Description of problem: Error appears when creating an instance: NoValidHost Note: the scheduler gets empty list of 'weighed_hosts' in filter_scheduler.py, List of hosts shall be returned, ie: [WeighedHost [host: bender.idm.lab.bos.redhat.com, weight: 1.0]] This issue does _NOT_ appear when using rabbitmq as a messaging service. This issue does _NOT_ appear when using python-qpid-0.24-1. Note: no errors seen in qpidd log Version-Release number of selected component (if applicable): python-oslo-messaging-1.3.0.2-2 python-qpid-0.18-12 openstack-packstack-2014.1.1-0.16.dev1100 How reproducible: 100% Steps to Reproduce: 1. yum -y install http://team.virt.bos.redhat.com/repos/rhos-release/rhos-release-latest.noarch.rpm 2. rhos-release 5 3. workaround 1103695 (change nagios-plugins-ping -> monitoring-plugins-ping) 4. packstack --allinone --amqp-backend=qpid 5. login to rhos5 Dashboard 6. Project -> Instances -> Launch instance 7. fill the requested fields and press 'Launch' 8. Error by creating the instance: No valid host was found File "/usr/lib/python2.7/site-packages/nova/scheduler/filter_scheduler.py", line 116, in schedule_run_instance raise exception.NoValidHost(reason="") Actual results: Unable to create rhos5 instance Unable to get list of hosts. Expected results: Propper list of hosts is returned. rhos5 instance may be successfully created. Additional info:
Note: this issue most probably causes also 'nova x509-create-cert' timeout. # . keystonerc_admin # nova x509-create-cert /root/accrc/admin/admin-pk.pem /root/accrc/admin/admin-cert.pem ERROR: The server has either erred or is incapable of performing the requested operation. (HTTP 500) (Request-ID: req-cfe83282-61ae-432a-9b19-cb07c62310a0) or (ERROR: HTTPConnectionPool(host='<host_ip>', port=8774): Request timed out. (timeout=600.0)) - seen in CI jobs. This shall be retested as well while verifying this bug. Expected output: # nova x509-create-cert /root/accrc/admin/admin-pk.pem /root/accrc/admin/admin-cert.pem Wrote private key to /root/accrc/admin/admin-pk.pem Wrote x509 certificate to /root/accrc/admin/admin-cert.pem
*** Bug 1106374 has been marked as a duplicate of this bug. ***
In addition to the oslo.messaging library, this fix needs to make it into the stable/icehouse branch of all projects that use impl_qpid from oslo-incubator. The projects affected are: Neutron, Heat, Ceilometer, Trove. I'll be cloning this bug to those projects to make sure the fix gets there if it hasn't already.
This is the stable/icehouse patch for oslo.messaging that needs to be backported for this bug: https://review.openstack.org/98882
Note: This issue is fixed. RHEL-OSP5 instances may be created without any issues. packages under test: python-oslo-messaging-1.3.0.2-4.el7ost openstack-neutron-2014.1-30.el7ost openstack-heat-2014.1-1.5.el7ost openstack-ceilometer-2014.1-2.3.el7ost openstack-puppet-modules-2014.1-14.1.el7ost openstack-packstack-puppet-2014.1.1-0.22.dev1117.el7ost openstack-packstack-2014.1.1-0.22.dev1117.el7ost qpid-cpp-server-0.18-24 python-qpid-0.18-12.el7
Issue fixed, packages available in latest puddle (2014-06-19.2). Comment 7 verified as well. packages under test: python-oslo-messaging-1.3.0.2-4.el7ost -> VERIFIED
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHEA-2014-0845.html