Using ATA Over Ethernet (AoE) On Ubuntu 12.10 (Initiator And Target)

This guide explains how you can set up an AoE target and an AoE initiator (client), both running Ubuntu 12.10. AoE stands for "ATA over Ethernet" and is a storage area network (SAN) protocol which allows AoE initiators to use storage devices on the (remote) AoE target using normal ethernet cabling. "Remote" in this case means "inside the same LAN" because AoE is not routable outside a LAN (this is a major difference compared to iSCSI). To the AoE initiator, the remote storage looks like a normal, locally-attached hard drive.

I do not issue any guarantee that this will work for you!

 

1 Preliminary Note

I'm using two Ubuntu 12.10 servers here:

  • server1.example.com (Initiator): IP address 192.168.0.100
  • server2.example.com (Target): IP address 192.168.0.101

 

2 Loading The aoe Kernel Module On Both Systems

server1/server2:

Before we start, we must make sure that the the kernel supports AoE:

grep ATA_OVER /boot/config-`uname -r`

This should display something like this:

root@server1:~# grep ATA_OVER /boot/config-`uname -r`
CONFIG_ATA_OVER_ETH=m
root@server1:~#

This means that AoE was built as a kernel module. Let's check if the module is already loaded:

lsmod | grep aoe

If you get nothing back, this means it's not loaded. In this case we can load it as follows:

modprobe aoe

Let's check again if the module is loaded:

lsmod | grep aoe
root@server1:~# lsmod | grep aoe
aoe                    26960  0
root@server1:~#

To have the module loaded automatically when the system boots, we add the aoe module to /etc/modules:

vi /etc/modules
# /etc/modules: kernel modules to load at boot time.
#
# This file contains the names of kernel modules that should be loaded
# at boot time, one per line. Lines beginning with "#" are ignored.

loop
lp
rtc
aoe

 

3 Setting Up The Target (server2)

server2:

First we set up the target (server2):

apt-get install vblade

We can use unused logical volumes, image files, hard drives (e.g. /dev/sdb), hard drive partitions (e.g. /dev/sdb1) or RAID devices (e.g. /dev/md0) for the storage. In this example I will create a logical volume of 20GB named storage1 in the volume group vg0:

lvcreate -L20G -n storage1 vg0

(If you want to use an image file, you can create it as follows:

mkdir /storage
dd if=/dev/zero of=/storage/storage1.img bs=1024k count=20000

This creates the image file /storage/storage1.img with a size of 20GB.

)

Now we export our storage device as follows:

vbladed 0 1 eth0 /dev/vg0/storage1

The first number (0) is the shelf number (major), the second (1) the slot number (minor), change these numbers to your liking. Each AoE device is identified by a couple major/minor which must be unique (if you are exporting multiple devices), with major between 0-65535 and minor between 0-255. The eth0 part tells vbladed which ethernet device to use (if you ethernet device is eth1, then use eth1 - you can find out about your ethernet devices by running

ifconfig

).

To start the export automatically whenever you boot the target, open /etc/rc.local...

vi /etc/rc.local

... and add the following line to it (before the exit 0 line):

[...]
vbladed 0 1 eth0 /dev/vg0/storage1
[...]

 

4 Setting Up The Initiator (server1)

server1:

On server1, we install the initiator:

apt-get install aoetools

Now we check what AoE storage devices are available:

aoe-discover

The command

aoe-stat

should now show the storage devices:

root@server1:~# aoe-stat
      e0.1        21.474GB   eth0 up
root@server1:~#

At this point we have a new block device available on the client box named /dev/etherd/e0.1. If we have a look at the /dev tree a new node appears:

ls -la /dev/etherd/
root@server1:~# ls -la /dev/etherd/
total 0
drwxr-xr-x  2 root root     160 2012-10-01 16:24 .
drwxr-xr-x 16 root root    3800 2012-10-01 16:21 ..
c-w--w----  1 root disk 152,  3 2012-10-01 16:21 discover
brw-rw----  1 root disk 152, 16 2012-10-01 16:24 e0.1
cr--r-----  1 root disk 152,  2 2012-10-01 16:21 err
c-w--w----  1 root disk 152,  6 2012-10-01 16:21 flush
c-w--w----  1 root disk 152,  4 2012-10-01 16:21 interfaces
c-w--w----  1 root disk 152,  5 2012-10-01 16:21 revalidate
root@server1:~#

In the output of

fdisk -l

you should now also find the new hard drive:

root@server1:~# fdisk -l

Disk /dev/sda: 32.2 GB, 32212254720 bytes
255 heads, 63 sectors/track, 3916 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00016be9

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1          32      248832   83  Linux
Partition 1 does not end on cylinder boundary.
/dev/sda2              32        3917    31205377    5  Extended
/dev/sda5              32        3917    31205376   8e  Linux LVM

Disk /dev/etherd/e0.1: 21.5 GB, 21474836480 bytes
255 heads, 63 sectors/track, 2610 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

Disk /dev/etherd/e0.1 doesn't contain a valid partition table
root@server1:~#

To use that device, we must format it:

fdisk /dev/etherd/e0.1

root@server1:~# fdisk /dev/etherd/e0.1
Device contains neither a valid DOS partition table, nor Sun, SGI or OSF disklabel
Building a new DOS disklabel with disk identifier 0x3093ae28.
Changes will remain in memory only, until you decide to write them.
After that, of course, the previous content won't be recoverable.

Warning: invalid flag 0x0000 of partition table 4 will be corrected by w(rite)

WARNING: DOS-compatible mode is deprecated. It's strongly recommended to
         switch off the mode (command 'c') and change display units to
         sectors (command 'u').

