April 2023 - openstack-discuss - lists.openstack.org

[glance][cinder]create volume from image too slow
by Nguyễn Hữu Khôi 14 Apr '23

14 Apr '23

Hello guys. I am using nfs for glance and SAN for cinder but when i create 1 gb volume with 100mb image(Converted 112.00 MB image at 25.4 MB/s) then it will fast. however, if I create 100gb volume with 100mb image(Converted 112.00 MB image at 8.24 MB/s) then it will very slow. Pls give me some advices. ty Nguyen Huu Khoi

1 1

[neutron] Drivers meeting agenda
by Rodolfo Alonso Hernandez 14 Apr '23

14 Apr '23

Hello Neutrinos: Remember we have the Neutron drivers meeting today at 14UTC. The agenda is here: https://wiki.openstack.org/wiki/Meetings/NeutronDrivers Regards.

1 0

[Controller/Infra node removal] [openstack-ansible] Need to remove infra/controller node from environment.
by Danish Khan 14 Apr '23

14 Apr '23

Hello Team, Is it possible to remove an infra host from environment? If yes, can you please share the exact steps that need to be followed. I am trying to remove it manually by destroying all lxc containers from infra host which needs to be removed and then running setup openstack playbook. but it is not working properly. Can someone please share the exact steps needed to be followed to remove infra host from the environment. Openstack version: Wallaby or Yoga. Regards, Danish

2 1

[kolla-ansible][yoga] is there a way to deploy or reconfigure a container to run in privileged mode?
by wodel youchi 14 Apr '23

14 Apr '23

Hi, Is there a way to deploy or reconfigure a container to run in privileged mode? I need to make glance-api run in privileged mode. Regards.

2 1

[ptl] Need PTL volunteer for OpenStack Winstackers
by Ghanshyam Mann 13 Apr '23

13 Apr '23

Hi Lukas, I am reaching out to you as you were PTL for OpenStack Winstackers project in the last cycle. There is no PTL candidate for the next cycle (2023.2), and it is on the leaderless project list. Please check if you or anyone you know would like to lead this project. - https://etherpad.opendev.org/p/2023.2-leaderless Also, if anyone else would like to help leading this project, this is time to let TC knows. -gmann

7 14

[qa][patrole] Retiring Patrole project
by Ghanshyam Mann 13 Apr '23

13 Apr '23

Hello Everyone, You might know that Patrole project[1] which was created to test the RBAC but had some design challenges, and we never got much attention to this project to align it with the new SRBAC or efficient way to test the RBAC. Since the last couple of years (After AT&T's focus was changed from OpenStack), we hardly had any maintainer in this project. In this cycle, the QA team decided to retire this project. There is always an option to bring it back in OpenStack if anyone is interested in maintaining it. I have pushed the changes required for retirement, including governance change. If anyone using or interested in this project, feel free to reply here or in governance change[2]. [1] https://wiki.openstack.org/wiki/QA#Patrole_.28RETIRED.29 [2] https://review.opendev.org/c/openstack/governance/+/880014 -gmann

1 0

[heat] Performance scaling with number of resources
by John Bartelme 13 Apr '23

13 Apr '23

Hello, I'm wondering about heat performance when I scale to a larger number of resources. A single resource deploys quite quickly but when I deploy the below template it takes 2 minutes for this stack to deploy. This occurs on Antelope via DevStack on Ubuntu 22.04 as well as a full production cluster(yoga) deployed with OpenStack-Ansible on RHEL 8. Not only does this template take 2 minutes to deploy it also takes 2 minutes to update without any changes. Is there anything in the stack configuration or heat configuration I can do to speed up the process of updating the stack when very little changes? Or this something inherit in the design of heat? --- heat_template_version: 2021-04-16 description: test resources: test-resources: type: OS::Heat::ResourceGroup properties: count: 100 resource_def: type: OS::Heat::None --- A little more detail: Obviously I have no need to deploy this template but it is a much more simplified case of what I really want to deploy. I'm deploying resource groups that contain hundreds of custom resources(containing network/compute/storage etc.) that are conditionally enabled or disabled based on a JSON property array(in general only about 20 or so resources are deployed but depending on needs any one of the resources might be needed in the near future). The JSON properties array is complex enough that I really want to have it all defined so that changing a boolean value is all that is needed to turn on and off resources without having to fully write new complex heat templates. Additionally the nature of the workflow is that I want to add future resources to the existing stack. I don't change the number of the resources to the resource group, i.e. count=20 instead of 100 because when I conditionally add another resource it causes all the existing resources to be renumbered. Ideally I could rapidly modify one value to the json array and then update the stack and have it quickly modified. Thanks, john

1 0

[manila] Cancelling next week's weekly meeting
by Carlos Silva 13 Apr '23

13 Apr '23

Hello everyone! As we will have a significant drop in attendance in next week's IRC weekly meeting (Apr 20th), we will be cancelling it. The next IRC weekly meeting will be on Apr 27th. Thanks, carloss

1 0

[oslo.messaging][nova][neutron] quorum queues
by Arnaud Morin 13 Apr '23

13 Apr '23

Hey team, I am currently looking into using quorum queues (see patch [1] introduced in oslo_messaging) for openstack services (nova, neutron mostly). After switching the value to true and bootstraping a new cluster, all queues are quorum except the fanout and reply queues. I understand that transient queues are managed differently in openstack, but why is it so? I am asking this because when using rabbit in cluster mode, the quorum queues are replicated accross the cluster. But because the fanout and reply queues are not replicated, if one of the cluster is shutting down during an instance boot, the boot may endup in error (the reply queue may be on the faulty node). Is there any reason to not enable quorum on transient queues? [1] https://github.com/openstack/oslo.messaging/commit/7e8acbf87070354cd2438f3a…

1 0

Updating OS-TRAITS in placement stable/yoga and stable/zed
by Karl Kloppenborg 13 Apr '23

13 Apr '23

Hello Placement and OS-Traits teams. Karl from Openstack-Helm. After debugging a number of issues in Cyborg, it has become necessary to update the pinned OS-TRAITs version in stable/yoga and stable/zed. I have submitted the following two reviews for updating the pinned versions to 2.10.0 https://review.opendev.org/c/openstack/placement/+/880249 https://review.opendev.org/c/openstack/placement/+/880249 By merging these changes, it will allow us to continue to use vGPU capabilities in cyborg in the stable/yoga and stable/zed onwards branches. Kind Regards, -- Karl Kloppenborg, (BCompSc, CNCF-[KCNA, CKA, CKAD], LFCE, CompTIA Linux+ XK0-004) Managing Director, Invention Labs.

3 2