Ask Your Question
1

cinder volume not being seen by scheduler

asked 2015-03-11 22:49:50 -0500

Liam Haworth gravatar image

Afternoon Ask Openstack,

I have a [kinda] HA setup on my openstack cluster, 2 controller nodes running all the APIs, schedulers and other "controller" services and 1 compute node running nova-compute, neutron-ovs and cinder-volume.

Everything is working absolutely fine apart from volumes, they can't be made. cinder-volume and other services are all only and running perfectly fine as seen bellow

+------------------+--------------------------+-----------+----------+-------+----------------------------+  
|      Binary      |           Host           |    Zone   |  Status  | State |         Updated_at         |  
+------------------+--------------------------+-----------+----------+-------+----------------------------+  
| cinder-scheduler | os-ctrl1.bluereef.com.au |    nova   | enabled  |   up  | 2015-03-12T03:43:02.000000 |  
| cinder-scheduler |         os-ctrl2         |    nova   | enabled  |   up  | 2015-03-12T03:43:00.000000 |   
|  cinder-volume   |         vmhost3          | OfficeSAN | enabled  |   up  | 2015-03-12T03:43:01.000000 |  
+------------------+--------------------------+-----------+----------+-------+----------------------------+

But whenever someone tries making a volume cinder-scheduler throws the following error

2015-03-12 13:52:50.527 28573 ERROR cinder.scheduler.flows.create_volume [...] Failed to schedule_create_volume: No valid host was found.

It doesn't even say that a host was down or dissabled, just straight out no hosts found.

I have made sure that the time and date is synced across all nodes, I have stopped rabbitmq and cleaned all the queues out and started it again, I am at a dead end right now.

If anyone knows what is up, please let me know A.S.A.P.

Thanks,

Liam.

edit retag flag offensive close merge delete

3 answers

Sort by ยป oldest newest most voted
0

answered 2015-03-12 00:50:37 -0500

It looks like cinder-volume is in a different Zone than your scheduler. I not 100% sure but if I remember correctly you need to have a volume service and a scheduler in the same zone. If you have 2 zones then you will need one of each in each zone.

edit flag offensive delete link more
0

answered 2015-03-11 23:58:35 -0500

Vinoth gravatar image

updated 2015-03-12 00:49:44 -0500

Try restarting the cinder service, especially the tgt service...

service cinder-volume restart

service tgt restart

service cinder-scheduler restart

service cinder-api restart

This is my cinder.conf file for your reference:

[DEFAULT]

control_exchange = cinder

notification_driver = cinder.openstack.common.notifier.rpc_notifier

rootwrap_config = /etc/cinder/rootwrap.conf

api_paste_confg = /etc/cinder/api-paste.ini

iscsi_helper = tgtadm

volume_name_template = volume-%s

volume_group = cinder-volumes

verbose = True

auth_strategy = keystone

state_path = /var/lib/cinder

lock_path = /var/lock/cinder

volumes_dir = /var/lib/cinder/volumes

rpc_backend = cinder.openstack.common.rpc.impl_kombu

rabbit_host = controller

rabbit_port = 5672

rabbit_userid = guest

rabbit_password = RABBIT_PASS

glance_host = controller

[database]

connection = mysql://cinder:CINDER_DBPASS@controller/cinder

[keystone_authtoken]

auth_uri = http://controller:5000

auth_host = controller

auth_port = 35357

auth_protocol = http

admin_tenant_name = service

admin_user = cinder

admin_password = CINDER_PASS

edit flag offensive delete link more

Comments

Nope, still got the same error from scheduler

Liam Haworth gravatar imageLiam Haworth ( 2015-03-12 00:13:54 -0500 )edit

hi

i ahve done everything and i get this error "Unable to update stats, NetAppCmodeNfsDriver -1.0.0 (config name netapp-driver) driver is uninitialized."

does anyone have idea.

bR Rushi

rushi gravatar imagerushi ( 2016-05-16 14:55:01 -0500 )edit
0

answered 2015-03-12 16:55:13 -0500

Liam Haworth gravatar image

updated 2015-03-12 17:11:34 -0500

Odd, I have had cinder-volume in it's own zone for a while now and never encountered a problem with it, I shall see if this is the problem.

EDIT: Turns out this was the problem, I don't know why it took so long to show its self but it is now up and going again, thanks for the help everyone!

#Solution:

The problem ended up being that cinder-volume was configured to be in a different availability zone than cinder-api and cinder-scheduler, I changed cinder-volume back to being in nova and it started working again

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: 2015-03-11 22:49:50 -0500

Seen: 1,126 times

Last updated: Mar 12 '15