not bootable disk about addon-lxdone

[dann1]

https://github.com/OpenNebula/addon-lxdone/blob/master/image-handling/build-img.sh is intended for creating a base image from scratch, if you have an existing LXC container read https://github.com/OpenNebula/addon-lxdone/blob/master/Image.md there are instructions for that case. Also mount the image created and list the contents inside it

[niksfirefly]

I tried both ways - none works
First I build lxdone.img with build-img.sh and then on sunstone image creation
uploaded or added via /var/tmp/lxdone.img
Then i created virtual appliance with
https://github.com/OpenNebula/addon-lxdone/blob/master/Image.md
same effect.

No errors on every step from tutorial
https://github.com/OpenNebula/addon-lxdone/blob/master/Setup.md
until dreadful VM not bootable info

I am using latest clone from github
https://github.com/OpenNebula/addon-lxdone

below image from ONE

NAME : lxdone
USER : oneadmin
GROUP : oneadmin
DATASTORE : default
TYPE : OS
REGISTER TIME : 04/03 20:05:25
PERSISTENT : No
SOURCE : /var/lib/one//datastores/1/96955803857eb6e97f71f08b9a6264b1
PATH : /var/tmp/lxdone.img
FSTYPE : raw
SIZE : 600M
STATE : rdy
RUNNING_VMS : 0

PERMISSIONS
OWNER : um-
GROUP : ---
OTHER : ---

IMAGE TEMPLATE
DEV_PREFIX="vd"
DRIVER="raw"

also ONE template

ID : 7
NAME : lxd
USER : oneadmin
GROUP : oneadmin
REGISTER TIME : 04/03 20:32:56

PERMISSIONS
OWNER : um-
GROUP : ---
OTHER : ---

TEMPLATE CONTENTS
CONTEXT=[
NETWORK="YES",
SSH_PUBLIC_KEY="$USER[SSH_PUBLIC_KEY]" ]
CPU="0.1"
DISK=[
IMAGE="lxdone",
IMAGE_UNAME="oneadmin" ]
GRAPHICS=[
LISTEN="0.0.0.0",
TYPE="VNC" ]
MEMORY="1024"
NIC=[
NETWORK="lxdbr0",
NETWORK_UNAME="oneadmin" ]

[dann1]

Let's inspect the contents of the image you created. Try

sudo losetup /dev/loop0  /var/lib/one//datastores/1/96955803857eb6e97f71f08b9a6264b1
mount /dev/loop0 /mnt
ls -lh /mnt

If there were no errors during the execution of build-img.sh I'd expect the image is fine. The below commands are supposed to do that. Also give me the screenshot of the VNC error. I suspect your host is using KVM because that particular output

Booting from Hard Disk
Boot failed: not a bootable disk

I've seen it when using virt-manager

[niksfirefly]

yes of course
I can mount images

drwx------ 2 root root 16K Apr 3 19:39 lost+found
-rw-r--r-- 1 root root 1.2K Apr 3 19:44 metadata.yaml
drwxr-xr-x 21 root root 4.0K Apr 3 19:44 rootfs
drwxr-xr-x 2 root root 4.0K Apr 3 19:44 templates

[dann1]

Try
ls -lh /mnt/rootfs/

[niksfirefly]

/mnt:
total 28K
drwx------ 2 root root 16K Apr 3 19:39 lost+found
-rw-r--r-- 1 root root 1.2K Apr 3 19:44 metadata.yaml
drwxr-xr-x 21 root root 4.0K Apr 3 19:44 rootfs
drwxr-xr-x 2 root root 4.0K Apr 3 19:44 templates

/mnt/rootfs/:
total 76K
drwxr-xr-x 2 root root 4.0K Apr 3 19:44 bin
drwxr-xr-x 2 root root 4.0K Apr 12 2016 boot
drwxr-xr-x 4 root root 4.0K Apr 3 19:41 dev
drwxr-xr-x 63 root root 4.0K Apr 3 19:44 etc
drwxr-xr-x 2 root root 4.0K Apr 12 2016 home
drwxr-xr-x 11 root root 4.0K Apr 3 19:43 lib
drwxr-xr-x 2 root root 4.0K Apr 3 19:40 lib64
drwxr-xr-x 2 root root 4.0K Apr 3 19:39 media
drwxr-xr-x 2 root root 4.0K Apr 3 19:39 mnt
drwxr-xr-x 2 root root 4.0K Apr 3 19:39 opt
drwxr-xr-x 2 root root 4.0K Apr 12 2016 proc
drwx------ 2 root root 4.0K Apr 3 19:49 root
drwxr-xr-x 6 root root 4.0K Apr 3 19:44 run
drwxr-xr-x 2 root root 4.0K Apr 3 19:44 sbin
drwxr-xr-x 2 root root 4.0K Apr 3 19:39 srv
drwxr-xr-x 2 root root 4.0K Feb 5 2016 sys
drwxrwxrwt 2 root root 4.0K Apr 3 19:44 tmp
drwxr-xr-x 10 root root 4.0K Apr 3 19:39 usr
drwxr-xr-x 11 root root 4.0K Apr 3 19:39 var

/mnt/templates/:
total 12K
-rw-r--r-- 1 root root 39 Apr 3 19:44 hostname.tpl
-rw-r--r-- 1 root root 262 Apr 3 19:44 hosts.tpl
-rw-r--r-- 1 root root 7 Apr 3 19:44 upstart-override.tpl

[niksfirefly]

and screenshot

[dann1]

well it seems the image is OK, and as I suspected you are not using LXD in OpenNebula. Give me a screenshot of the host tab you are using. Take this one as a reference.

[dann1]

Read and follow carefully https://github.com/OpenNebula/addon-lxdone/blob/master/Setup.md#42-virtualization-node seems you made some mistake there

[niksfirefly]

Thanks for your suggestions.
Indeed Vms were initialized with default KVM host so I change in Vm template
SCHED_REQUIREMENTS = "ID="4""
where ID is the proper lxd Host definition.
Now Vms are created but without network configuration in it.
I follow exactly
https://github.com/OpenNebula/addon-lxdone/blob/master/Setup.md

I wonder is it normal behaviour and I must create network config manually in every lxd VM?
Why we remove eth0 as is in tutorial ?
lxc profile device remove default eth0

[jmdelafe]

The VMs were initialized in that host because the Opennebula scheduler's decision. You must specify an LXD host (or a cluster) within it's template to avoid such behavior.

We remove eth0 from the default profile because we want to, inside the containers, only interfaces that we attached from it's template. If you skip that step, containers will have an interface that was not declared inside Opennebula. Contextualization, for example, will not work.

[dann1]

if you leave eth0 in the default lxd profile that network interface won't be controlled by OpenNebula, and that's not what we aim.