Hide Forgot
Description of problem: When deploying: 2968:2016-10-09 17:41:29.769 1532 DEBUG nova.scheduler.filters.ram_filter [req-9582bca7-ba7a-4099-98a0-25e8303f8f45 d5d5a01d3a9a4779b8 7cd3da53830d08 72510ce285f7480693ef1bd9fd46ae80 - - -] (undercloud1.hackfest.apac, 5c9d4adc-0356-4815-ac2d-07dd5254c337) ram: 0MB disk : 0MB io_ops: 0 instances: 0 does not have 4096 MB usable ram before overcommit, it only has 0 MB. host_passes /usr/lib/python2.7/site -packages/nova/scheduler/filters/ram_filter.py:45 During introspection the correct value or RAM/disk is reported to ironic but after deployment the node goes to no valid host found cause nova-scheduler thinks it 0 disk and 0 ram. Version-Release number of selected component (if applicable): Latest puddle (2016-10-07.4). How reproducible: Every time 2016-10-06.1 puddle is not affected.
Hi Alexandre, It may take a while to the values to be propagated from Ironic to Nova or if the node in Ironic is in maintenance Nova will set it all to 0 so that it won't be able to schedule an instance to that node. Can you please provide the output of the following commands: $ (for each node) ironic node-show <node name/UUID> $ nova hypervisor-stats Also the logs from nova-compute, nova-scheduler and ironic-conductor services ? Thanks
Closing this as requested info not received for 9 months.