5

I'm getting the following error on my Openstack (DevStack) every time I try to launch an image other than cirrOS. Walking through internet leads me to:

  1. Openstack cannot allocate RAM, CPU resources. It's not true because I have a lot of RAM, disk space and CPU available.
  2. set in nova.conf -> scheduler_default_filters=AllHostsFilter Tried without success.

This hapends to any image in any format that is other than cirrOS.

Update: Now it is clear that there is no direct answer to this question. Lets hope Openstack guys will provide more specific information in this error message

ZuzEL
  • 12,768
  • 8
  • 47
  • 68

5 Answers5

4

Make sure the flavour size you select is size "small" or larger, cirros uses tiny by default, as do others if not changed

c0nsaw
  • 39
  • 3
  • Yes, and also check the hypervisor resources - in my case I had too much occupied RAM, so I had to clear out some old VMs. – metakermit Apr 24 '14 at 20:59
1

For me, I got this same error because I mistakenly added an ubuntu image and set the metadata "hypervisor" tag to be "KVM" and not "QEMU". My host only had QEMU capability, of course. When I went to launch it, it gave that "No Valid Host was found". I'd say make sure the tags on the image aren't preventing the host from thinking "I can't run this". Simply changing the image tag back to QEMU fixed it for me.

Geese
  • 51
  • 4
1
  1. check the core service is running by typing command " netstat -an | grep LISTENING". In the controller node,it should contains listening port 8778(placement_api service), 8774(compute-service),9292(Image service),9696(network),5000(Identify service),5672(rabbitmq server), 11211( memcache server),35357(Identify service) at least if you don't modify the default config. if you install Ocata by offical guide line by line ,You must start placement-api service manually。
  2. In compute node,you can run command "virt-host-validate" to check your host that whether it supports hardware virtualization.If fails ,edit the file "/etc/nova/nova.conf",set virt_type=qemu.
  3. Ensure your host owns enough cpu,Memory,disk resources.
  4. if All the steps is ok ,Open Debug log message By set debug=true int /etc/nova/nova.conf。you can find more information in the directory /var/log/nova/
CurlDown
  • 11
  • 2
0

I don't know WHY but after a while I can launch Ubuntu

saucy-server-cloudimg-i386-disk1.img — Ubuntu 13.10 x32

but can not

saucy-server-cloudimg-amd64-disk1.img — Ubuntu 13.10 x64

and vise versa, I can launch

precise-server-cloudimg-amd64-disk1.img — Ubuntu 13.04 x64

and cannot

precise-server-cloudimg-i386-disk1.img — Ubuntu 13.04 x32

ZuzEL
  • 12,768
  • 8
  • 47
  • 68
0

The error can be due to many reasons. As you have told that it works with cirros, try this. Run the command "glance index". you will get the images you have in your glance. Now do a "glance show (your-glance-id)" Compare this between Cirros image and the rest.

Chetan
  • 1,217
  • 2
  • 13
  • 27