| opendevreview | huanghailun proposed openstack/neutron master: Add allowed_address_pairs add/remove atomic operations https://review.opendev.org/c/openstack/neutron/+/901895 | 02:46 |
|---|---|---|
| opendevreview | Miro Tomaska proposed openstack/neutron master: DHCP agent get ovn dhcp port from server https://review.opendev.org/c/openstack/neutron/+/903572 | 02:49 |
| opendevreview | huanghailun proposed openstack/neutron master: Add allowed_address_pairs add/remove atomic operations https://review.opendev.org/c/openstack/neutron/+/901895 | 07:00 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: [DNM] Testing error in "neutron-functional-with-uwsgi-fips" stable/yoga https://review.opendev.org/c/openstack/neutron/+/903592 | 08:25 |
| opendevreview | Merged openstack/ovn-octavia-provider master: Remove unnecessary ceilometer service overrides https://review.opendev.org/c/openstack/ovn-octavia-provider/+/902554 | 09:18 |
| opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/2023.1: [OVN] Enhanced external port scheduling https://review.opendev.org/c/openstack/neutron/+/903510 | 09:31 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: [DNM] Testing error in "neutron-functional-with-uwsgi-fips" stable/yoga https://review.opendev.org/c/openstack/neutron/+/903592 | 09:48 |
| opendevreview | huanghailun proposed openstack/neutron master: Add allowed_address_pairs add/remove atomic operations https://review.opendev.org/c/openstack/neutron/+/901895 | 09:48 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: [DNM] Testing error in "neutron-functional-with-uwsgi-fips" stable/yoga https://review.opendev.org/c/openstack/neutron/+/903592 | 11:04 |
| opendevreview | Lucas Alvares Gomes proposed openstack/neutron master: Fix IGMP inconsistency across drivers https://review.opendev.org/c/openstack/neutron/+/901753 | 11:48 |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent master: Add documentation about NB DB driver https://review.opendev.org/c/openstack/ovn-bgp-agent/+/903407 | 12:00 |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent master: Fix watcher logic for exposing OVN LB with t he NB DB driver https://review.opendev.org/c/openstack/ovn-bgp-agent/+/903408 | 12:02 |
| Continuity | Is anyone aware of anyone working on OVN+DVR and VLAN tenant networks. | 12:05 |
| Continuity | https://bugs.launchpad.net/neutron/+bug/2035332 | 12:05 |
| Continuity | We have found a workaround of sorts listed in the bug above, but would be good to understand if anyone is looking at OVN+DVR and VLAN networks | 12:05 |
| frickler | Continuity: seems that that bug was missed by the usual bug triaging. ralonsoh maybe you can take a look? | 12:18 |
| ralonsoh | I was doing that now | 12:18 |
| ralonsoh | Continuity, please check https://review.opendev.org/c/openstack/neutron/+/871252 | 12:18 |
| Continuity | :D | 12:18 |
| Continuity | cheers | 12:18 |
| frickler | jinx ;) | 12:18 |
| Continuity | So we have that patched deployed, we still end up in the same situation as explained in the bug i linked | 12:19 |
| ralonsoh | Continuity, if you check the commit message, the 3rd bullet point | 12:20 |
| ralonsoh | "reside-on-redirect-chassis to False, but with ovn-chassis-mac-mappings | 12:20 |
| ralonsoh | configured" | 12:20 |
| ralonsoh | this is for VLAN networks with DVR | 12:20 |
| ralonsoh | ltomasbo, ^^ some help here? | 12:21 |
| ralonsoh | ah, https://review.opendev.org/c/openstack/neutron/+/878450 should be considered too (but I see this flag in the bug report) | 12:22 |
| Continuity | yeah this is the issue i have had, its a bit of a mess, and hard to work out :D | 12:23 |
| ralonsoh | do you have the mac mappings configured (in each compute node)? | 12:23 |
| Continuity | yes | 12:23 |
| Continuity | i believe so | 12:23 |
| ralonsoh | now it is not possible (with DVR) to mix vlan/flat and tunnelled networks | 12:23 |
| ralonsoh | the configuration combinations are strict on this | 12:23 |
| Continuity | So we have a combo that works, but requires a change after the networks are deployed | 12:24 |
| Continuity | which is explained in the bug i linked above | 12:24 |
| Continuity | i mean it might be working through luck.... :D | 12:24 |
| ralonsoh | Continuity, why do you change the chassis priority? | 12:26 |
| ralonsoh | this is done by Neutron | 12:26 |
| ralonsoh | should not be modified manually | 12:26 |
| ralonsoh | "New routers with vlan networks attached don't create all three ports on the controllers." What does it means (sorry, i don't get it) | 12:27 |
| Continuity | So each of the controllers should have a chassis_gateway created, sometimes (not sure why) those chassis_gateways were not created correctly on all three controllers | 12:28 |
| Continuity | then the next issue we had after that was that the ha-chassis groups didnt match the chassis_gateways | 12:29 |
| Continuity | which is why we had to modify them to get them into the correct order | 12:29 |
| Continuity | the first two issues may have been related to patches that have been fixed since we were troubleshooting this | 12:29 |
| ralonsoh | what patches? | 12:30 |
| ralonsoh | because what you are saying is that the OVN L3 scheduler is maybe not working fine | 12:30 |
| ralonsoh | (and that's a problem for sure) | 12:30 |
| ralonsoh | I would focus first on find why, when you create a LRP, the OVN L3 scheduler is not creatign the corresponding 3 Gateway_Chassis registers | 12:31 |
| Continuity | yes, we have had an issues with the OVN L3 scheduler, but since we had those issues, we have patched, and I remember seeing patches that addressed those issues | 12:32 |
