| opendevreview | liuyulong proposed openstack/neutron-tempest-plugin master: Re-enable the router test_migration cases https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/936318 | 01:16 |
|---|---|---|
| opendevreview | liuyulong proposed openstack/neutron master: Remove port resource change after unplug https://review.opendev.org/c/openstack/neutron/+/936235 | 01:17 |
| opendevreview | liuyulong proposed openstack/neutron master: Add basical functionalities for metadata path extension https://review.opendev.org/c/openstack/neutron/+/881535 | 01:25 |
| opendevreview | liuyulong proposed openstack/neutron master: Add metadata path extension openflows https://review.opendev.org/c/openstack/neutron/+/888097 | 01:25 |
| opendevreview | liuyulong proposed openstack/neutron master: Fullstack case for metadata path https://review.opendev.org/c/openstack/neutron/+/888098 | 01:25 |
| opendevreview | liuyulong proposed openstack/neutron master: Add devstack plugin to enable ovs metadata_path https://review.opendev.org/c/openstack/neutron/+/928586 | 01:25 |
| opendevreview | Jakub Libosvar proposed openstack/ovn-bgp-agent master: WIP: Introduce multinode tempest job https://review.opendev.org/c/openstack/ovn-bgp-agent/+/936968 | 01:32 |
| opendevreview | liuyulong proposed openstack/neutron master: Add meter bandwidth limit support https://review.opendev.org/c/openstack/neutron/+/860767 | 01:57 |
| opendevreview | Ihar Hrachyshka proposed openstack/neutron master: Add option to configure live migration activation strategy for OVN https://review.opendev.org/c/openstack/neutron/+/938106 | 02:22 |
| opendevreview | yatin proposed openstack/neutron master: [DNM] wsgi switch higher workers eventlet https://review.opendev.org/c/openstack/neutron/+/937367 | 06:19 |
| opendevreview | Eduardo Olivares proposed x/whitebox-neutron-tempest-plugin master: Use run_on_master_controller to run command when possible https://review.opendev.org/c/x/whitebox-neutron-tempest-plugin/+/938050 | 06:45 |
| ralonsoh | slaweq, hello! if you have a couple of mins, I finished the UTs: https://review.opendev.org/c/openstack/neutron/+/900236 | 07:50 |
| ralonsoh | thanks! | 07:50 |
| amorin | hey ralonsoh | 08:02 |
| ralonsoh | amorin, hello | 08:05 |
| amorin | i was asking about remote cache because, we can see a lot of sync_routers RPC calls coming from our L3 agent | 08:05 |
| amorin | and it seems that, from time to time, it could be heavy on database | 08:06 |
| amorin | and in some situation, we see similar sync_routers coming from a lot of l3 agents | 08:07 |
| ralonsoh | let me check that code | 08:08 |
| amorin | l3 agent side is coming from here: https://opendev.org/openstack/neutron/src/branch/master/neutron/agent/l3/agent.py#L131 | 08:09 |
| amorin | ending here on rpc server side: https://opendev.org/openstack/neutron/src/branch/master/neutron/api/rpc/handlers/l3_rpc.py#L114 | 08:09 |
| amorin | then going through _ensure_host_set_on_ports | 08:10 |
| amorin | then _ensure_host_set_on_port | 08:10 |
| ralonsoh | that happens in two situations: when a new router event happens (_process_router_update) and when doing a full sync | 08:10 |
| amorin | true | 08:10 |
| amorin | full sync is "expected" lets say. _process_router_update depends on customer action | 08:10 |
| amorin | any route added to a router | 08:11 |
| amorin | trigger a sync router | 08:11 |
| amorin | not only on the affected router, but also on sides routers | 08:11 |
| ralonsoh | but you are expecting "_process_router_update" to have some kind of cache | 08:11 |
| amorin | (we have using DVR btw) | 08:11 |
| ralonsoh | "_process_router_update" triggers a full sync? | 08:12 |
| amorin | yes, I am asking if I am alone in that situation or if you remember having issues like this | 08:12 |
| ralonsoh | when that happens? | 08:12 |
| amorin | it's triggering stuff like this: | 08:12 |
| amorin | Queued a router update for 5d97562c-9a47-4154-b976-8702fe93e9f7 (related router 26630892-3ea8-4ffc-8009-d1219ed71a8a) | 08:13 |
