When I run "nova --debug resize instanceid flavorid", nova scheduler will raise NoValidHost exception If there is no enough resource available on any computer node for the flavor requested. Then nova conductor will catch the exception and report Warning messages as follows:
WARNING nova.scheduler.utils [req-c0d5f130-c5a9-41b7-8fe4-4d08be4cc774 9ed1534f040c43e98293f6bc6b632e96 bd5848810607480d968b6d1ca9a36637] Failed to compute_task_migrate_server: No valid host was found.
However, nova api does not catch exception but response 202. I think this is bug that nova resize failed but return 202 which is Normal Response Code.
By the way, one question is why "CoreFilter" isn't default in "scheduler_default_filters"? That cause scheduler do not check the vcpu resource when run nova resize.
cfg.ListOpt('scheduler_default_filters',
default=[
'RetryFilter',
'AvailabilityZoneFilter',
'RamFilter',
'ComputeFilter',
'ComputeCapabilitiesFilter',
'ImagePropertiesFilter'
],
Fix proposed to branch: master /review. openstack. org/90970
Review: https:/