| opendevreview | Miguel Lavalle proposed openstack/neutron master: [PoC][DNM] Enable HA for OVN router flavors https://review.opendev.org/c/openstack/neutron/+/901513 | 01:25 |
|---|---|---|
| ralonsoh | jamesdenton__, hi, are you saying that GARPs from other hosts are coming into a second host ovnmeta namespace? | 08:25 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Add the bridge name and datapath type to the port VIF details https://review.opendev.org/c/openstack/neutron/+/903494 | 09:04 |
| ralonsoh | bcafarel, hi! please check https://review.opendev.org/c/openstack/neutron/+/903897. The upper 3 patches are already approved and waiting for this one | 09:05 |
| ralonsoh | slaweq, hello! if you have less than 1 min: https://review.opendev.org/q/Idfae87cf6c60e9e18ede91ea20857cea5322738c | 09:06 |
| ralonsoh | thanks in advance folks! | 09:06 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Add a start method to the OVN agent extensions https://review.opendev.org/c/openstack/neutron/+/903977 | 09:11 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: WIP == DNM == [OVN] Implement OVN agent metadata extension https://review.opendev.org/c/openstack/neutron/+/898238 | 09:11 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2023.1: Increase port name size and type to internal https://review.opendev.org/c/openstack/neutron/+/904818 | 09:16 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2023.1: [OVN] Retrieve the OVN agent extensions correctly https://review.opendev.org/c/openstack/neutron/+/903933 | 09:17 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: Forbid the subnet gateway IP deletion if a router interface is attached https://review.opendev.org/c/openstack/neutron/+/904713 | 09:25 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2023.2: [OVN] Add the bridge name and datapath type to the port VIF details https://review.opendev.org/c/openstack/neutron/+/904252 | 09:27 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2023.1: [OVN] Add the bridge name and datapath type to the port VIF details https://review.opendev.org/c/openstack/neutron/+/904254 | 09:30 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: Reset device namespace when adding to the namespace fails https://review.opendev.org/c/openstack/neutron/+/904825 | 11:22 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/xena: Reset device namespace when adding to the namespace fails https://review.opendev.org/c/openstack/neutron/+/904826 | 11:23 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/wallaby: Reset device namespace when adding to the namespace fails https://review.opendev.org/c/openstack/neutron/+/904827 | 11:23 |
| ralonsoh | bcafarel ^^ if you have 1 min | 11:24 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2023.1: Add sleep before checking if ovs port is in the namespace https://review.opendev.org/c/openstack/neutron/+/904829 | 11:30 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/zed: Add sleep before checking if ovs port is in the namespace https://review.opendev.org/c/openstack/neutron/+/904830 | 11:31 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: Add sleep before checking if ovs port is in the namespace https://review.opendev.org/c/openstack/neutron/+/904831 | 11:31 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/xena: Add sleep before checking if ovs port is in the namespace https://review.opendev.org/c/openstack/neutron/+/904832 | 11:31 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/wallaby: Add sleep before checking if ovs port is in the namespace https://review.opendev.org/c/openstack/neutron/+/904833 | 11:31 |
| ralonsoh | (zuul is going to start charging me...) | 11:33 |
| Continuity | lol | 11:38 |
| Continuity | Pay as you commit :D | 11:38 |
| ralonsoh | NEUTRINOS: as commented during the last Neutron meeting, I'm pinging you to spend some time reviewing the open specs | 11:40 |
| ralonsoh | https://review.opendev.org/q/project:openstack%252Fneutron-specs+status:open | 11:40 |
| ralonsoh | so please, during the next week we need to review and approve (or not) the open ones | 11:41 |
| ralonsoh | thanks! | 11:41 |
| opendevreview | Felipe Figueroa Vergara proposed openstack/tap-as-a-service master: Enables external network mirroring https://review.opendev.org/c/openstack/tap-as-a-service/+/896515 | 12:49 |
| jamesdenton___ | ralonsoh It looks like ARP requests to the subnet's gateway can be seen in the ovnmeta namespace on tap90748bb9-c1, and on the host itself on tap90748bb9-c0 (in br-int), but aren't seen on the provider interface. ARP requests for other hosts CAN be seen; it's only the gateway IP acting this way. This is familiar; i think i reported this symptom late last year, lemme find it | 13:49 |
| jamesdenton___ | FWIW the network is a VLAN provider network not attached to any routers and configured as the provisioning network for Ironic | 13:59 |
| ralonsoh | I'll try to reproduce this issue locally with VLAN networks | 14:00 |
| ralonsoh | #startmeeting neutron_drivers | 14:01 |
| opendevmeet | Meeting started Fri Jan 5 14:01:01 2024 UTC and is due to finish in 60 minutes. The chair is ralonsoh. Information about MeetBot at http://wiki.debian.org/MeetBot. | 14:01 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 14:01 |
| opendevmeet | The meeting name has been set to 'neutron_drivers' | 14:01 |
| ralonsoh | Ping list: ykarel, mlavalle, mtomaska, slawek, obondarev, tobias-urdin, lajoskatona, amotoki | 14:01 |
| ralonsoh | hello? | 14:02 |
| seba | heyjo :) | 14:02 |
| mtomaska | o/ | 14:02 |
| ykarel | o/ | 14:02 |
| ralonsoh | ok, due to the evident lack of quorum, I'm cancelling this meeting. I'll send a mail about that | 14:05 |
| ralonsoh | #endmeeting | 14:05 |
