On 08/02/2012 04:30 PM, Stephen Liu wrote: > > > - snip - > >> Ok, requests to 10.0.0.2 are going somewhere, you may need to fire up >> traceroute to see where the requests are going, 10.0.0.2 might clash >> with the IP range on your local network or across a VPN. If that the >> case then you should use another range in the call to > > $ traceroute 10.0.0.2 > traceroute to 10.0.0.2 (10.0.0.2), 30 hops max, 60 byte packets > 1 * * * > 2 * * * > 3 * * * > 4 * * * > 5 * * * > 6 * * * > 7 * * * > 8 * * * > 9 * * * > 10 * * * > 11 * * * > 12 * * * > 13 * * * > 14 * * * > 15 * * * > 16 * * * > ....... All of the packets are timing out, do you still get a ssh response to this ipaddress? > > >> "Failed to schedule_run_instance: No valid host was found." >> >> this usually means the scheduler couldn't find a suitable compute host >> on which to run a VM, an example of why this could happen is that your >> compute node doesn't have enough memory to run a VM on. Can you delete >> the VM in the error state and try nova boot with --flavor 1 which will >> request less memory. > > Pls explain in more detail on "which VM to be deleted"? All other VMs were not running when I tested Fedora 17 Sorry, what I meant was, can you delete the openstack instance that that failed to start. Run > nova list > nova delete <uuid of instance in Error state> then start a new instance with flavor = 1 > >>> nova delete <uuid> >>> nova boot .... >> >> If memory isn't the problem it may also have something todo with the >> error logged in cert.log above > > memory assigned to Fedora 17 is 1536M. I can assign more memory to it. I have 8G memory onboard. > > $ tail /var/log/nova/cert.log my previous email mentioned checking permission of /var/lib/nova/CA to make sure everything is owned by nova, can you check that also please and if it isn't change the owner and restart nova-cert thanks, Derek. _______________________________________________ cloud mailing list cloud@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/cloud
