Ask Your Question
0

cinder-volume node questions

asked 2013-03-08 02:32:52 -0500

jmurphy-s gravatar image

I am setting up my first cluster and i am having trouble figuring out how to set up cinder with the api and scheduler on the management node and cinder-volume on its own node. I got it all to work in a "all in one setup" but i cant seem to figure out how to make the api communicate with the remote cinder-volume. I am trying to centralize all of the api's on the management node. Right now i have the management node with 2 nics, one public and one for management. The cinder node with 2 nics, one for management and one for storage network. the glance node basically just a nfs server to the glance api on the management node (better ways?) the quantum node with 3 nics, one public one management and the compute node. Any tips for this kind of setup?

Thanks for any help

edit retag flag offensive close merge delete

8 answers

Sort by ยป oldest newest most voted
0

answered 2013-03-19 00:14:54 -0500

jmurphy-s gravatar image

The vg is persistant. It was made at installation.

edit flag offensive delete link more
0

answered 2013-03-09 07:30:15 -0500

thingee gravatar image

The API server is ultimately going to communicate the request through the message queue. The scheduler is going to then hand that off to a host. I would double check your cinder.conf to make sure things are all configured right. It's hard to say what the problem could be without seeing your configuration file and the error you're getting.

edit flag offensive delete link more
0

answered 2013-03-18 17:26:35 -0500

The cinder.conf needs to be set up to point to the corresponding controller's interface for communication.

service_host=%keystone API interface IP address% auth_host=%keystone management interface IP address%

also, for the mysql and rabbit connections have to be configured with your management IPs and respective authentications. please correct me if I'm wrong.

edit flag offensive delete link more
0

answered 2013-03-18 17:35:36 -0500

my mistake, service_host and auth_host are found in the "/etc/cinder/api-paste.ini" not cinder.conf

edit flag offensive delete link more
0

answered 2013-03-18 18:11:44 -0500

jmurphy-s gravatar image

Thanks everyone. I figured it out but now I have another question, of course. I made a volume group which is an entire raid volume and when I reboot cinder volume fails with critical got unknown keyword args to utils.execute, when it tries to export the existing volumes. The vg is persistant. Cinder volume

edit flag offensive delete link more
0

answered 2013-03-18 23:32:56 -0500

yes it will lose it after a reboot. to make the volume group to stay persistant, follow the instruction on the link:

https://github.com/mseknibilel/OpenStack-Folsom-Install-guide/blob/master/Tricks%26Ideas/load_volume_group_after_system_reboot.rst (https://github.com/mseknibilel/OpenSt...)

I used method 1 from the link, for example: I inserted this line - "losetup /dev/loop2 cinder-volumes" #as cinder-volumes is the name of my volume group

edit flag offensive delete link more
0

answered 2013-03-19 15:01:07 -0500

jmurphy-s gravatar image

sorry i was half asleep on that last comment. I created the volume group on /dev/sdb1 when i installed the os so the lv's and vg show up uppon reboot. Being setup this way do i still need the losetup?

root@cinder1:/var/log/cinder# vgdisplay --- Volume group --- VG Name cinder-volumes System ID Format lvm2 Metadata Areas 1 Metadata Sequence No 84 VG Access read/write VG Status resizable MAX LV 0 Cur LV 1 Open LV 0 Max PV 0 Cur PV 1 Act PV 1 VG Size 4.45 TiB PE Size 4.00 MiB Total PE 1166679 Alloc PE / Size 2560 / 10.00 GiB Free PE / Size 1164119 / 4.44 TiB VG UUID zNuWxh-IamW-EHlb-m8ZU-TK8p-1Yvw-A2RxkU

root@cinder1:/var/log/cinder# lvdisplay --- Logical volume --- LV Name /dev/cinder-volumes/volume-299187cc-e7b8-49a0-a0d1-f0e72adb80ec VG Name cinder-volumes LV UUID fYRctf-2C4q-XRwD-PSEA-g0Fv-KPEK-YmdYSa LV Write Access read/write LV Status available # open 0 LV Size 10.00 GiB Current LE 2560 Segments 1 Allocation inherit Read ahead sectors auto - currently set to 256 Block device 252:0

edit flag offensive delete link more
0

answered 2013-03-20 00:39:02 -0500

jmurphy-s gravatar image

Well i switched to tgt and everything works fine.

Thanks!

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2013-03-08 02:32:52 -0500

Seen: 123 times

Last updated: Mar 20 '13