Software, Physics, Data, Mountains

...and other random associations

Debuggin Eucalyptus Aoe

mostly from http://open.eucalyptus.com/forum/attaching-ebs-image and http://www.debian-administration.org/articles/553


problem1

vtund wasn’t in sync with the pid in /var/run/eucalyptus and so it “wasn’t able to bind to socket” in /var/log/syslog


problem2

Volume is created and available from the SC ps awux | grep vblade should return something… something running on the br0 interface… but they seem to be stuck on eth0


general method for exporting things using aoe

On the storage controller (“target” in AoE lingo)…

Load the module modprobe aoe

Then vbladed 0 1 eth0 /dev/sdd5 or for euca vblade 0 5 br0 /dev/vg-TldY4A../lv-hrh7uQ..

Then on the NC (“initiator”)…

discover aoe-discover (this wasn’t necessary on a NC… just worked once the target had the right device)

status aoe-stat ls /dev/etherd/

this showed a e0.5 entry that can be treated like a regular partition

Note: iSCSI is over TCP (layer 4?), AoE is over Ethernet (layer 2) so iscsi can be routed, aoe can’t (at least not without ebtables and l2 switches) aoe is designed to be used in a tighter lan environment


Had to manually change the files in /var/lib/eucalyptus/db/ to use br0 instead of eth0

This is probably the biggest problem with setting up eucalyptus!!


vtun?

dunno wtf… I assume it’s something going on with tunneling traffic between SC and NC instances?

Comments

If you have any questions or feedback, please feel free to share it with me on Twitter: @m_3