Ask Your Question

JICarretero's profile - activity

2015-07-22 23:13:10 -0600 received badge  Famous Question (source)
2015-05-09 21:58:17 -0600 received badge  Notable Question (source)
2015-05-09 21:58:17 -0600 received badge  Popular Question (source)
2015-04-15 04:11:25 -0600 commented question Error synchronizing routers on DVR l3_agent

It seems that adding "neutron-servers" the problem allevaites a bit.

2015-04-13 01:52:36 -0600 asked a question Error synchronizing routers on DVR l3_agent

I've configured an Openstack Juno under Ubuntu 14.04 and I've configured distributed router (DVR) in all compute nodes. So, in every compute node and in the neutron node (l3-sdvr) there seem to be some errors related to RPC tmeouts.

I think this problem is related to some strange behaviours with floating IPs, once I associate a virtual IP with an instance it takes quite long time to respond. It seems to work, but 5, 10 or even 20 minutes later.

"neutron-server" proccess uses about 90% CPU. I have almost 40 routers and over 60 networks.

The error message I can read several times this message:

2015-04-13 08:21:02.335 33360 ERROR neutron.agent.l3_agent [-] Failed synchronizing routers due to RPC error
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent Traceback (most recent call last):
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr/lib/python2.7/dist-packages/neutron/agent/l3_agent.py", line 1896, in _sync_routers_task
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     context, router_ids)
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr/lib/python2.7/dist-packages/neutron/agent/l3_agent.py", line 105, in get_routers
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     router_ids=router_ids))
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr/lib/python2.7/dist-packages/neutron/common/log.py", line 34, in wrapper
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     return method(*args, **kwargs)
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr/lib/python2.7/dist-packages/neutron/comg/_drivers/amqpdriver.py", line 408, in send
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     retry=retry)
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr/lib/python2.7/dist-packages/oslo/messaging/_drivers/amqpdriver.py", line 397, in _send
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     result = self._waiter.wait(msg_id, timeout)
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr/lib/python2.7/dist-packages/oslo/messaging/_drivers/amqpdriver.py", line 285, in wait
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     reply, ending = self._poll_connection(msg_id, timeout)
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr/lib/python2.7/dist-packages/oslo/messaging/_drivers/amqpdriver.py", line 235, in _poll_connection
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     % msg_id)
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent MessagingTimeout: Timed out waiting for a reply to message ID caa0f2dc36c54787b91fb2f3c9df8259mon/rpc.py", line 161, in call
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     context, msg, rpc_method='call', **kwargs)
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr/lib/python2.7/dist-packages/neutron/common/rpc.py", line 187, in __call_rpc_method
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent     return func(context, msg['method'], **msg['args'])
2015-04-13 08:21:02.335 33360 TRACE neutron.agent.l3_agent   File "/usr ...
(more)
2015-03-27 04:54:35 -0600 received badge  Enthusiast
2015-03-19 10:30:30 -0600 answered a question TCP/SSH not working with VM

The problem is the MTU. Inside the VM (maybe you can use the console to log in) you can set the mtu of the instance to 1400. So, to correct this problem in the living instance you have to lower the mtu.

 sudo ip link set eth0  mtu 1400

In your neutron controller, you must configure your dhcp_agent.ini

[DEFAULT]
....
dnsmasq_config_file = /etc/neutron/dnsmasq-neutron.conf

And

echo dhcp-option-force=26,1400" >> /etc/neutron/dnsmasq-neutron.conf

Restart the neutron-server, neutron-dhcp-agent and the instances you create from this moment on will be accesible using ssh. And the existing instances will need a reboot.

2014-12-16 19:31:53 -0600 received badge  Nice Question (source)
2014-04-15 19:19:12 -0600 received badge  Student (source)
2014-03-31 01:57:22 -0600 received badge  Famous Question (source)
2014-03-10 11:47:47 -0600 received badge  Teacher (source)
2014-03-06 01:33:52 -0600 answered a question error when running nova boot-no valid host was found

Make sure you have AMPQ Queues (qpid in your case) running and well configured in your Openstack components and the Centos Firewall opened to accept connectios to QPID Port.

