opendevreview | Merged openstack/neutron stable/wallaby: Replace "get_routing_table" with "list_ip_routes" https://review.opendev.org/c/openstack/neutron/+/868087 | 00:33 |
---|---|---|
opendevreview | norman shen proposed openstack/neutron master: Filter out unsatisfied routers in SQL https://review.opendev.org/c/openstack/neutron/+/842141 | 00:37 |
opendevreview | Dmitrii Shcherbakov proposed openstack/neutron master: Allow Multiple External Gateways https://review.opendev.org/c/openstack/neutron/+/873593 | 00:40 |
opendevreview | Dmitrii Shcherbakov proposed openstack/neutron master: Allow Multiple External Gateways https://review.opendev.org/c/openstack/neutron/+/873593 | 00:44 |
opendevreview | norman shen proposed openstack/neutron master: Filter out unsatisfied routers in SQL https://review.opendev.org/c/openstack/neutron/+/842141 | 01:20 |
opendevreview | Merged openstack/neutron-fwaas master: Make tempest-multinode job voting https://review.opendev.org/c/openstack/neutron-fwaas/+/873088 | 02:25 |
opendevreview | yatin proposed openstack/neutron master: [DNM] Check issue and fix with ovn source installs https://review.opendev.org/c/openstack/neutron/+/874112 | 04:11 |
opendevreview | yatin proposed openstack/neutron master: [DNM] Check issue and fix with ovn source installs https://review.opendev.org/c/openstack/neutron/+/874112 | 07:16 |
opendevreview | Luis Tomas Bolivar proposed openstack/ovn-octavia-provider master: Ensure HM also apply to FIPs associated to LB VIPs https://review.opendev.org/c/openstack/ovn-octavia-provider/+/873860 | 07:21 |
opendevreview | Luis Tomas Bolivar proposed openstack/ovn-octavia-provider master: Ensure HM also apply to FIPs associated to LB VIPs https://review.opendev.org/c/openstack/ovn-octavia-provider/+/873860 | 07:23 |
opendevreview | Frode Nordahl proposed openstack/neutron master: Allow Multiple External Gateways https://review.opendev.org/c/openstack/neutron/+/873593 | 08:01 |
opendevreview | Frode Nordahl proposed openstack/neutron master: [ovn] Add helper for retrieving LR associated with LRP https://review.opendev.org/c/openstack/neutron/+/873698 | 08:01 |
opendevreview | Frode Nordahl proposed openstack/neutron master: [ovn] Apply soft anti-affinity for LRs with multiple LRPs when scheduling https://review.opendev.org/c/openstack/neutron/+/873699 | 08:01 |
opendevreview | Frode Nordahl proposed openstack/neutron master: [ovn] Implement support for external-gateway-multihoming extension https://review.opendev.org/c/openstack/neutron/+/874199 | 08:01 |
opendevreview | Rodolfo Alonso proposed openstack/neutron master: Improve "sync_ha_chassis_group" method https://review.opendev.org/c/openstack/neutron/+/872023 | 08:46 |
opendevreview | Merged openstack/neutron master: [ovn-ipv6] Temporary exclude ``TestNetworkBasicOps`` class https://review.opendev.org/c/openstack/neutron/+/874093 | 08:50 |
opendevreview | Rodolfo Alonso proposed openstack/ovsdbapp master: Add Interface paramteres to ``OvsdbIdl.add_port`` method. https://review.opendev.org/c/openstack/ovsdbapp/+/873566 | 08:55 |
opendevreview | Bernard Cafarelli proposed openstack/neutron stable/xena: Do not ignore attributes in bulk port create https://review.opendev.org/c/openstack/neutron/+/874130 | 09:18 |
opendevreview | Bernard Cafarelli proposed openstack/neutron stable/wallaby: Do not ignore attributes in bulk port create https://review.opendev.org/c/openstack/neutron/+/874131 | 09:20 |
opendevreview | Luis Tomas Bolivar proposed openstack/ovn-octavia-provider master: Ensure HM also apply to FIPs associated to LB VIPs https://review.opendev.org/c/openstack/ovn-octavia-provider/+/873860 | 10:50 |
opendevreview | Rodolfo Alonso proposed openstack/neutron master: Improve "sync_ha_chassis_group" method https://review.opendev.org/c/openstack/neutron/+/872023 | 11:34 |
opendevreview | Rodolfo Alonso proposed openstack/neutron master: Improve "sync_ha_chassis_group" method https://review.opendev.org/c/openstack/neutron/+/872023 | 11:35 |