Command (m for help):
 <-- n
Command action
   e   extended
   p   primary partition (1-4)

<-- p
Partition number (1-4): <-- 1
First cylinder (1-2610, default 1): <-- ENTER
Using default value 1
Last cylinder, +cylinders or +size{K,M,G} (1-2610, default 2610):
 <-- ENTER
Using default value 2610

Command (m for help):
 <-- t
Selected partition 1
Hex code (type L to list codes):
 <-- 83

Command (m for help): <-- w
The partition table has been altered!

Calling ioctl() to re-read partition table.
Syncing disks.
root@server1:~#

Afterwards, the output of

fdisk -l

should look as follows:

root@server1:~# fdisk -l

Disk /dev/sda: 32.2 GB, 32212254720 bytes
255 heads, 63 sectors/track, 3916 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00016be9

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1          32      248832   83  Linux
Partition 1 does not end on cylinder boundary.
/dev/sda2              32        3917    31205377    5  Extended
/dev/sda5              32        3917    31205376   8e  Linux LVM

Disk /dev/etherd/e0.1: 21.5 GB, 21474836480 bytes
255 heads, 63 sectors/track, 2610 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x3093ae28

            Device Boot      Start         End      Blocks   Id  System
/dev/etherd/e0.1p1               1        2610    20964793+  83  Linux
root@server1:~#

Now we create a filesystem on /dev/etherd/e0.1p1...

mkfs.ext4 /dev/etherd/e0.1p1

... and mount it for test purposes:

mount /dev/etherd/e0.1p1 /mnt

You should now see the new device in the outputs of...

mount
root@server1:~# mount
/dev/mapper/server1-root on / type ext4 (rw,errors=remount-ro)
proc on /proc type proc (rw,noexec,nosuid,nodev)
none on /sys type sysfs (rw,noexec,nosuid,nodev)
none on /sys/fs/fuse/connections type fusectl (rw)
none on /sys/kernel/debug type debugfs (rw)
none on /sys/kernel/security type securityfs (rw)
none on /dev type devtmpfs (rw,mode=0755)
none on /dev/pts type devpts (rw,noexec,nosuid,gid=5,mode=0620)
none on /dev/shm type tmpfs (rw,nosuid,nodev)
none on /var/run type tmpfs (rw,nosuid,mode=0755)
none on /var/lock type tmpfs (rw,noexec,nosuid,nodev)
none on /lib/init/rw type tmpfs (rw,nosuid,mode=0755)
none on /var/lib/ureadahead/debugfs type debugfs (rw,relatime)
/dev/sda1 on /boot type ext2 (rw)
/dev/etherd/e0.1p1 on /mnt type ext4 (rw)
root@server1:~#

... and

df -h
root@server1:~# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/server1-root
                       18G  838M   16G   5% /
none                  243M  180K  242M   1% /dev
none                  247M     0  247M   0% /dev/shm
none                  247M   32K  247M   1% /var/run
none                  247M     0  247M   0% /var/lock
none                  247M     0  247M   0% /lib/init/rw
none                   18G  838M   16G   5% /var/lib/ureadahead/debugfs
/dev/sda1             228M   17M  199M   8% /boot
/dev/etherd/e0.1p1     20G  172M   19G   1% /mnt
root@server1:~#

You can unmount it like this:

umount /mnt

To have the device mounted automatically at boot time, e.g. in the directory /storage, we create that directory...

mkdir /storage

... and add the following line to /etc/fstab:

vi /etc/fstab
[...]
/dev/etherd/e0.1p1       /storage        ext4    defaults,auto,_netdev 0 0

This alone isn't enough to have the device mounted at boot time because the AoE stuff gets loaded after /etc/fstab is read. Therefore we open /etc/rc.local...

vi /etc/rc.local

... and add the following lines to it (before the exit 0 line):

[...]
aoe-discover
sleep 5
mount -a
[...]

For test purposes, you can now reboot the system:

reboot

After the reboot, the device should be mounted:

mount
root@server1:~# mount
/dev/mapper/server1-root on / type ext4 (rw,errors=remount-ro)
proc on /proc type proc (rw,noexec,nosuid,nodev)
none on /sys type sysfs (rw,noexec,nosuid,nodev)
none on /sys/fs/fuse/connections type fusectl (rw)
none on /sys/kernel/debug type debugfs (rw)
none on /sys/kernel/security type securityfs (rw)
none on /dev type devtmpfs (rw,mode=0755)
none on /dev/pts type devpts (rw,noexec,nosuid,gid=5,mode=0620)
none on /dev/shm type tmpfs (rw,nosuid,nodev)
none on /var/run type tmpfs (rw,nosuid,mode=0755)
none on /var/lock type tmpfs (rw,noexec,nosuid,nodev)
none on /lib/init/rw type tmpfs (rw,nosuid,mode=0755)
none on /var/lib/ureadahead/debugfs type debugfs (rw,relatime)
/dev/sda1 on /boot type ext2 (rw)
/dev/etherd/e0.1p1 on /storage type ext4 (rw,_netdev)
root@server1:~#
df -h
root@server1:~# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/server1-root
                       18G  838M   16G   5% /
none                  243M  180K  242M   1% /dev
none                  247M     0  247M   0% /dev/shm
none                  247M   32K  247M   1% /var/run
none                  247M     0  247M   0% /var/lock
none                  247M     0  247M   0% /lib/init/rw
none                   18G  838M   16G   5% /var/lib/ureadahead/debugfs
/dev/sda1             228M   17M  199M   8% /boot
/dev/etherd/e0.1p1     20G  172M   19G   1% /storage
root@server1:~#

 

Share this page:

1 Comment(s)