2014-03-06 01:17:56 -0600 answered a question Failed when I try to start openstack-compute

I had that issue on Havana using Ubuntu 12.04 after updating the Compute node to the latests updates but not doing the same on the nova-controller node -- So, I'd try to be sure that the latests versions are running in both nodes -- I mean the latest Havana Release on your compute-node ( Feb 13, 2014 ) seems to break something if you have the previous one in the controller node ( Dec 16, 2013 ).

2014-03-03 01:48:58 -0600 commented question Issue with nova image-list?

Do you have 4 glance services? (These are the IDs from your keystone endpoint-list: 9057b7241d0e4e75b3cc69051ee2007f, 48d46adbb3214c6a8b2d1dfb1efe3c67, cbc9695cec004841b122a389405af96a, af728ad3d517458381235602f7f54b08)

2014-03-01 13:56:45 -0600 received badge  Notable Question (source)
2014-02-28 23:08:21 -0600 received badge  Popular Question (source)
2014-02-28 05:41:01 -0600 answered a question nova hypervisor-show node display qemu

This is explained at KVM FAQ:

http://www.linux-kvm.org/page/FAQ#How_can_I_check_that_I.27m_not_falling_back_to_QEMU_with_no_hardware_acceleration.3F (link text)

2014-02-28 05:09:05 -0600 commented question How do nova-cells really works?

Thank you smafulli, I've elaborated my question.

2014-02-28 05:08:32 -0600 answered a question How do nova-cells really works?

Thank you Smaffulli, I've elaborated my question.

2014-02-28 05:07:04 -0600 received badge  Editor (source)
2014-02-27 13:18:28 -0600 asked a question How do nova-cells really works?

I'm trying to do some testing with nova-cells, however, whenever I try to deploy a new instance It goes to scheduling state and it never goes any further.

In one host (172.30.8.203) I've installed nova-api, nova-cells and nova-conductor: Binary Host Zone Status State Updated_At nova-conductor havana-novaController internal enabled :-) 2014-02-28 10:42:53 nova-cells havana-novaController internal enabled :-) 2014-02-28 10:42:54 nova-cert havana-novaController internal enabled :-) 2014-02-28 10:42:53

In another host (172.30.8.210) I've installed (appart from rabbitmq) nova-conductor havana-compute-1 internal enabled :-) 2014-02-28 10:45:35 nova-cert havana-compute-1 internal enabled :-) 2014-02-28 10:45:30 nova-scheduler havana-compute-1 internal enabled :-) 2014-02-28 10:45:35 nova-compute havana-compute-1 nova enabled :-) 2014-02-28 10:45:35 nova-cells havana-compute-1 internal enabled :-) 2014-02-28 10:45:28

I've defined the cells:

 Id  Name        Type    Username    Hostname         Port   VHost     
---  ----------  ------  ----------  ---------------  -----  ----------
  5  api         parent  guest       172.30.8.205     5672   None      
---  ----------  ------  ----------  ---------------  -----  ----------

And (yes, RabbitMQ is there for API).

 Id  Name        Type    Username    Hostname         Port   VHost     
---  ----------  ------  ----------  ---------------  -----  ----------
  5  api         parent  guest       172.30.8.205     5672   None      
---  ----------  ------  ----------  ---------------  -----  ----------

When I start the nova-cells services, I can see it connects to the other cell rabbitMQ for a moment, but it stays unconnected most of the time. (Is this the way it should work?)

I try to deploy a new instance and the API starNo error-logs are shown. The API queries glance for the image ID, neutron for the network id, everything goes well... But it stops. There's no communication between cells, no messages to AMQP, no error logs, nothing... but the instance goes to "BUILD/scheduling" state and it never moves to another state.

I've been trying to figure out how nova-cells works and understand how this should go forward, but the fact is that I haven't find anything helpful.

I'd like to find something more profound about nova-cells than the basic configurations I've found in order to understand this technology better.

So, any help would be very welcome.

Thank you in advance. José Ignacio.