| amorin | an update on 26630892-3ea8-4ffc-8009-d1219ed71a8a | 08:13 |
| amorin | can trigger an update on A LOT of other routers | 08:13 |
| amorin | then doing a sync_router for all of them | 08:14 |
| amorin | and because we can have customer with a pretty big private network, splitted on a lot of computes... | 08:15 |
| ralonsoh | in the OVS agent there is an implementation of a cached RPC | 08:16 |
| ralonsoh | OVSPluginApi | 08:16 |
| ralonsoh | inherits from CacheBackedPluginApi | 08:16 |
| ralonsoh | this RPC class is listening to some resource events (ports, SGs, etc) | 08:17 |
| ralonsoh | and caching the resource values | 08:17 |
| amorin | should I take this as a starting point to implement in the l3? | 08:17 |
| ralonsoh | so when the OVS agent is making an RPC call to retrieve some resource, if this resource is caches, the value is returned form the local agent cache | 08:18 |
| ralonsoh | and is not issues to the API | 08:18 |
| ralonsoh | that could be an improvement, for sure | 08:18 |
| ralonsoh | I would make this implementation configurable, if possible | 08:18 |
| ralonsoh | but I think you can also use some kind of cache, that was discussed in a PTG (I don't remember which one) | 08:19 |
| amorin | like building a local {} in l3-agent and store values there for an amount of time | 08:19 |
| ralonsoh | slaweq, ^ do you recall something? could be possible to use olso.cache? | 08:19 |
| ralonsoh | amorin, that's the goal of CacheBackedPluginApi | 08:20 |
| ralonsoh | it will listen (as now) to any resource update (for example you can limit that to routers in a POC) | 08:20 |
| ralonsoh | and any router event (create, update, delete) will update the local cache | 08:20 |
| amorin | that sounds what I need | 08:21 |
| ralonsoh | then no API calls are done to the server if the local cache has this resource | 08:21 |
| ralonsoh | investigate the OVS agent and how it works | 08:21 |
| amorin | I'll try a poc on my side | 08:22 |
| opendevreview | liuyulong proposed openstack/neutron master: Add meter bandwidth limit support https://review.opendev.org/c/openstack/neutron/+/860767 | 09:00 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] QoS max and min rules should be defined in LSP for phynet ports https://review.opendev.org/c/openstack/neutron/+/934418 | 09:31 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Use the Neutron port ID in ``UpdateLSwitchPortQosOptionsCommand`` https://review.opendev.org/c/openstack/neutron/+/937682 | 09:31 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] QoS max and min rules should be defined in LSP for phynet ports https://review.opendev.org/c/openstack/neutron/+/934418 | 10:04 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Use the Neutron port ID in ``UpdateLSwitchPortQosOptionsCommand`` https://review.opendev.org/c/openstack/neutron/+/937682 | 10:04 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Improve ``_ovn_qos_rule`` method, removing unnecesary argument https://review.opendev.org/c/openstack/neutron/+/938120 | 10:25 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Improve ``_ovn_qos_rule`` method, removing unnecesary argument https://review.opendev.org/c/openstack/neutron/+/938120 | 10:26 |
| opendevreview | Slawek Kaplonski proposed openstack/neutron master: QinQ implementation for the ML2/OVN backend https://review.opendev.org/c/openstack/neutron/+/937633 | 10:44 |
| opendevreview | Slawek Kaplonski proposed openstack/neutron master: QinQ implementation for the ML2/OVN backend https://review.opendev.org/c/openstack/neutron/+/937633 | 10:46 |
| opendevreview | Slawek Kaplonski proposed openstack/neutron master: Add .mypy_cache directory to the gitignore file https://review.opendev.org/c/openstack/neutron/+/938123 | 10:55 |
| slaweq | ralonsoh hi, please check ^^ when You will have time | 10:56 |
| slaweq | the reason why I proposed that small patch you can see in https://review.opendev.org/c/openstack/neutron/+/937633/7 :D | 10:56 |
