October 2022 - openstack-discuss - lists.openstack.org

[nova] Proposing EOL of nova project's old stable branches (Stein, Rocky, Queens)
by Elõd Illés 24 Oct '22

24 Oct '22

Hi, At the 2023.1 Antelope PTG, Nova team discussed that currently open Queens, Rocky and Stein branches for Nova teams' repositories will be moved to End of Life. There are multiple reasons behind this decision: - gate of these branches are broken for a couple of months now - zuul job definitions for these branches use the pre-"zuul-v3" [1] style, which makes the maintenance harder of these jobs - CI on Rocky and Queens use Ubuntu Xenial, which is also beyond its public maintenance window date And last but not least bugfix backports to these branches don't really get enough reviews anymore. If anyone has any interest to keep any of the above branches open then this is the last opportunity to step up, otherwise, as Nova team decided on the PTG, these branches will transition to End of Life. Please let us know before November 7th, we will proceed with the transition in case no responses will come till that day. [1] https://docs.openstack.org/project-team-guide/zuulv3.html Thanks in advance, Előd Illés irc:elodilles @ #openstack-nova

1 0

[telemetry][cloudkitty][ceilometer] Billing windows instances
by Thomas Goirand 24 Oct '22

24 Oct '22

Hi there! We're using telemetry+cloudkitty for our rating. We're happy of it, though we're having the issue that we would like to bill instances running Windows. The example given in the Cloudkitty doc shows how to bill more when an instance runs Windows. That works in theory, though in practice, Microsoft has a license model based on how many vCPU the instannce runs. So the model that the Cloudkitty documentation shows simply doesn't work with the SPLA thingy. We've looked at options. One way would of course writing a new custom pollster, but we don't like the idea: this would mean polling for all of the thousands of instances that are running in our deployments. So we would very much prefer having ceilometer running on compute node (with polling_namespaces=compute) to do the work, as this scales a way better. However, Ceilometer polls libvirt, which only has the information about the image ID, not the metadata associated with the image (like the property os_type=windows, for example). So, is there a better way than a dynamic pollster? Can this be done with ceilometer on the compute nodes? Cheers, Thomas Goirand (zigo)

2 3

Openstack cluster cannot create instances when 1 of 3 rabbitmq cluster node down
by Nguyễn Hữu Khôi 24 Oct '22

24 Oct '22

Description =========== I set up 3 controllers and 3 compute nodes. My system cannot work well when 1 rabbit node in cluster rabbitmq is down, cannot launch instances. It stucked at scheduling. Steps to reproduce =========== Openstack nodes point rabbit://node1:5672,node2:5672,node3:5672// * Reboot 1 of 3 rabbitmq node. * Create instances then it stucked at scheduling. Workaround =========== Point to rabbitmq VIP address. But We cannot share the load with this solution. Please give me some suggestions. Thank you very much. I did google and enabled system log's debug but I still cannot understand why. Nguyen Huu Khoi

5 9

Openstack Cinder and Nova Services cannot work when rabbitmq cluter node down
by Nguyễn Hữu Khôi 24 Oct '22

24 Oct '22

Title: Openstack cluster cannot create when 1 of 3 rabbitmq cluster node down Bug description: Description =========== I set up 3 controllers and 3 compute nodes. My system cannot work when 1 rabbit node in cluster rabbitmq is down, cannot create volume or launch instance. It stucked at creating and scheduling respectively. Steps to reproduce =========== Openstack nodes point rabbit://node1:5672,node2:5672,node3:5672// * Reboot 1 of 3 rabbitmq node. * Create volume or launch instance then it stucked at creating and scheduling respectively. Workaround =========== I need reboot cinder and nova services to create volume and launch instance . More Info: I see in cinder_scheduler, it looks like cinder cannot change to another rabbitmq node. I hope we have ideas for that.. 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task Traceback (most recent call last): 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_messaging/_drivers/amqpdriver.py", line 441, in get 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task return self._queues[msg_id].get(block=True, timeout=timeout) 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/eventlet/queue.py", line 322, in get 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task return waiter.wait() 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/eventlet/queue.py", line 141, in wait 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task return get_hub().switch() 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/eventlet/hubs/hub.py", line 313, in switch2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task return self.greenlet.switch() 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task _queue.Empty 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task During handling of the above exception, another exception occurred: 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task Traceback (most recent call last): 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_service/periodic_task.py", line 216, in run_periodic_tasks 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task task(self, context) 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/nova/compute/manager.py", line 9716, in _sync_power_states 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task db_instances = objects.InstanceList.get_by_host(context, self.host, 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_versionedobjects/base.py", line 175, in wrapper 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task result = cls.indirection_api.object_class_action_versions( 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/nova/conductor/rpcapi.py", line 240, in object_class_action_versions 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task return cctxt.call(context, 'object_class_action_versions', 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_messaging/rpc/client.py", line 189, in call 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task result = self.transport._send( 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_messaging/transport.py", line 123, in _send 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task return self._driver.send(target, ctxt, message, 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_messaging/_drivers/amqpdriver.py", line 689, in send 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task return self._send(target, ctxt, message, wait_for_reply, timeout, 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_messaging/_drivers/amqpdriver.py", line 678, in _send 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task result = self._waiter.wait(msg_id, timeout, 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_messaging/_drivers/amqpdriver.py", line 567, in wait 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task message = self.waiters.get(msg_id, timeout=timeout) 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task File "/var/lib/kolla/venv/lib/python3.8/site-packages/oslo_messaging/_drivers/amqpdriver.py", line 443, in get 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task raise oslo_messaging.MessagingTimeout( 2022-10-24 14:23:01.945 7 ERROR oslo_service.periodic_task oslo_messaging.exceptions.MessagingTimeout: Timed out waiting for a reply to message ID c8a676a9709242908dcff97046d7976d Nguyen Huu Khoi

