Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

Compute nodes not report the proper VMs

Hi Everyone,

Few days ago my OpenStack cluster was restarted. Since then the compute nodes are not reporting the proper instance list, they have. So I have 20+ VMs distributed on 5 compute nodes. If i check the Instance list on the dashboard, I can see them properly, however if I go to Admin -> System -> Hypervisors, the Hypervisor summary says there is only 1 VM in the cluster.

I checked all the nova logs on the controller and the compute nodes too, but they don't report any error.

The {1} nova-compute log snippet represents that it only reports 1 VM. However if I check the /var/lib/nova/instances I can clearly see the other VMs running in the compute node.

What could be the problem? Thanks for your help.

{1} 2018-02-28 22:17:20.021 3856 INFO nova.compute.resource_tracker [req-b54c0c1d-1f1d-44a9-a278-6d30139e2976 - - - - -] Auditing locally available compute resources for node decent-dane 2018-02-28 22:17:20.466 3856 INFO nova.compute.resource_tracker [req-b54c0c1d-1f1d-44a9-a278-6d30139e2976 - - - - -] Total usable vcpus: 12, total allocated vcpus: 2 2018-02-28 22:17:20.466 3856 INFO nova.compute.resource_tracker [req-b54c0c1d-1f1d-44a9-a278-6d30139e2976 - - - - -] Final resource view: name=decent-dane phys_ram=64404MB used_ram=16896MB phys_disk=1039GB used_disk=40GB total_vcpus=12 used_vcpus=2 pci_stats=[] 2018-02-28 22:17:20.509 3856 INFO nova.compute.resource_tracker [req-b54c0c1d-1f1d-44a9-a278-6d30139e2976 - - - - -] Compute_service record updated for decent-dane:decent-dane