| opendevreview | Merged openstack/neutron stable/2024.1: Check unspecified flavor in user defined driver https://review.opendev.org/c/openstack/neutron/+/914251 | 00:51 |
|---|---|---|
| opendevreview | Miguel Lavalle proposed openstack/neutron master: Enable HA for OVN router flavors https://review.opendev.org/c/openstack/neutron/+/901513 | 01:03 |
| opendevreview | Miguel Lavalle proposed openstack/neutron master: Enable HA for OVN router flavors https://review.opendev.org/c/openstack/neutron/+/901513 | 01:08 |
| *** mklejn_ is now known as mklejn | 06:43 | |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: Remove the Windows OS libraries https://review.opendev.org/c/openstack/neutron/+/914271 | 07:20 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Create an OVN DB transaction context decorator https://review.opendev.org/c/openstack/neutron/+/910538 | 07:22 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] "Logical_Router" pinned to chassis, OVN L3 scheduler https://review.opendev.org/c/openstack/neutron/+/909194 | 07:22 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Provide HA functionality to "Logical_Router" chassis pinning https://review.opendev.org/c/openstack/neutron/+/909437 | 07:22 |
| opendevreview | Merged openstack/neutron master: Update jobs based on testing runtime for 2024.2 https://review.opendev.org/c/openstack/neutron/+/913923 | 07:34 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: Unify the Metadata proxy handler code https://review.opendev.org/c/openstack/neutron/+/914428 | 08:22 |
| opendevreview | Lajos Katona proposed openstack/networking-bgpvpn master: Update jobs based on testing runtime for 2024.2 https://review.opendev.org/c/openstack/networking-bgpvpn/+/914273 | 08:36 |
| opendevreview | Lajos Katona proposed openstack/networking-bagpipe master: Update jobs based on testing runtime for 2024.2 https://review.opendev.org/c/openstack/networking-bagpipe/+/914272 | 08:42 |
| opendevreview | Lajos Katona proposed openstack/neutron-vpnaas master: Update jobs based on testing runtime for 2024.2 https://review.opendev.org/c/openstack/neutron-vpnaas/+/914280 | 08:45 |
| opendevreview | Merged openstack/neutron master: Disable skip-level jobs in check queue https://review.opendev.org/c/openstack/neutron/+/913451 | 09:26 |
| jakeyip | Hi, I have a question, is linuxbridge + ovn mechanism drivers actually a supported configuration? we are running that in our deployment, and a recent change [1] broke our deployment. | 10:07 |
| jakeyip | [1] https://review.opendev.org/c/openstack/neutron/+/886988/11/neutron/agent/linux/dhcp.py#1486 | 10:07 |
| jakeyip | just wondering if it is something worth filing a bug for... or it's not supported | 10:08 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron-lib master: WIP == Add the field "rbac_actions" to the model query hooks https://review.opendev.org/c/openstack/neutron-lib/+/914473 | 10:46 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron-lib master: WIP == Add the field "rbac_actions" to the model query hooks https://review.opendev.org/c/openstack/neutron-lib/+/914473 | 10:46 |
| ralonsoh | jakeyip, no, OVN and LB cannot run together | 10:48 |
| ralonsoh | only OVN + SRIOV | 10:48 |
| ralonsoh | lajoskatona, hi! | 10:51 |
| ralonsoh | I think gmann commented not to add py312 in the gate while not voting | 10:51 |
| ralonsoh | https://review.opendev.org/c/openstack/networking-bgpvpn/+/914273/3/.zuul.yaml | 10:51 |
| jakeyip | ralonsoh: thanks! any idea if this is in a doc? https://docs.openstack.org/neutron/latest/admin/config-ml2.html#mechanism-drivers doesn't seem to have this info? | 10:53 |
| ralonsoh | jakeyip, not that I'm aware. Linux Bridge, OVS and OVN are not tested working together and I don't think that was supposed to work | 10:56 |