| Continuity | ralonsoh: i will try building some new networks in the next day or so, and test that | 12:32 |
| ralonsoh | what patches? just to know | 12:32 |
| Continuity | er.... :D | 12:32 |
| ralonsoh | if possible, please add extra logs in the code | 12:32 |
| ralonsoh | and print them, please | 12:32 |
| Continuity | will do | 12:32 |
| ralonsoh | that could help us a lot | 12:32 |
| Continuity | if we are hacing issues with the OVN l3 scheduler Ill open another bug report and place that data in there, and split the issues in to smaller bugs, rather than it all being in oen big one. | 12:34 |
| stephenfin | slaweq: ralonsoh: Could one of you sanity check https://review.opendev.org/c/openstack/python-openstackclient/+/903672 when you have time? | 12:34 |
| ralonsoh | Continuity, perfect | 12:34 |
| ralonsoh | stephenfin, let me check | 12:34 |
| Continuity | np | 12:34 |
| stephenfin | ralonsoh: Thanks | 12:43 |
| stephenfin | I also recently rewrote the SG docs for nova. You can find them at https://docs.openstack.org/nova/latest/user/security-groups.html Not remotely urgent but if you have time to take a skim through and sanity check those also it would be helpful to all | 12:43 |
| ykarel | Continuity, your external network is also of vlan type? | 13:10 |
| Continuity | ykarel: yes | 13:11 |
| ykarel | Continuity, thx for confirmation | 13:12 |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent master: Fix watcher logic for exposing OVN LB with t he NB DB driver https://review.opendev.org/c/openstack/ovn-bgp-agent/+/903408 | 13:46 |
| ltomasbo | sorry ralonsoh I was having lunch and then meeting... reading now | 13:49 |
| ltomasbo | Continuity, ralonsoh there was an issue with FIPs at core ovn when redirect-type=bridge was used | 13:53 |
| ralonsoh | ltomasbo, is there a LP bug? because so far, I though that option was working | 13:54 |
| ralonsoh | so is it possible right now to use VLAN + DVR? | 13:54 |
| ltomasbo | Continuity, perhaps related to this one? https://bugzilla.redhat.com/show_bug.cgi?id=2007120 | 13:54 |
| ltomasbo | this was fixed, but pernaps it is not in the ovn version used | 13:55 |
| ralonsoh | okidoki | 13:55 |
| ltomasbo | Continuity, ralonsoh: from current experience, the real distribution of provider tenant vlan networks is with the redirect-type=bridge flag, the reside-on-redirect-chassis is still centralized, the different between true and false is if it goes tunneled or not to the gateway port | 14:01 |
| Continuity | thanks | 14:02 |
| opendevreview | Lucas Alvares Gomes proposed openstack/neutron master: Fix IGMP inconsistency across drivers https://review.opendev.org/c/openstack/neutron/+/901753 | 14:11 |
| *** haleyb|out is now known as haleyb | 14:24 | |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent master: Add documentation about NB DB driver https://review.opendev.org/c/openstack/ovn-bgp-agent/+/903407 | 14:34 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: [stable-only] Workaround in ``TestPortBindingChassisEvent`` https://review.opendev.org/c/openstack/neutron/+/903592 | 14:55 |
| ralonsoh | bcafarel, ^^ if you have time | 15:18 |
| ralonsoh | ykarel, | 15:18 |
| ykarel | ralonsoh, thx, you can drop those Log imports too | 15:21 |
| ralonsoh | ahh yes, sorry | 15:26 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: [stable-only] Workaround in ``TestPortBindingChassisEvent`` https://review.opendev.org/c/openstack/neutron/+/903592 | 15:26 |
| opendevreview | Harald Jensås proposed openstack/neutron stable/2023.2: [OVN] DB sync host/physnet - filter on agent_type https://review.opendev.org/c/openstack/neutron/+/903682 | 15:29 |
| opendevreview | Harald Jensås proposed openstack/neutron stable/2023.1: [OVN] DB sync host/physnet - filter on agent_type https://review.opendev.org/c/openstack/neutron/+/903683 | 15:29 |
| *** jamesdenton_ is now known as jamesdenton | 15:37 | |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent master: Add documentation about NB DB driver https://review.opendev.org/c/openstack/ovn-bgp-agent/+/903407 | 15:56 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: Improve the SG RPC callback ``security_group_info_for_ports`` https://review.opendev.org/c/openstack/neutron/+/903707 | 15:59 |
| opendevreview | Luis Tomas Bolivar proposed openstack/ovn-bgp-agent master: Fix watcher logic for exposing OVN LB with the NB DB driver https://review.opendev.org/c/openstack/ovn-bgp-agent/+/903408 | 15:59 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: Improve the SG RPC callback ``security_group_info_for_ports`` https://review.opendev.org/c/openstack/neutron/+/903707 | 16:00 |
| opendevreview | Merged openstack/ovn-bgp-agent master: Fix watcher logic for exposing OVN LB with the NB DB driver https://review.opendev.org/c/openstack/ovn-bgp-agent/+/903408 | 16:17 |
| opendevreview | Merged openstack/neutron stable/zed: get_hosts_mapped_with_segments add filter agt_type https://review.opendev.org/c/openstack/neutron/+/901790 | 19:16 |
| opendevreview | Harald Jensås proposed openstack/neutron stable/zed: [OVN] DB sync host/physnet - filter on agent_type https://review.opendev.org/c/openstack/neutron/+/903685 | 21:16 |
| *** dmellado2 is now known as dmelladoo | 21:55 | |
| *** dmelladoo is now known as dmellado | 21:58 | |
| opendevreview | Merged openstack/neutron master: [OVN] Fix UT assertion https://review.opendev.org/c/openstack/neutron/+/903584 | 22:31 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!