February 2022 - openstack-discuss

[infra] Gerrit email outage?
by Stephen Finucane 23 Feb '22

23 Feb '22

I haven't received any emails from Gerrit since midday (UTC) yesterday. Chatting on IRC, it seems I'm not alone. Has something changed in Gerrit or is this a simple outage? Cheers, Stephen

2 2

[kolla][rabbitmq] Trouble with RMQ 3.8 in Train
by Albert Braden 23 Feb '22

23 Feb '22

We are running kolla-ansible Train and I successfully followed the recommendations in [1] to change our RMQ settings, and our RMQ works a lot better now, but one of the recommendations is to run RMQ 3.8 and so far I haven’t been able to make it work. Our working RMQ container has these settings in the Dockerfile to force 3.7 and the appropriate erlang version: 'erlang-21.3*', 'rabbitmq-server-3.7.23*' I changed those settings to allow 3.8: 'rabbitmq-server-3.*.*el8ost*' Now my RMQ container fails to start with: {"log":"20:29:28.440 [error] Supervisor rabbit_prelaunch_sup had child prelaunch started with rabbit_prelaunch:run_prelaunch_first_phase() at undefined exit with reason no_configuration_schema_found in context start_error\r\n","stream":"stdout","time":"2022-02-22T20:29:28.441064637Z"} What am I missing? [1] https://wiki.openstack.org/wiki/Large_Scale_Configuration_Rabbit

2 1

[qa] Zed PTG planning
by Martin Kopec 23 Feb '22

23 Feb '22

Hi everyone, virtual PTG for the Zed cycle will be held between 4th - 8th April [1]. I've created a Doodle [2] to collect the best time slots for our team. Please fill it in rather sooner than later (the deadline for team signup is March 11th). I've also created an etherpad [3] to collect ideas/topics for the PTG sessions. If you have anything to discuss, please don't hesitate to write it there. [1] http://lists.openstack.org/pipermail/openstack-discuss/2022-February/027051… [2] https://doodle.com/poll/q5yzatzh45brt5v2 [3] https://etherpad.opendev.org/p/qa-zed-ptg Thanks, -- Martin Kopec Senior Software Quality Engineer Red Hat EMEA IM: kopecmartin

1 0

[neutron] Stale Tap Ports
by Ammad Syed 23 Feb '22

23 Feb '22

Hi, I am using neutron 19.1 and ovn 21.09. There is a strange behavior I am seeing on compute nodes that ovs logs shows below logs 2022-02-23T06:14:06.426Z|00226|bridge|WARN|could not open network device tapea0dca9a-32 (No such device) 2022-02-23T06:14:06.428Z|00227|bridge|WARN|could not open network device tapaeaa1749-b5 (No such device) 2022-02-23T06:14:06.429Z|00228|bridge|WARN|could not open network device tap53f06ea0-70 (No such device) 2022-02-23T06:14:06.431Z|00229|bridge|WARN|could not open network device tap724cf0cc-7b (No such device) 2022-02-23T06:14:06.433Z|00230|bridge|WARN|could not open network device tap18864125-6d (No such device) In the ovs-vsctl show I also see alot of ports with these status. Port tapaeaa1749-b5 Interface tapaeaa1749-b5 error: "could not open network device tapaeaa1749-b5 (No such device)" Port tap18864125-6d Interface tap18864125-6d error: "could not open network device tap18864125-6d (No such device)" Port tap54415fc0-dd Interface tap54415fc0-dd error: "could not open network device tap54415fc0-dd (No such device)" I have checked in neutron port list and there is no port found for these tap interfaces. openstack port list --long | grep -i 18864125-6d I have checked in database in ports table, there is also no ports there with these ids. mysql> select * from ports where id like '%18864125-6d%'; These ports even show on compute node even if there is no instance running on it. Are these stale ports ? If yes, is there any tool or utility to run audit and delete these ports ? Ammad

1 0

Re: [Tempest] OpenSack Powered * vs OpeStack Compatible
by yahoo yahho 23 Feb '22

23 Feb '22

Sent from Yahoo Mail on Android

1 0

[all] [requirements] Artificially inflated requirements in os-brick
by Thomas Goirand 22 Feb '22