| ralonsoh | slaweq, sure | 10:57 |
| opendevreview | Slawek Kaplonski proposed openstack/neutron master: Add limit of tags for every resource https://review.opendev.org/c/openstack/neutron/+/937887 | 11:19 |
| ralonsoh | slaweq, please check https://review.opendev.org/c/openstack/neutron/+/900236 | 11:33 |
| ralonsoh | once merged, I'll propose the backports | 11:33 |
| slaweq | ralonsoh approved | 11:35 |
| ralonsoh | thanks! | 11:35 |
| slaweq | thx for additional test coverage | 11:35 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2024.2: Allow network owner reader to get subnets https://review.opendev.org/c/openstack/neutron/+/937966 | 11:36 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2024.1: Allow network owner reader to get subnets https://review.opendev.org/c/openstack/neutron/+/937967 | 12:01 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2023.2: Allow network owner reader to get subnets https://review.opendev.org/c/openstack/neutron/+/937968 | 12:02 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Isolate test_ovn_db_sync.TestOvnNbSync.* FTs https://review.opendev.org/c/openstack/neutron/+/937992 | 12:07 |
| andreykurilin | Hi folks! Can someone point me the place in neutron’s code that sends RMQ notifications described in https://docs.openstack.org/neutron/2024.1/admin/archives/adv-operational-features.html ? | 13:02 |
| opendevreview | Pavlo Shchelokovskyy proposed openstack/neutron-lib master: Fix network-ip-availability API samples https://review.opendev.org/c/openstack/neutron-lib/+/938132 | 14:46 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: WIP == [OVN] Assign HA_Chassis_Group to all LRPs for tunnelled gateway https://review.opendev.org/c/openstack/neutron/+/938134 | 15:13 |
| opendevreview | Slawek Kaplonski proposed openstack/neutron master: WIP Make API policies for tags to be working with resource attributes https://review.opendev.org/c/openstack/neutron/+/938135 | 15:15 |
| ralonsoh | slaweq, please check https://review.opendev.org/c/openstack/neutron/+/937992 | 15:19 |
| ralonsoh | because no one else is here right now, if that is ok for you, I'll merge it | 15:20 |
| ralonsoh | to avoid the recurrent issues in the CI FTs we have | 15:20 |
| opendevreview | Rodolfo Alonso proposed openstack/neutron master: WIP == [OVN] Assign HA_Chassis_Group to all LRPs for tunnelled gateway https://review.opendev.org/c/openstack/neutron/+/938134 | 16:27 |
| opendevreview | Ihar Hrachyshka proposed openstack/neutron master: Add option to configure live migration activation strategy for OVN https://review.opendev.org/c/openstack/neutron/+/938106 | 16:42 |
| opendevreview | Jakub Libosvar proposed openstack/ovn-bgp-agent master: WIP: Introduce multinode tempest job https://review.opendev.org/c/openstack/ovn-bgp-agent/+/936968 | 17:40 |
| opendevreview | Merged openstack/neutron master: [OVN] Isolate test_ovn_db_sync.TestOvnNbSync.* FTs https://review.opendev.org/c/openstack/neutron/+/937992 | 18:41 |
| opendevreview | Jakub Libosvar proposed openstack/ovn-bgp-agent master: WIP: Introduce multinode tempest job https://review.opendev.org/c/openstack/ovn-bgp-agent/+/936968 | 19:42 |
| opendevreview | Jakub Libosvar proposed openstack/ovn-bgp-agent master: WIP: Introduce multinode tempest job https://review.opendev.org/c/openstack/ovn-bgp-agent/+/936968 | 20:50 |
| opendevreview | Jakub Libosvar proposed openstack/ovn-bgp-agent master: WIP: Introduce multinode tempest job https://review.opendev.org/c/openstack/ovn-bgp-agent/+/936968 | 21:12 |
| opendevreview | Jakub Libosvar proposed openstack/ovn-bgp-agent master: WIP: Introduce multinode tempest job https://review.opendev.org/c/openstack/ovn-bgp-agent/+/936968 | 22:44 |
| opendevreview | Jakub Libosvar proposed openstack/ovn-bgp-agent master: WIP: Introduce multinode tempest job https://review.opendev.org/c/openstack/ovn-bgp-agent/+/936968 | 23:50 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!