| jakeyip | ralonsoh: and I think OVN + baremetal is supported? e.g. https://review.opendev.org/c/openstack/neutron/+/886988 | 10:57 |
| ralonsoh | it is yes | 10:57 |
| jakeyip | I'm not familiar with baremetal but I think might be https://docs.openstack.org/networking-baremetal/latest/ ? | 10:58 |
| jakeyip | if someone will to send patches for linuxbridge + OVN, will that be considered? | 10:59 |
| ralonsoh | jakeyip, I'm not familiar neither with baremetal, I would ask ironic folks | 10:59 |
| jakeyip | I am not sure what's the neutron team take on it... the support for OVN + baremetal broke our OVN + linuxbridge | 10:59 |
| ralonsoh | jakeyip, as commented, LB, OVS and OVN are not supposed to work together | 10:59 |
| ralonsoh | and LB is deprecated in Neutron | 10:59 |
| ralonsoh | but if your patches do not affect any working code, please feel free to open a Launchpad bug describing the issue and the rationale of the code to be pushed | 11:00 |
| jakeyip | ok if LB is going to be deprecated; we can also look towards migrating away, thanks. | 11:01 |
| jakeyip | what are considered the 'first class' mech drivers now? | 11:02 |
| ralonsoh | OVS and OVN (and SRIOV) | 11:03 |
| jakeyip | ok good to know, we can plan accordingly | 11:09 |
| jakeyip | ralonsoh: thanks for taking the time to answer my queries, appreciate it :) | 11:11 |
| ralonsoh | yw | 11:13 |
| lajoskatona | ralonsoh: Hi, but py312 job is now by default non-voting | 12:44 |
| lajoskatona | ralonsoh: perhaps I am false and it is not non-voting by default, let me check it | 12:44 |
| ralonsoh | lajoskatona, yes, is non voting by default | 12:46 |
| ralonsoh | but gmann commented to remove it from the gate | 12:46 |
| ralonsoh | we should not have non-voting jobs in the gate | 12:47 |
| ralonsoh | lajoskatona, and re-add again this job to the experimental/periodic queue, as in PS2 | 12:47 |
| *** mklejn__ is now known as mklejn | 13:15 | |
| opendevreview | Lajos Katona proposed openstack/neutron-vpnaas master: Update jobs based on testing runtime for 2024.2 https://review.opendev.org/c/openstack/neutron-vpnaas/+/914280 | 13:17 |
| frickler | if it is passing and you want to avoid regressions you could explicitly make it voting for neutron | 13:17 |
| opendevreview | Lajos Katona proposed openstack/networking-bgpvpn master: Update jobs based on testing runtime for 2024.2 https://review.opendev.org/c/openstack/networking-bgpvpn/+/914273 | 13:36 |
| f0o | What's the correct way of doing Self-Service IPv6 in Neutron? I saw https://docs.openstack.org/neutron/latest/admin/config-ipv6.html but it states many experimental/deprecated warning blobs heh | 13:47 |
| frickler | f0o: if you intend to have public IPv6, I suggest to look at https://docs.openstack.org/neutron-dynamic-routing/latest/install/usecase-ipv6.html | 13:53 |
| frickler | I also recommend to stay away from prefix delegation | 13:54 |
| f0o | yeah PD seems a lot of headache | 13:55 |
| frickler | and I recommend to have a consistent MTU of 1500 for tenant networks by increasing the MTU on the underlay accordingly | 13:55 |
| f0o | frickler: BGP is nice and all but in this case the network nodes already have full tables as they're edge routers. BGP to localhost seems a bit excessive here. Is there a way to get the subnetpooling done without bgp advertisements? | 13:55 |
| frickler | f0o: full table on network nodes, that's an interesting concept, but I don't see how it is supposed to work since the neutron networking is usually separated from routing on the host | 13:57 |
| jrosser | the tenant subnets sit behind neutron routers that have an address on the public network, the routes are needed to stitch those two thing together | 13:58 |