22 Feb '22

Hi, I've been quite surprised to see $topic. Most requirements for os-brick are now to be found in Yoga only. I've asked why, and I've been told that the reason is because "we've been testing with the new versions only for the number of months". Well, the reason why someone would increase a minimum requirement *must* be only because of the need of a new feature, and should be treated with care. Otherwise, this makes upgrading very dangerous and annoying to do. As much as possible, I would strongly recommend that any dependency in Stable can be found in Stable-1 (only when a new feature is mandatory, then it's ok-ish to require that new version, though I would advocate for a possible fallback if that's not too complex to write). Now, if that's the path we're taking, all is going backward 5 years ago, and then my thinking is: we must reintroduce lower-constraints testing ASAP. Your thoughts everyone? Cheers, Thomas Goirand (zigo) P.S: Please note that this is absolutely not about os-brick, I'm not pointing my finger to any team, it's a global thinking for the whole OpenStack project.

5 10

LetsEncrypt OS Ansible Ussuri
by Marc-Antoine Godde 22 Feb '22

22 Feb '22

Hello, I have a question on how to setup LetsEncrypt with OpenStack Ansible. We are still on OpenStack Ussuri. We added the following variables to user_variables.yml. ================================================================================== haproxy_ssl_letsencrypt_enable: True haproxy_ssl_letsencrypt_install_method: "distro" haproxy_ssl_letsencrypt_setup_extra_params: "--http-01-address {{ ansible_host }} --http-01-port 8888" haproxy_ssl_letsencrypt_email: email(a)example.com haproxy_interval: 2000 user avatar user avatar haproxy_extra_services: # an internal only service for acme-challenge whose backend is certbot on the haproxy host - service: haproxy_service_name: letsencrypt haproxy_backend_nodes: - name: localhost ip_addr: {{ ansible_host }} #certbot binds to the internal IP backend_rise: 1 #quick rise and fall time for multinode deployment to succeed backend_fall: 2 haproxy_bind: - 127.0.0.1 #bind to 127.0.0.1 as the local internal address will be used by certbot haproxy_port: 8888 #certbot is configured with http-01-port to be 8888 haproxy_balance_type: http ================================================================================== Yet, Horizon config for HAproxy is already defined in the default vars (https://github.com/openstack/openstack-ansible/blob/stable/ussuri/inventory… <https://github.com/openstack/openstack-ansible/blob/stable/ussuri/inventory…>) and we don’t know where ta add the required ACL to redirect the traffic from 80 port to 8888: ==================================== haproxy_frontend_acls: #use a frontend ACL specify the backend to use for acme-challenge letsencrypt-acl: rule: "path_beg /.well-known/acme-challenge/" backend_name: letsencrypt ==================================== We know that this is fixed in OpenStack Ansible Victoria. Is it possible with Ussuri tho ? Many thanks, Best, Marc-Antoine Godde

2 5

[cinder] reminder: this week's meeting in video+IRC
by Brian Rosmaita 22 Feb '22

22 Feb '22

Quick reminder that this week's Cinder team meeting on Wednesday 23 February, being the final meeting of the month, will be held in both videoconference and IRC at the regularly scheduled time of 1400 UTC. These are the video meeting rules we've agreed to: * Everyone will keep IRC open during the meeting. * We'll take notes in IRC to leave a record similar to what we have for our regular IRC meetings. * Some people are more comfortable communicating in written English. So at any point, any attendee may request that the discussion of the current topic be conducted entirely in IRC. * The meeting will be recorded. connection info: https://bluejeans.com/3228528973 meeting agenda: https://etherpad.opendev.org/p/cinder-yoga-meetings cheers, brian

