New Question
0

Error: Failed to launch instance "test": Please try again later [Error: No valid host was found. ].

asked 2015-01-13 11:32:33 +0300

Niladri gravatar image

updated 2015-01-13 14:42:16 +0300

Hi

I have two physical machines, One with Devstack deployment and another with Hyper-v. I have deployed Nova using Cloudbase: http://www.cloudbase.it/openstack/ope... When I try to spawn an instance named "test" I am getting the error :

Error: Failed to launch instance "test": Please try again later [Error: No valid host was found. ].

I have followed the following steps to create the instance (I would be better if I could explain it with screenshots, but anyway)

1) First I downloaded a mini ISO image of ubuntu.

2) Uploaded the image in openstack as VHD format. It uploaded successfully.

3) Next I clicked on the "Launch" button and it began to create an instance "test" (Flavor selected "m1.tiny")

4) Next on Instance page the task appeared as "Scheduling" and a success message appeared as "Success: Launched instance named "test" ".

5) Next, the task shifted to "Spawing", and after that, it ended in an error, flashing the above message, and the Status is set as "Error".

the log file in /var/log/nova/nova-compute.log contained the following error log:

2015-01-13 02:59:58.758 4106 AUDIT nova.compute.resource_tracker [-] Auditing locally available compute resources

2015-01-13 02:59:59.003 4106 AUDIT nova.compute.resource_tracker [-] Total physical ram (MB): 1841, total allocated virtual ram (MB): 512

2015-01-13 02:59:59.003 4106 AUDIT nova.compute.resource_tracker [-] Free disk (GB): 26

2015-01-13 02:59:59.004 4106 AUDIT nova.compute.resource_tracker [-] Total usable vcpus: 1, total allocated vcpus: 0

2015-01-13 02:59:59.004 4106 AUDIT nova.compute.resource_tracker [-] PCI stats: []

2015-01-13 02:59:59.004 4106 INFO nova.compute.resourcetracker [-] Computeservice record updated for localhost.localdomain:localhost.localdomain

2015-01-13 03:00:15.595 4106 AUDIT nova.compute.manager [req-e458f80c-f53b-40b6-88ee-1ecfa8be66c8 None] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] Starting instance...

2015-01-13 03:00:15.789 4106 WARNING nova.compute.resource_tracker [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] Host field should not be set on the instance until resources have been claimed.

2015-01-13 03:00:15.790 4106 WARNING nova.compute.resource_tracker [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] Node field should not be set on the instance until resources have been claimed.

2015-01-13 03:00:15.799 4106 AUDIT nova.compute.claims [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] Attempting claim: memory 512 MB, disk 1 GB

2015-01-13 03:00:15.800 4106 AUDIT nova.compute.claims [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] Total memory: 1841 MB, used: 512.00 MB

2015-01-13 03:00:15.800 4106 AUDIT nova.compute.claims [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] memory limit: 2761.50 MB, free: 2249.50 MB

2015-01-13 03:00:15.800 4106 AUDIT nova.compute.claims [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] Total disk: 26 GB, used: 0.00 GB

2015-01-13 03:00:15.800 4106 AUDIT nova.compute.claims [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] disk limit not specified, defaulting to unlimited

2015-01-13 03:00:15.871 4106 AUDIT nova.compute.claims [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] Claim successful

2015-01-13 03:00:16.371 4106 INFO nova.scheduler.client.report [-] Compute_service record updated for ('localhost.localdomain', 'localhost.localdomain')

2015-01-13 03:00:17.062 4106 INFO nova.virt.libvirt.driver [-] [instance: 5d2580c2-5f29-4a46-a7b3-d757ffd48de7] Creating image

2015-01-13 03 ... (more)

edit retag flag offensive close merge delete

1 answer

Sort by » oldest newest most voted
0

answered 2015-01-13 19:19:46 +0300

ader1990 gravatar image

Hello,

First of all, m1.tiny is too small for a Windows vm. m1.small or higher(at least 32GB) is preferable.

Secondly, when you create the Windows image you can specify a property: --property hypervisor_type=hyperv, in order for the nova scheduler to choose the correct compute node. More info on creating the image you can find here: http://www.cloudbase.it/filtering-gla...

For more debugging, you can take a look at: nova service-list, to check if all the nodes are up.

Hope it helps.

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

[hide preview]

Question Tools

1 follower

Stats

Asked: 2015-01-13 11:32:33 +0300

Seen: 956 times

Last updated: Jan 13 '15