March 2020 - openstack-discuss - lists.openstack.org

[Horizon] Unable to access the dashboard page
by Rahul Sharma 09 Mar '20

09 Mar '20

Hi, I am trying to get OpenStack (Train) up and running on CentOS 7 by following the "OpenStack Installation Guide" provided on OpenStack's website and have completed installation of below components - Keystone - Glance - Placement - Nova - Networking After installing each component, I have also verified its operation and it seems to be working successfully. However, there is a problem I am facing after installing "Horizon" for dashboard services. In order to verify its operation, one is supposed to browse to the URL "http://controller/horizon/" where "controller" could be the hostname or IP address of the Node which is running the controller Browsing to the above URL throws an error "The requested URL /horizon was not found on this server." In the apache access logs, I see below error: " 103.44.50.92 - - [07/Mar/2020:13:14:52 +0000] "GET /horizon/ HTTP/1.1" 404 206 "-" "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.132 Safari/537.36" 103.44.50.92 - - [07/Mar/2020:13:14:52 +0000] "GET /favicon.ico HTTP/1.1" 404 209 "http://3.21.90.63/horizon/" "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.132 Safari/537.36" " If I browse to the URL "http://controller/", the default "Testing123" page of apache gets loaded. Please assist. Thanks, Rahul

4 7

[CINDER] Snapshots export
by Alfredo De Luca 09 Mar '20

09 Mar '20

Hi all. We have our env with Openstack (Train) and cinder with CEPH (nautilus) backend. We are creating automatic volumes snapshots and now we'd like to export them as a backup/restore plan. After exporting the snapshots we will use Acronis as backup tool. I couldn't find the right steps/commands to exports the snapshots. Any info? Cheers -- *Alfredo*

5 6

[all][tc] What happened in OpenStack Governance recently
by Jean-Philippe Evrard 09 Mar '20

09 Mar '20

Hello, It's been a while there was no update on "what happened in governance recently?" (February). So here is a summary for you! We had two meetings since the last community update, February [1] and March [2]. We have established our first (business-focused) upstream investment opportunities for 2020, which currently consists of: - Goal champions - Consistent and secure policy defaults - QA developers You can see the latest version in [3]. The timing for the next PTL and TC elections is now determined [4]. We've introduced the ideas framework, allowing you to propose, follow, and track large scale changes of OpenStack, with their conversations history [5]. We are starting to split OpenDev out of OpenStack-infra. Our next release names are Victoria and Wallaby. In terms of community goals [6], we've changed the U goal for documentation, so please have a look if you haven't already [7][8], and/or talk to the goal champion Kendall (diablo_rojo). The 'Switch legacy Zuul jobs to native' goal has been selected for Victoria. Please talk to its champion, Luigi Toscano (tosky)! On the projects' side, we've clarified the guidelines to drop official project teams [9]. We want to retire teams more actively if necessary, to allow us to increase our focus. Some projects have changed: - networking-ovn won't get releases anymore, as it's merged back into neutron in U [10] - neutron-lbaas was retired in Ussuri [11] Have a good day to you all! Regards, Jean-Philippe Evrard (evrardjp) [1]: http://eavesdrop.openstack.org/meetings/tc/2020/tc.2020-02-06-14.00.html [2]: http://eavesdrop.openstack.org/meetings/tc/2020/tc.2020-03-05-14.00.html [3]: https://governance.openstack.org/tc/reference/upstream-investment-opportuni… [4]: https://review.opendev.org/#/c/708470/3/configuration.yaml [5]: http://lists.openstack.org/pipermail/openstack-discuss/2020-February/012847… [6]: https://governance.openstack.org/tc/goals/selected/index.html [7]: https://review.opendev.org/#/c/708672/ [8]: https://review.opendev.org/#/c/709617/ [9]: https://review.opendev.org/#/c/707421/ [10]: https://review.opendev.org/#/c/705781/ [11]: https://review.opendev.org/#/c/705780/

1 0

[tc] April meeting
by Jean-Philippe Evrard 09 Mar '20

09 Mar '20

Hello everyone, It would be nice if you could update the agenda on the wiki [1] for the April meeting, happening on April, 2nd. Thank you in advance, Jean-Philippe Evrard (evrardjp). [1]: https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Next_Meeting

1 0

[tc] March meeting
by Jean-Philippe Evrard 09 Mar '20

09 Mar '20

Hello, Here are the minutes for the march meeting [1]. There are a few action points for you, please have a look! Regards, Jean-Philippe Evrard (evrardjp) [1]: http://eavesdrop.openstack.org/meetings/tc/2020/tc.2020-03-05-14.00.html

1 0

[queens] [neutron]security_groups_log]
by Ignazio Cassano 08 Mar '20

08 Mar '20

Hello, I have queens installation based on centos7. Before implementing security groups logs, I had the following configuration in /etc/neutron/plugins/ml2/openvswitch_agent.ini: firewall_driver = iptables_hybrid Enabling security groups log I had to change it in : firewall_driver = openvswitch It seems to work end security logs are logged . After restarting kvm nodes and controllers, virtual machines do not live migrate. The firewall driver change could be the cause of my problem ? firewall_driver = openvswitch is mandatory for security groups log ? Please, any help ? I cannot reproduce the problem rebooting all my nodes. I rebooted them because I hat to transfer from a rack to another. Ignazio

