5aa99600a)


Cinder API@controller:

Cinder Scheduler@controller:

Cinder Volume@compute node:

Please let me know if you need anymore information / any more logs, and thankyou for your time.

(Edit, noticed the log pastes had been cut off, I assume there is some kind of max limit to openstack's pastebin, so I've made new pastes that show only the last couple days of log messages)

(Edit 2, more logs)








edit retag flag offensive close merge delete


do you use ceph or lvm as backend?

yasin lachiny gravatar imageyasin lachiny ( 2018-11-03 11:06:32 -0600 )edit

The Volume uses LVM and NFS.

Hobnob gravatar imageHobnob ( 2018-11-03 13:38:52 -0600 )edit

When you have problems with images when launching an instance, first check Nova and Glance logs, not Cinder. Grep for the image ID from the error message.

What are the launch parameters? Do you use ephemeral storage or is your instance root disk a volume?

Bernd Bausch gravatar imageBernd Bausch ( 2018-11-03 13:58:46 -0600 )edit

Regarding the “already mounted” messages, I find a reference to a possible permissions problem on SUSE’s support site: There are other hits when you search for this string.

Bernd Bausch gravatar imageBernd Bausch ( 2018-11-03 14:04:29 -0600 )edit

Not seeing any obvious errors in the glance or nova logs, i'll upload them as well

(too big to fit in comment, edited post with logs)

Hobnob gravatar imageHobnob ( 2018-11-03 14:35:15 -0600 )edit

2 answers

Sort by » oldest newest most voted

answered 2018-11-03 14:12:02 -0600

yasin lachiny gravatar image

create your image again and try again

edit flag offensive delete link more


Remade the image, same error.

Hobnob gravatar imageHobnob ( 2018-11-03 14:19:02 -0600 )edit

openstack image list

does show the uuid of your image?

yasin lachiny gravatar imageyasin lachiny ( 2018-11-04 02:25:31 -0600 )edit

in glance-api.conf what is the backend you define?

yasin lachiny gravatar imageyasin lachiny ( 2018-11-04 02:26:14 -0600 )edit

The glance backend uses NFS, and when i type that command i see both the origional image i was trying to use, and the new one i made as suggested by you.

Hobnob gravatar imageHobnob ( 2018-11-04 02:37:42 -0600 )edit

in /etc/glance/glance-api.conf

[glance_store] stores = file,http default_store = file filesystem_store_datadir = /var/lib/glance/images/

this is my configuration. what is your conf?

yasin lachiny gravatar imageyasin lachiny ( 2018-11-04 03:20:31 -0600 )edit

answered 2018-11-05 06:09:31 -0600

Hobnob gravatar image

I have no idea what the cause of this issue is, I set up openstack myself with no tools to speed up installation, so I assume the issue was some incorrect/misunderstood cinder config. I have "fixed" my issue by starting fresh with installing openstack, and this time using Kolla Ansible to install it for me.

It now all works without issue, so i guess this issue is closed now.

edit flag offensive delete link more

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2018-11-03 06:44:38 -0600

Seen: 1,188 times

Last updated: Nov 05 '18