1 0

[large-scale][oslo.messaging] RPC workers and db connection
by Arnaud Morin 24 Oct '22

24 Oct '22

Hey all, TLDR: How can I fine tune the number of DB connection on OpenStack services? Long story, with some inline questions: I am trying to figure out the maximum number of db connection we should allow on our db cluster. For this short speech, I will use neutron RPC as example service, but I think nova is acting similar. So, to do so, I identified few parameters that I can tweak: rpc_workers [1] max_pool_size [2] max_overflow [3] executor_thread_pool_size [4] rpc_worker default is half CPU threads available (result of nproc) max_pool_size default is 5 max_overflow default is 50 executor_thread_pool_size is 64 Now imagine I have a server with 40 cores, So rpc_worker will be 20. Each worker will have a DB pool with 5+50 connections available. Each worker will use up to 64 "green" thread. The theorical max connection that I should set on my database is then: rpc_workers*(max_pool_size+max_overflow) = 20*(5+50) = 1100 Q1: am I right here? I have the feeling that this is huge. Now, let's assume each thread is consuming 1 connection from the DB pool. Under heavy load, I am affraid that the 64 threads could exceed the number of max_pool_size+max_overflow. Also, I noticed that some green threads were consuming more than 1 connection from the pool, so I can reach the max even sooner! Another thing, I notice that I have 21 RPC workers, not 20. Is it normal? [1] https://docs.openstack.org/neutron/latest/configuration/neutron.html#DEFAUL… [2] https://docs.openstack.org/neutron/latest/configuration/neutron.html#databa… [3] https://docs.openstack.org/neutron/latest/configuration/neutron.html#databa… [4] https://docs.openstack.org/neutron/latest/configuration/neutron.html#DEFAUL… Cheers, Arnaud.

3 4

RE: RE: [yoga][cinder] Cinder NFS backend: Compute service cannot access volume file (UID/GID problem)
by Felix Hüttner 24 Oct '22

