Cannot start instance after reboot on RDO OpenStack [closed]

asked 2015-05-10 14:49:18 -0500

OmegaHarvest gravatar image

Just setup RDO OpenStack all in one setup on a single workstation (CentOS 7). Everything works fine except when I reboot and try to launch an existing instance.

Prior to shutting down I'll shut down my test CirrOS instance from the GUI and then shutdown the host via an acpi shutdown. When the host is back up, I go to start the shutoff instance in the UI but it fails after a few seconds to start. I notice the following of interest in the nova-compute log:

ERROR oslo.messaging.rpc.dispatcher [req-3e94838d-e627-4278-8cc1-2f69435813c8 ] Exception during message handling: Unexpected error while running command. Command: sudo nova-rootwrap /etc/nova/rootwrap.conf iscsiadm -m node -T iqn.2010-10.org.openstack:volume-4b7ccef7-18fe-4980-a36a-6f4ed0227611 -p 172.16.1.5:3260 --rescan 
Exit code: 21 
Stdout: u''
Stderr: u'iscsiadm: No session found.\n'

I wondered if there was a correct procedure for shutting down my particular config?

I created the instance with a 10 gig volume which I see before and after rebooting if I run 'lvdisplay' on the host. The host has 1 single SSD formatted using LVM and ext4 if this matters?

Thanks for any help you can provide.

OH

edit retag flag offensive reopen merge delete

Closed for the following reason question is not relevant or outdated by rbowen
close date 2016-06-21 13:43:37.334761

Comments

If your are using lvm , run losetup -a or vgs are root and see does cinder-volumes group show up ?

dbaxps gravatar imagedbaxps ( 2015-05-10 23:34:46 -0500 )edit

If you enabled real LVMiSCSI backend for cinder , then service target is supposed to be enabled to support reboots

dbaxps gravatar imagedbaxps ( 2015-05-10 23:40:37 -0500 )edit

I ran losetup -a commands and got the following:

/dev/loop0: [64768]:10223620 (/srv/loopback-device/swiftloopback)
/dev/loop1: [64768]:8389772 (/var/lib/cinder/cinder-volumes)

All seems ok to me?

OmegaHarvest gravatar imageOmegaHarvest ( 2015-05-11 13:40:13 -0500 )edit

I didnt enabled LVMiSCSI explicitly, although maydbe the allinone install of packstack does this by default? I do have the option to set volume types as iscsi. Should I be using this? Seems odd as its all on one host though.

OmegaHarvest gravatar imageOmegaHarvest ( 2015-05-11 13:41:20 -0500 )edit

I also have no free space on the only hard disk for this system? Should I have left some for cinder to use for LVM?

OmegaHarvest gravatar imageOmegaHarvest ( 2015-05-11 15:30:22 -0500 )edit