*** dasm|off is now known as dasm | 14:11 | |
opendevreview | Bence Romsics proposed openstack/neutron-lib master: port-hint-ovs-tx-steering: api definition https://review.opendev.org/c/openstack/neutron-lib/+/873112 | 14:22 |
opendevreview | Bence Romsics proposed openstack/neutron master: port-hint-ovs-tx-steering: agent side https://review.opendev.org/c/openstack/neutron/+/872905 | 14:26 |
opendevreview | Bence Romsics proposed openstack/neutron master: port-hint-ovs-tx-steering: shim extension https://review.opendev.org/c/openstack/neutron/+/873113 | 14:26 |
opendevreview | Bence Romsics proposed openstack/neutron master: DNM debug logs and dev helper scripts https://review.opendev.org/c/openstack/neutron/+/872906 | 14:27 |
opendevreview | Merged openstack/ovn-octavia-provider master: Avoid use of ovn metadata port IP for HM checks https://review.opendev.org/c/openstack/ovn-octavia-provider/+/873426 | 14:31 |
opendevreview | Merged openstack/neutron stable/yoga: Do not ignore attributes in bulk port create https://review.opendev.org/c/openstack/neutron/+/873963 | 14:34 |
opendevreview | Rodolfo Alonso proposed openstack/neutron master: Stop "neutron-ovn-tempest-ipv6-only-ovs-release" CI job https://review.opendev.org/c/openstack/neutron/+/874253 | 14:38 |
opendevreview | Rodolfo Alonso proposed openstack/neutron master: Improve "sync_ha_chassis_group" method https://review.opendev.org/c/openstack/neutron/+/872023 | 15:07 |
opendevreview | yatin proposed openstack/neutron master: [DNM] Check issue and fix with ovn source installs https://review.opendev.org/c/openstack/neutron/+/874112 | 15:50 |
opendevreview | Lajos Katona proposed openstack/tap-as-a-service master: Doc: add documentation for usage and flow examples for OVS https://review.opendev.org/c/openstack/tap-as-a-service/+/828382 | 15:57 |
haleyb | ralonsoh: have a minute for a bug discussion? | 15:57 |
ralonsoh | haleyb, sure | 15:58 |
haleyb | So I had triaged this bug, https://bugs.launchpad.net/neutron/+bug/1988069 and at first seemed like a kernel issue, but yesterday confirmed it was worse | 15:59 |
haleyb | so i started a patch to test some things, https://review.opendev.org/c/openstack/neutron/+/874167 | 15:59 |
haleyb | in short, setting a small mtu on a network can kill the dhcp agent | 16:00 |
haleyb | without changing the API to detect the call with cause an issue, we're left with working around it in init_l3() | 16:01 |
ralonsoh | haleyb, min MTU for ipv4 is 576 | 16:01 |
ralonsoh | https://en.wikipedia.org/wiki/Maximum_transmission_unit | 16:01 |
ralonsoh | so --mtu 70 is not a valid value | 16:01 |
ralonsoh | not because of Neutron but because of tcp | 16:02 |
haleyb | right, but IPv6 is 1280, and the kernel is not very forgiving when you drop below and try and configure an IPV6 address | 16:02 |
haleyb | ralonsoh: look at some of the later comments, 1279 is the unfortunate mtu value that causes issues | 16:02 |
ralonsoh | but that is expected | 16:03 |
ralonsoh | not TCP, but IP | 16:03 |
ralonsoh | The Internet Protocol requires that hosts must be able to process IP datagrams of at least 576 bytes (for IPv4) or 1280 bytes (for IPv6). | 16:03 |
haleyb | but it causes the dhcp-agent to go into a loop, affecting all tenants | 16:03 |
ralonsoh | sorry but I don't see any issue there | 16:05 |
haleyb | ralonsoh: on a devstack, change the mtu of the private network to 1279 while watching the dhcp-agent log, it will fail adding it's ipv6 global address and loop forever trying | 16:06 |
haleyb | at first i thought it was user error, but we have to stop it from affectinng other tenants | 16:06 |
ralonsoh | maybe we can check the MTU value in the DHCP agent and log an error (and don't spawn a dnsmasq process) | 16:07 |
ralonsoh | but this is a misconfiguration in first place | 16:07 |
haleyb | i understand that, and i would love to change the API to complain but we know we cant. in my WIP patch i did detect the mtu and work around it, i can continue down that path as the fix | 16:08 |
ralonsoh | we can go this way, including IPv4 max mtu too | 16:09 |
ralonsoh | and logging an error | 16:09 |
haleyb | i.e. if network has an ipv6 subnet don't allow a small mtu | 16:09 |