24 Oct '22

Sorry, no idea about that, for us the group also has write permissions -- Felix Huettner From: 박경원 <park0kyung0won(a)dgist.ac.kr> Sent: Monday, October 24, 2022 9:24 AM To: Felix Hüttner <felix.huettner(a)mail.schwarz>; openstack-discuss(a)lists.openstack.org Subject: RE: RE: [yoga][cinder] Cinder NFS backend: Compute service cannot access volume file (UID/GID problem) Hello Felix Thank you very much for kind reply Do I also need to change permission setting of volume file in /var/lib/nova/mnt/... ? By default its: drwxr-x--- 2 64061 64061 11 Oct 24 04:19 99c4f7e8b15983b65e20cb7d37db899f group has only read and execute permission, no write permission ---------- 원본 메일 ---------- 보낸사람: "Felix Hüttner" <felix.huettner(a)mail.schwarz<mailto:felix.huettner@mail.schwarz>> 받는사람: "park0kyung0won(a)dgist.ac.kr<mailto:park0kyung0won@dgist.ac.kr>" <park0kyung0won(a)dgist.ac.kr<mailto:park0kyung0won@dgist.ac.kr>>, "openstack-discuss(a)lists.openstack.org<mailto:openstack-discuss@lists.openstack.org>" <openstack-discuss(a)lists.openstack.org<mailto:openstack-discuss@lists.openstack.org>> 날짜: 2022-10-24 (월) 16:12:48 제목: RE: [yoga][cinder] Cinder NFS backend: Compute service cannot access volume file (UID/GID problem) Hi, we are solving this issue for us by creating a “cinder” group on all hypervisors with the same gid (64061 in your case). Then we add the nova user to the cinder group and we are fine afterwards. You might need set “dynamic_ownership = 0" In your libvirt qemu.conf -- Felix Huettner From: 박경원 <park0kyung0won(a)dgist.ac.kr<mailto:park0kyung0won@dgist.ac.kr>> Sent: Monday, October 24, 2022 7:21 AM To: openstack-discuss(a)lists.openstack.org<mailto:openstack-discuss@lists.openstack.org> Subject: [yoga][cinder] Cinder NFS backend: Compute service cannot access volume file (UID/GID problem) Hi I'm trying to setup cinder-volume service with NFS backend When I create a new VM instance with a volume from web UI, cinder-volume service on storage node creates volume file just fine But I get the following error on compute node and instance fails to spawn. 2022-10-24 02:14:25.347 402789 ERROR nova.compute.manager [req-47ec9fb1-9daa-4c24-8673-538797a217cc 8769cfaf608349bd9fbb36f92b188fe3 e1e8e8397cde49899b00d09dec76b29e - default default] [instance: 5acb1dc3-0685-4980-977b-b6dfff6dfb45] Instance failed to spawn: libvirt.libvirtError: internal error: process exited while connecting to monitor: 2022-10-24T02:14:24.819644Z qemu-system-x86_64: -blockdev {"driver":"file","filename":"/var/lib/nova/mnt/99c4f7e8b15983b65e20cb7d37db899f/volume-8f478992-dde3-4c20-9005-61cd34eacf30","aio":"native","node-name":"libvirt-2-storage","cache":{"direct":true,"no-flush":false},"auto-read-only":true,"discard":"unmap"}: Could not open '/var/lib/nova/mnt/99c4f7e8b15983b65e20cb7d37db899f/volume-8f478992-dde3-4c20-9005-61cd34eacf30': Permission denied I've added appropriate configs to apparmor profile. (Using Ubuntu 22.04) Apparmor isn't blocking this access. While the instance is spawning, I've checked ownership of the volume file on compute node: root@compute-node:/var/lib/nova/mnt$ ls -al total 17 drwxr-xr-x 3 nova nova 4096 Oct 24 04:19 . drwxr-xr-x 12 nova nova 4096 Oct 24 02:14 .. drwxr-x--- 2 64061 64061 11 Oct 24 04:19 99c4f7e8b15983b65e20cb7d37db899f It seems like cinder user on storage node creates volume file with UID/GID of 64061 (cinder user's UID/GID) But nova user on compute node has UID/GID of 64060, therefore cannot open volume file(/var/lib/nova/mnt/99c4f7e8b15983b65e20cb7d37db899f/volume-8f478992-dde3-4c20-9005-61cd34eacf30) Should I manually set the UID/GID of nova user on compute node to 64061, so both nova user on compute node and cinder user on storage node would have the same UID/GID? Feels like this duct taping isn't a proper solution. Did I miss something? Thank you Diese E Mail enthält möglicherweise vertrauliche Inhalte und ist nur für die Verwertung durch den vorgesehenen Empfänger bestimmt. Sollten Sie nicht der vorgesehene Empfänger sein, setzen Sie den Absender bitte unverzüglich in Kenntnis und löschen diese E Mail. Hinweise zum Datenschutz finden Sie hier<https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.daten…>. Diese E Mail enthält möglicherweise vertrauliche Inhalte und ist nur für die Verwertung durch den vorgesehenen Empfänger bestimmt. Sollten Sie nicht der vorgesehene Empfänger sein, setzen Sie den Absender bitte unverzüglich in Kenntnis und löschen diese E Mail. Hinweise zum Datenschutz finden Sie hier<https://www.datenschutz.schwarz>.

1 0

RE: RE: [yoga][cinder] Cinder NFS backend: Compute service cannot access volume file (UID/GID problem)
by 박경원 24 Oct '22

24 Oct '22

1 0

[yoga][cinder] Cinder NFS backend: Compute service cannot access volume file (UID/GID problem)
by 박경원 24 Oct '22

24 Oct '22

2 1

[largescale-sig] No meeting this week
by Thierry Carrez 23 Oct '22

23 Oct '22

Hi everyone, With the PTG just over, the Large Scale SIG will not be meeting this week. Our next regular IRC meeting will be November 9, at 1500utc on #openstack-operators on OFTC. Feel free to add topics to the agenda: https://etherpad.openstack.org/p/large-scale-sig-meeting Regards, -- Thierry Carrez

1 0

single Network interface
by Parsa Aminian 23 Oct '22

23 Oct '22

Hello I use kolla ansible wallaby version . my compute node has only one port . is it possible to use this server ? as I know openstack compute need 2 port one for management and other for external user network . Im using provider_networks and it seems neutron_external_interface could not be the same as network_interface because openvswitch need to create br-ex bridge on separate port is there any solution that i can config my compute with 1 port ?

7 9