| opendevmeet | Meeting ended Fri Jan 5 14:05:16 2024 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 14:05 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/neutron_drivers/2024/neutron_drivers.2024-01-05-14.01.html | 14:05 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/neutron_drivers/2024/neutron_drivers.2024-01-05-14.01.txt | 14:05 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/neutron_drivers/2024/neutron_drivers.2024-01-05-14.01.log.html | 14:05 |
| seba | ralonsoh, do you still want to discuss the gateway ip allocation thing re https://review.opendev.org/c/openstack/neutron/+/904783 ? | 14:06 |
| ralonsoh | not today, the meeting was cancelled. I'll add it to the neutron meeting next week, on tuesday at 1400UTC | 14:07 |
| ralonsoh | most probably people will be back from holidays | 14:07 |
| ralonsoh | seba, in any case, I'm standing on what I commented in the review | 14:08 |
| ralonsoh | this is now currently limited for non-admin users | 14:08 |
| jamesdenton___ | ralonsoh for reference, here's a bug that had a similar issue with OVN "owning" the gateway address despite it not being attached to a router. But I'm not seeing all of those symptoms here, only what i described above. https://bugs.launchpad.net/neutron/+bug/2007167 | 14:08 |
| ralonsoh | and you can also specify rules to limit that for anyone, if needed | 14:08 |
| seba | yeah, makes sense if in vanilla openstack this is an admin-only action. didn't occur to me, as we have a custom polcy there | 14:08 |
| seba | well, I have users that use this feature, so I'll have to open a discussion if this is something we can forbid by policy, but it'll be either that or a downstream patch | 14:09 |
| ralonsoh | jamesdenton___, but are you talking about baremetal nodes or VMs? | 14:10 |
| seba | but if this is not something that might be usable by other people, we also don't need to get this upstream, then. | 14:10 |
| seba | other option would be to hide it behind a featureflag | 14:10 |
| ralonsoh | jamesdenton___, in any case, I think it could be required to limit the ARP table of the ovn meta to only the required IPs | 14:10 |
| jamesdenton___ | Well, this is all related to baremetal provisioning activities, but the VLAN provider network in question is used for both VMs and baremetal nodes in one of these two environments. In both cases, the ovnmeta namespace can't reach the gateway of the respective subnet; it's ARP requests for that gateway IP fail to get forwarded. Not a HUGE issue, i suppose, but sent me down a lengthy troubleshooting path. | 14:12 |
| ralonsoh | seba, what you are proposing is an API change, that could be enabled via an extension. But in any case, this is a feature already covered via policies | 14:12 |
| seba | ralonsoh, the feature I'm looking for is "allow users to specify the fip they want to allocate, but disallow allocating the network's subnet's gateway ip" - is that possible via policy? | 14:14 |
| ralonsoh | jamesdenton___, sorry, why the ovnmeta namespace needs to reach the GW IP? | 14:14 |
| ralonsoh | seba, no, it is not possible to make this kind of check via policies | 14:15 |
| jamesdenton___ | ralonsoh no need to jump on it now, i will file a bug and spell out the issue and you can determine if it's acting as designed or worth fixing. Also, ovnmeta wouldn't usually need to reach gateway IP but when trying to test connectivity while working out other issues it was evident that ovnmeta couldn't reach the (known) gateway ip and exhibited the behavior described. So, spent some time going down that rabbit hole and thought maybe | 14:16 |
| jamesdenton___ | it was a bigger issue | 14:16 |
| mtomaska | Hello. Just want to check my understanding. It looks like we always start RPC server listeners no matter what type of Ml2 back end we use. Am I correct? https://opendev.org/openstack/neutron/src/branch/master/neutron/plugins/ml2/plugin.py#L282-L309 Meaning, even If I have a ML2/OVN deployment without metadata, dhcp, etc. agents , the server side of the RPC is always started and listening. https://opendev.org/op | 15:29 |
| mtomaska | enstack/neutron/src/branch/master/neutron/plugins/ml2/plugin.py#L282-L309 | 15:29 |
| mtomaska | ralonsoh ^ is my understanding correct? | 15:40 |
| ralonsoh | mtomaska, yes, we create the default rpc resources | 15:44 |
| ralonsoh | for networks and ports (for example, for the sriov agent with OVN) | 15:45 |
| ralonsoh | https://paste.opendev.org/show/bQzss2JmZXmYrJjSjMkJ/ | 15:45 |
| ralonsoh | ^ the RPC worker (configured only 1 in my env) is running | 15:45 |
| mtomaska | thank you | 16:01 |
| opendevreview | Miro Tomaska proposed openstack/neutron master: [WIP] Make get_ports RPC method conmon for the DHCP and Metadata agent This WIP patch is the initial implementation on the suggestion from this patch[1] https://review.opendev.org/c/openstack/neutron/+/904872 | 18:53 |
| opendevreview | Merged openstack/neutron stable/2023.1: Improve "sync_ha_chassis_group" method https://review.opendev.org/c/openstack/neutron/+/903897 | 23:59 |
| opendevreview | Merged openstack/neutron stable/2023.1: Fix the common/ovn functional tests https://review.opendev.org/c/openstack/neutron/+/903898 | 23:59 |
| opendevreview | Merged openstack/neutron stable/2023.1: Reintroduce agent bridge resync test https://review.opendev.org/c/openstack/neutron/+/903938 | 23:59 |
| opendevreview | Merged openstack/neutron stable/2023.1: ovn-metadata: Refactor events https://review.opendev.org/c/openstack/neutron/+/903509 | 23:59 |
| opendevreview | Merged openstack/neutron stable/2023.1: Enhanced external port scheduling https://review.opendev.org/c/openstack/neutron/+/903510 | 23:59 |
| opendevreview | Merged openstack/neutron stable/2023.1: Update the External Ports documentation https://review.opendev.org/c/openstack/neutron/+/903899 | 23:59 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!