2 5

[queens] [neutron]security_groups_log] issues
by Ignazio Cassano 07 Mar '20

07 Mar '20

Hello, I have queens installation based on centos7. Before implementing security groups logs, I had the following configuration in /etc/neutron/plugins/ml2/openvswitch_agent.ini: firewall_driver = iptables_hybrid Enabling security groups log I had to change it in : firewall_driver = openvswitch It seems to work end security logs are logged . After restarting kvm nodes and controllers, virtual machines do not live migrate. I rolled back my configuration without security groups logs feature and now all works fine. The firewall driver change could be the cause of my problem ? firewall_driver = openvswitch is mandatory for security groups log ? Please, any help ? I cannot reproduce the problem rebooting all my nodes. I rebooted them because I hat to transfer from a rack to another. Ignazio

1 0

[nova][qa] When creating a instance using a flavor with "hw:cpu_policy=dedicated" "hw:cpu_realtime=yes" and "hw:cpu_realtime_mask=^0-1" failed.
by Juntingqiu Qiujunting (邱军婷) 06 Mar '20

06 Mar '20

When I create a instance using a flavor with "hw:cpu_policy=dedicated" "hw:cpu_realtime=yes" and "hw:cpu_realtime_mask=^0-1". The error is "libvirtError:Cannot set scheduler parameters for pid 3666:operation not permitted". https://bugs.launchpad.net/starlingx/+bug/1866311 The instance xml as following: <vcpu placement='static'>4</vcpu> <cputune> <vcpupin vcpu='0' cpuset='3'/> <vcpupin vcpu='1' cpuset='11'/> <vcpupin vcpu='2' cpuset='0'/> <vcpupin vcpu='3' cpuset='8'/> <emulatorpin cpuset='3,11'/> <vcpusched vcpus='2' scheduler='fifo' priority='1'/> <vcpusched vcpus='3' scheduler='fifo' priority='1'/> </cputune> The error log as following: ERROR nova.compute.manager [instance: 74dcc0a1-9ed4-4d13-bef7-ac9623bca2d0] File "/usr/lib64/python2.7/site-packages/libvirt.py", line 1099, in createWithFlags ERROR nova.compute.manager [instance: 74dcc0a1-9ed4-4d13-bef7-ac9623bca2d0] if ret == -1: raise libvirtError ('virDomainCreateWithFlags() failed', dom=self) ERROR nova.compute.manager [instance: 74dcc0a1-9ed4-4d13-bef7-ac9623bca2d0] libvirtError: Cannot set scheduler parameters for pid 6504: Operation not permitted ERROR nova.compute.manager [instance: 74dcc0a1-9ed4-4d13-bef7-ac9623bca2d0]

2 1

Re: [sdk][congress][octavia][designate][magnum][senlin][swift] Adding project-specific cores to SDK
by Eric Fried 06 Mar '20

06 Mar '20

Any chance of reusing the “PTL ack” bot that has recently appeared in the releases repo? But as a “SME ack” that would recognize anyone from $project’s core team? (How does the releases bot know which project the patch is for? Might have to be a bit fuzzy on that logic for SDK/OSC.) Then the team could adopt a policy of single core approval if the patch has this SME +1, and no real danger of “abuse”. Eric Fried

4 4

[sdk][congress][octavia][designate][magnum][senlin][swift] Adding project-specific cores to SDK
by Monty Taylor 06 Mar '20

06 Mar '20

Heya, I’d like to try something. I’d like to try adding some project-specific people to the core team so that they can more directly help maintain the support for their service in SDK. In some of these cases the person I’m suggestion has next to no review experience in SDK. I think let’s be fine with that for now - we’re still a 2x +2 in general thing - but I know currently when reviewing neutron or ironic changes I always want to see a +2 from slaweq or dtantsur … so in the spirit of trying new things and trying to move the project forward in a healthy and welcoming way - how about we give this a try? The idea here is that we’re trusting people to use their good judgement and to only use their new +2 powers for good in their project. Over time, if they feel like they’ve gotten a handle on things more widely, there’s nothing stopping them from reviewing other patches - but I think that most of us aren’t looking for additional review work anyway. Specifically this would be: Shogo Saito - congress Adam Harwell - octavia Graham Hayes - designate Bharat Kumar - magnum Erik Olof Gunnar Andersson - senlin Tim Burke - swift I think we should also add a file in the repo that lists “subject matter experts” for each service we’ve got support for, where we have them. My list of current cores who I’d ping for specific service suitability are: Sean McGinnis - cinder Slawek Kaplonski - neutron Dmitry Tantsur - ironic Eric Fried - nova (at least until tomorrow my friend) How does that sound to folks? Monty

4 4