Hmm, so I did this query:
http://logstash.openstack.org/#eyJzZWFyY2giOiJtZXNzYWdlOlwiRmlsdGVyIENvbXB1dGVGaWx0ZXIgcmV0dXJuZWQgMCBob3N0c1wiIEFORCBmaWxlbmFtZTpcImxvZ3Mvc2NyZWVuLW4tc2NoLnR4dFwiIiwiZmllbGRzIjpbXSwib2Zmc2V0IjowLCJ0aW1lZnJhbWUiOiI2MDQ4MDAiLCJncmFwaG1vZGUiOiJjb3VudCIsInRpbWUiOnsidXNlcl9pbnRlcnZhbCI6MH0sInN0YW1wIjoxMzg2MzM3Mzc0Njk0fQ==
Which also turns this up:
2013-12-05 23:39:35.662 DEBUG nova.scheduler.filters.compute_filter [req-416eeb14-f13d-41a5-90f0-84fe1f589100 ServersAdminTestXML-tempest-546887800-user ServersAdminTestXML-tempest-546887800-tenant] (devstack-precise-check-hpcloud-region-b-801504, devstack-precise-check-hpcloud-region-b-801504) ram:7154 disk:113664 io_ops:0 instances:5 is disabled, reason: None host_passes /opt/stack/new/nova/nova/scheduler/filters/compute_filter.py:43
2013-12-05 23:39:35.662 INFO nova.filters [req-416eeb14-f13d-41a5-90f0-84fe1f589100 ServersAdminTestXML-tempest-546887800-user ServersAdminTestXML-tempest-546887800-tenant] Filter ComputeFilter returned 0 hosts
2013-12-05 23:39:35.731 WARNING nova.scheduler.driver [req-416eeb14-f13d-41a5-90f0-84fe1f589100 ServersAdminTestXML-tempest-546887800-user ServersAdminTestXML-tempest-546887800-tenant] [instance: b44877c0-cd63-4812-b679-9d1594fcf3e4] Setting instance to ERROR state.
The difference is rather than seeing "is disabled or has not been heard from in a while" for host, we see "is disabled, reason: None" - but still results in the failure. That's coming from this log:
http://logs.openstack.org/47/54647/7/check/check-tempest-dsvm-postgres-full/70389b4/logs/screen-n-sch.txt
Hmm, so I did this query:
http:// logstash. openstack. org/#eyJzZWFyY2 giOiJtZXNzYWdlO lwiRmlsdGVyIENv bXB1dGVGaWx0ZXI gcmV0dXJuZWQgMC Bob3N0c1wiIEFOR CBmaWxlbmFtZTpc ImxvZ3Mvc2NyZWV uLW4tc2NoLnR4dF wiIiwiZmllbGRzI jpbXSwib2Zmc2V0 IjowLCJ0aW1lZnJ hbWUiOiI2MDQ4MD AiLCJncmFwaG1vZ GUiOiJjb3VudCIs InRpbWUiOnsidXN lcl9pbnRlcnZhbC I6MH0sInN0YW1wI joxMzg2MzM3Mzc0 Njk0fQ= =
Which also turns this up:
2013-12-05 23:39:35.662 DEBUG nova.scheduler. filters. compute_ filter [req-416eeb14- f13d-41a5- 90f0-84fe1f5891 00 ServersAdminTes tXML-tempest- 546887800- user ServersAdminTes tXML-tempest- 546887800- tenant] (devstack- precise- check-hpcloud- region- b-801504, devstack- precise- check-hpcloud- region- b-801504) ram:7154 disk:113664 io_ops:0 instances:5 is disabled, reason: None host_passes /opt/stack/ new/nova/ nova/scheduler/ filters/ compute_ filter. py:43
2013-12-05 23:39:35.662 INFO nova.filters [req-416eeb14- f13d-41a5- 90f0-84fe1f5891 00 ServersAdminTes tXML-tempest- 546887800- user ServersAdminTes tXML-tempest- 546887800- tenant] Filter ComputeFilter returned 0 hosts
2013-12-05 23:39:35.731 WARNING nova.scheduler. driver [req-416eeb14- f13d-41a5- 90f0-84fe1f5891 00 ServersAdminTes tXML-tempest- 546887800- user ServersAdminTes tXML-tempest- 546887800- tenant] [instance: b44877c0- cd63-4812- b679-9d1594fcf3 e4] Setting instance to ERROR state.
The difference is rather than seeing "is disabled or has not been heard from in a while" for host, we see "is disabled, reason: None" - but still results in the failure. That's coming from this log:
http:// logs.openstack. org/47/ 54647/7/ check/check- tempest- dsvm-postgres- full/70389b4/ logs/screen- n-sch.txt