Neutron fails to schedule network to VM guest

[Nicolae Paladi]

When using VLAN network type, quantum fails to schedule network, with the following warning:

(example from quantum log)

2013-09-17 16:50:58 WARNING [quantum.db.agentschedulers_db] Fail scheduling network {'status': u'ACTIVE', 'subnets': [u'792aa923-a38a-4918-a108-bff0ec11f6f5'], 'name': u'fe_net01', 'provider:physical_network': u'physnet2', 'admin_state_up': True, 'tenant_id': u'8e6aa410c7e9464fa4d32b14f77f9571', 'provider:network_type': u'vlan', 'router:external': False, 'shared': False, 'id': u'409c22e7-d1ca-46d2-ad49-6e26f26d0286', 'provider:segmentation_id': 401L}
2013-09-17 16:51:11 WARNING [quantum.db.agentschedulers_db] Fail scheduling network {'status': u'ACTIVE', 'subnets': [u'792aa923-a38a-4918-a108-bff0ec11f6f5'], 'name': u'fe_net01', 'provider:physical_network': u'physnet2', 'admin_state_up': True, 'tenant_id': u'8e6aa410c7e9464fa4d32b14f77f9571', 'provider:network_type': u'vlan', 'router:external': False, 'shared': False, 'id': u'409c22e7-d1ca-46d2-ad49-6e26f26d0286', 'provider:segmentation_id': 401L}
2013-09-17 16:51:12 WARNING [quantum.db.agentschedulers_db] Fail scheduling network {'status': u'ACTIVE', 'subnets': [u'985000db-abdf-44c6-92f4-dabe786e9dde'], 'name': u'ext_net', 'provider:physical_network': u'physnet1', 'admin_state_up': True, 'tenant_id': u'8e6aa410c7e9464fa4d32b14f77f9571', 'provider:network_type': u'flat', 'router:external': True, 'shared': False, 'id': u'd333af77-aa2e-4a28-8fc1-7e609042833e', 'provider:segmentation_id': None}
2013-09-17 16:52:40 WARNING [quantum.db.agentschedulers_db] Fail scheduling network {'status': u'ACTIVE', 'subnets': [u'792aa923-a38a-4918-a108-bff0ec11f6f5'], 'name': u'fe_net01', 'provider:physical_network': u'physnet2', 'admin_state_up': True, 'tenant_id': u'8e6aa410c7e9464fa4d32b14f77f9571', 'provider:network_type': u'vlan', 'router:external': False, 'shared': False, 'id': u'409c22e7-d1ca-46d2-ad49-6e26f26d0286', 'provider:segmentation_id': 401L}
2013-09-17 16:52:40 WARNING [quantum.db.agentschedulers_db] Fail scheduling network {'status': u'ACTIVE', 'subnets': [u'985000db-abdf-44c6-92f4-dabe786e9dde'], 'name': u'ext_net', 'provider:physical_network': u'physnet1', 'admin_state_up': True, 'tenant_id': u'8e6aa410c7e9464fa4d32b14f77f9571', 'provider:network_type': u'flat', 'router:external': True, 'shared': False, 'id': u'd333af77-aa2e-4a28-8fc1-7e609042833e', 'provider:segmentation_id': None}
2013-09-17 16:59:01 WARNING [quantum.db.agentschedulers_db] Fail scheduling network {'status': u'ACTIVE', 'subnets': [u'792aa923-a38a-4918-a108-bff0ec11f6f5'], 'name': u'fe_net01', 'provider:physical_network': u'physnet2', 'admin_state_up': True, 'tenant_id': u'8e6aa410c7e9464fa4d32b14f77f9571', 'provider:network_type': u'vlan', 'router:external': False, 'shared': False, 'id': u'409c22e7-d1ca-46d2-ad49-6e26f26d0286', 'provider:segmentation_id': 401L}
2013-09-17 16:59:01 WARNING [quantum.db.agentschedulers_db] Fail scheduling network {'status': u'ACTIVE', 'subnets': [u'985000db-abdf-44c6-92f4-dabe786e9dde'], 'name': u'ext_net', 'provider:physical_network': u'physnet1', 'admin_state_up': True, 'tenant_id': u'8e6aa410c7e9464fa4d32b14f77f9571', 'provider:network_type': u'flat', 'router:external': True, 'shared': False, 'id': u'd333af77-aa2e-4a28-8fc1-7e609042833e', 'provider:segmentation_id': None}

similar problem in multiple email threads:
http://lists.openstack.org/pipermail/openstack/2013-August/000402.html

http://openstack.redhat.com/forum/discussion/310/fail-scheduling-network-in-quantum-with-vlan/p1

http://openstack.redhat.com/forum/discussion/230/warning-quantum-db-agentschedulers_db-fail-scheduling-network/p1

[ZhiQiang Fan (aji-zqfan)]

see: https://bugs.launchpad.net/neutron/+bug/1213394
review(abandoned, will be restored soon): https://review.openstack.org/#/c/42457/

or if you have better solution, please commit it and add me to the reviewer list

[Akihiro Motoki (amotoki)]

@bug reporter,
On the requested network, is enable_dhcp enabled?

If enable_dhcp is True, it is apparently not related to bug 1213394 mentioned by ZhiQiang Fan.

There is a bug report dhcp-agent healthcheck sometimes regarded as FAIL due to too short threshold.
In the default configuration, report_interval from agents is 4 and agent_down_time is 5 in Grizzly and Havana release.

During Icehouse development, agent_down_time is increased to 9 (larger than twice of report_interval).
Recently report_interval is increased to 30 seconds from performance perspective and agent_down_time is increated accordingly.

I would like to know more information from the bug reporter.

[Launchpad Janitor (janitor)]

[Expired for neutron because there has been no activity for 60 days.]