1 0

[nova] OOM Killed Processes
by Ammad Syed 22 Feb '22

22 Feb '22

Hi,, I am having trouble with my compute node that nova-compute and ovs process are being killed by OOM. I have alot memory available in system. # free -g total used free shared buff/cache available Mem: 1006 121 881 0 2 879 Swap: 7 0 7 But I am seeing process are being killed in dmesg. [Sat Feb 19 03:46:26 2022] Memory cgroup out of memory: Killed process 2080898 (ovs-vswitchd) total-vm:9474284kB, anon-rss:1076384kB, file-rss:11700kB, shmem-rss:0kB, UID:0 pgtables:2776kB oom_score_adj:0 [Sat Feb 19 03:47:01 2022] Memory cgroup out of memory: Killed process 2081218 (ovs-vswitchd) total-vm:9475332kB, anon-rss:1096988kB, file-rss:11700kB, shmem-rss:0kB, UID:0 pgtables:2780kB oom_score_adj:0 [Sat Feb 19 03:47:06 2022] Memory cgroup out of memory: Killed process 2081616 (ovs-vswitchd) total-vm:9473252kB, anon-rss:1073052kB, file-rss:11700kB, shmem-rss:0kB, UID:0 pgtables:2784kB oom_score_adj:0 [Sat Feb 19 03:47:16 2022] Memory cgroup out of memory: Killed process 2081940 (ovs-vswitchd) total-vm:9471236kB, anon-rss:1070920kB, file-rss:11700kB, shmem-rss:0kB, UID:0 pgtables:2776kB oom_score_adj:0 [Sat Feb 19 03:47:16 2022] Memory cgroup out of memory: Killed process 6098 (nova-compute) total-vm:3428356kB, anon-rss:279920kB, file-rss:9868kB, shmem-rss:0kB, UID:64060 pgtables:1020kB oom_score_adj:0 [Mon Feb 21 11:15:08 2022] Memory cgroup out of memory: Killed process 2082296 (ovs-vswitchd) total-vm:9475372kB, anon-rss:1162636kB, file-rss:11700kB, shmem-rss:0kB, UID:0 pgtables:2864kB oom_score_adj:0 Any advice on how to fix this ? Also any best practices document on configuring memory optimizations in nova compute node. Ammad

2 3

[openstacksdk][magnum] Adding get_coe_cluster_by_id() method to openstack.cloud._coe
by Florian Haas 22 Feb '22

22 Feb '22

Hi everyone, here's something I recently ran into while playing with openstacksdk against Magnum; I could use some help/feedback there if anyone's inclined. :) It looks as though openstacksdk, which presently doesn't include get_coe_cluster_by_id() in cloud._coe[1], will effectively only ever list clusters and then filter client-side, as opposed to looking up individual clusters directly. That means that it only ever hits the Magnum API's /v1/clusters URL[2], which provides only limited properties on each cluster it returns. Were it to implement get_coe_cluster_by_id(), then any Connection object that has use_direct_get set to True could hit /v1/clusters/<uuid> instead[3], which provides a much richer set of properties. I have submitted a patch for this[4], but I have a couple of open questions: (1) I'm not quite sure what's the proper way to test that get_coe_cluster() actually invokes get_coe_cluster_by_id() when passed a UUID. Normally I'd do that with unittest.mock's assert_called() method, but openstacksdk heavily wraps various mock.patch() calls so I'm sure there's a better, preferred way to do it for the openstacksdk codebase. (2) Even if implementing get_coe_cluster_by_id() is the correct approach (which I am entirely not sure of), it still strikes me as a bit unintuitive that there are additional details that could be retrieved only if use_direct_get is set on the Connection object. I think having an explicit "get me more detail" toggle on the call would be preferable, which would hit /v1/clusters first (to retrieve the cluster UUID from the name), and then hit /v1/clusters/<uuid> to retrieve the rich set of properties. Now, the docstring for search_coe_clusters() mentions a "detail" parameter[5], but that appears to be a bit of a red herring as that method doesn't actually use that param. So I'm a bit stuck and unsure what to do there. :) If anyone has thoughts on this they'd like to share, I'd be most grateful! Cheers, Florian References: [1] https://opendev.org/openstack/openstacksdk/src/branch/master/openstack/clou… [2] https://docs.openstack.org/api-ref/container-infrastructure-management/?exp… [3] https://docs.openstack.org/api-ref/container-infrastructure-management/?exp… [4] https://review.opendev.org/c/openstack/openstacksdk/+/828791 [5] https://opendev.org/openstack/openstacksdk/src/branch/master/openstack/clou…

1 1