| f0o | heya jrosser :D | 14:00 |
| jrosser | hi :) what frickler describes is exactly how i have v6 stuff setup fwiw | 14:00 |
| f0o | frickler: yeah I noticed as well as OVS having a long standing bug with full-tables which we (well OVS devs and my lab-bunny-self) resolved and pushed upstream with next release - so now neutron (read OVN Gateway node) can have full bgp tables and VRFs set up so they can coexist on Edge/TOR Routers | 14:01 |
| frickler | oh, another happy customer, nice ;) | 14:01 |
| f0o | But alright this just shows that my current architecture, while working fine for v4 and vlan-typed tenant networks, wont work for SelfService Geneve networks because of the missing glue-stack of OVS<>RAs | 14:03 |
| f0o | so if I can write a neutron driver that teaches my router to enable/start RAs on gevene ports from OVS this would then simulate what is happening on vlan-typed tenant networks | 14:03 |
| * f0o dreams | 14:03 | |
| f0o | to avoid collision I could use the VNI as subnet index and calculate the right /64 even | 14:04 |
| f0o | Hrm... ALl things considered, PD should still be working no? I mean dibbler might be deprecated but ISC Kea supports PD too, however I cant see neutron requesting any PD from the kea logs | 14:30 |
| f0o | but I might be missing something | 14:30 |
| f0o | such as, does neutron actually rely on dibbler-client | 14:31 |
| opendevreview | Brian Haley proposed openstack/neutron master: Make common Metadata Proxy classes https://review.opendev.org/c/openstack/neutron/+/914526 | 14:43 |
| f0o | actually cant find anything on my network-node's logs... and kea dbug doesnt show any requests either... but all this is on a ovs network iface which behaves differently from a real one.. so many failure points | 14:43 |
| opendevreview | Brian Haley proposed openstack/neutron master: Make common Metadata Proxy classes https://review.opendev.org/c/openstack/neutron/+/914526 | 14:50 |
| f0o | oh well... frickler the doc you linked me made me realize something; I honestly do not need dualstack. 1 v4 and 1 v6 interface is absolutely OK. I've been looking at this from an oldschool sysadmin view where I shouldnt. The interfaces are faux anyway so it simply does not matter if I have 2 to get v4+v6 connectivity. | 16:02 |
| f0o | That said we're just going to dump a /64 on a shared network that's v6 only and let tenants add ports to it | 16:03 |
| f0o | Security Groups are still applied, FWaaS is not used, so no biggie | 16:03 |
| opendevreview | Brian Haley proposed openstack/neutron master: Make common Metadata Proxy classes https://review.opendev.org/c/openstack/neutron/+/914526 | 17:52 |
| opendevreview | Terry Wilson proposed openstack/ovsdbapp stable/2024.1: ovn-sb: remove chassis from Chassis_Private on ChassisDelCommand https://review.opendev.org/c/openstack/ovsdbapp/+/914455 | 21:09 |
| opendevreview | Terry Wilson proposed openstack/ovsdbapp stable/2023.2: ovn-sb: remove chassis from Chassis_Private on ChassisDelCommand https://review.opendev.org/c/openstack/ovsdbapp/+/914456 | 21:10 |
| opendevreview | Terry Wilson proposed openstack/ovsdbapp stable/2023.1: ovn-sb: remove chassis from Chassis_Private on ChassisDelCommand https://review.opendev.org/c/openstack/ovsdbapp/+/914457 | 21:11 |
| opendevreview | Terry Wilson proposed openstack/ovsdbapp stable/zed: ovn-sb: remove chassis from Chassis_Private on ChassisDelCommand https://review.opendev.org/c/openstack/ovsdbapp/+/914458 | 21:11 |
| opendevreview | Merged openstack/neutron master: Enable HA for OVN router flavors https://review.opendev.org/c/openstack/neutron/+/901513 | 21:31 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!