ralonsoh | or a warning | 16:09 |
haleyb | ralonsoh: so maybe that is my question - we can fix the agent, and/or change the API to return an error (httpconflict?) | 16:11 |
haleyb | i always thought changing the API like this breaks our contract? | 16:11 |
ralonsoh | we can log this warning in the agent for now | 16:11 |
opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider stable/zed: Avoid use of ovn metadata port IP for HM checks https://review.opendev.org/c/openstack/ovn-octavia-provider/+/874270 | 16:12 |
ralonsoh | and we can check, when adding subnets to a network, if the MTU is big enough | 16:12 |
ralonsoh | if not, raise an exception and don't allow it | 16:12 |
opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider stable/yoga: Avoid use of ovn metadata port IP for HM checks https://review.opendev.org/c/openstack/ovn-octavia-provider/+/874271 | 16:12 |
opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider stable/xena: Avoid use of ovn metadata port IP for HM checks https://review.opendev.org/c/openstack/ovn-octavia-provider/+/874272 | 16:13 |
haleyb | or when changing the mtu too... ok | 16:13 |
ralonsoh | exactly | 16:13 |
haleyb | so what error should the API raise? in your opinion. like i said i was thinking an http conflict (is that 409?) | 16:14 |
opendevreview | Fernando Royo proposed openstack/ovn-octavia-provider stable/wallaby: Avoid use of ovn metadata port IP for HM checks https://review.opendev.org/c/openstack/ovn-octavia-provider/+/874273 | 16:15 |
ralonsoh | yes, maybe this is the best error code | 16:15 |
haleyb | ralonsoh: ack, i'll do that in a separate patch as i did with the doc warning. thanks for the discussion, didn't want to spend a lot of time if there was a better way | 16:18 |
ralonsoh | haleyb, thanks! | 16:19 |
* haleyb knows what he's doing the next few days at least :) | 16:19 | |
opendevreview | Brian Haley proposed openstack/neutron master: Move tox-py311 job override into job-templates.yaml https://review.opendev.org/c/openstack/neutron/+/873615 | 16:34 |
mnasiadka | Is there any second core willing to review https://review.opendev.org/c/openstack/neutron/+/855668 ? That bug is a bit irritating :-) | 17:07 |
ralonsoh | mnasiadka, but why don't you add this parameter to the config? | 17:11 |
mnasiadka | ralonsoh: well, I've been doing that on many openstack envs that we manage, but it's irritating to be forced to enable qos just for the db-sync script to do it's job | 17:13 |
opendevreview | Rodolfo Alonso proposed openstack/neutron stable/zed: Enable qos extension_driver in ovn_db_sync https://review.opendev.org/c/openstack/neutron/+/874142 | 17:15 |
opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: Enable qos extension_driver in ovn_db_sync https://review.opendev.org/c/openstack/neutron/+/874143 | 17:15 |
opendevreview | Rodolfo Alonso proposed openstack/neutron stable/xena: Enable qos extension_driver in ovn_db_sync https://review.opendev.org/c/openstack/neutron/+/874144 | 17:16 |
opendevreview | Rodolfo Alonso proposed openstack/neutron stable/wallaby: Enable qos extension_driver in ovn_db_sync https://review.opendev.org/c/openstack/neutron/+/874145 | 17:16 |
opendevreview | Merged openstack/networking-bgpvpn master: Imported Translations from Zanata https://review.opendev.org/c/openstack/networking-bgpvpn/+/858008 | 17:35 |
opendevreview | Merged openstack/neutron stable/zed: Improve scheduling L3/DHCP agents, missing lower binding indexes https://review.opendev.org/c/openstack/neutron/+/873619 | 20:52 |
opendevreview | Merged openstack/neutron stable/yoga: Improve scheduling L3/DHCP agents, missing lower binding indexes https://review.opendev.org/c/openstack/neutron/+/873622 | 20:53 |
opendevreview | Merged openstack/neutron stable/xena: Improve scheduling L3/DHCP agents, missing lower binding indexes https://review.opendev.org/c/openstack/neutron/+/873626 | 20:53 |
opendevreview | Merged openstack/neutron master: Enable qos extension_driver in ovn_db_sync https://review.opendev.org/c/openstack/neutron/+/855668 | 22:12 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!