)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":15197,"name":"Pierre Riteau","email":"pierre@stackhpc.com","username":"priteau","status":"StackHPC"},"change_message_id":"44ec8040264add66d33892dfe51a4d62ad68b187","unresolved":true,"context_lines":[{"line_number":13,"context_line":"Note this does not affect the VRRP based checks that"},{"line_number":14,"context_line":"keepalived is already doing. But it also means no one is"},{"line_number":15,"context_line":"checking the gateway is up. Although its unclear what extra"},{"line_number":16,"context_line":"benifit this would bring."},{"line_number":17,"context_line":""},{"line_number":18,"context_line":"Change-Id: If5d0054ecb3e4d73d9d6e533e71ce44e6b55220d"},{"line_number":19,"context_line":"Signed-off-by: John Garbutt \u003cjohn.garbutt@stackhpc.com\u003e"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"0447ed1c_8569e15b","line":16,"range":{"start_line":16,"start_character":0,"end_line":16,"end_character":7},"updated":"2026-02-06 15:19:37.000000000","message":"benefit","commit_id":"28f74b5557d70829d16ed5edd8398072fc11da01"},{"author":{"_account_id":17669,"name":"Doug Szumski","email":"doug@stackhpc.com","username":"DougSzumski"},"change_message_id":"6e892853c8f3b0fc53060cfe9e8cd71dd8dc79e8","unresolved":true,"context_lines":[{"line_number":13,"context_line":"Note this does not affect the VRRP based checks that"},{"line_number":14,"context_line":"keepalived is already doing. But it also means no one is"},{"line_number":15,"context_line":"checking the gateway is up. Although its unclear what extra"},{"line_number":16,"context_line":"benefit this would bring."},{"line_number":17,"context_line":""},{"line_number":18,"context_line":"Change-Id: If5d0054ecb3e4d73d9d6e533e71ce44e6b55220d"},{"line_number":19,"context_line":"Signed-off-by: John Garbutt \u003cjohn.garbutt@stackhpc.com\u003e"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":2,"id":"b461d6ff_dd5573f0","line":16,"updated":"2026-02-06 17:33:23.000000000","message":"There is some interesting reading on failure scenarios here: https://bugs.launchpad.net/neutron/+bug/1375625","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"}],"/PATCHSET_LEVEL":[{"author":{"_account_id":17669,"name":"Doug Szumski","email":"doug@stackhpc.com","username":"DougSzumski"},"change_message_id":"6e892853c8f3b0fc53060cfe9e8cd71dd8dc79e8","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"9fbd1e87_26522172","updated":"2026-02-06 17:33:23.000000000","message":"The config override has been there for a long time [1]. I\u0027m curious if the instability has arrived in recent Neutron versions. I\u0027m not against setting it back to the default.\n\n[1] https://bugs.launchpad.net/kolla/+bug/1698992","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"},{"author":{"_account_id":17669,"name":"Doug Szumski","email":"doug@stackhpc.com","username":"DougSzumski"},"change_message_id":"2db7daafb5572392c45b8f3e21424f7bd251eb0e","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"9153d3d4_f156ce10","updated":"2026-02-09 09:27:42.000000000","message":"recheck - CI fail is unrelated, possible bug with compute service registration","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"0d3b654383ce27593f231c13070334204375d8c8","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"2f3acf5c_beacbdce","in_reply_to":"083c77f9_d2835f3b","updated":"2026-02-09 14:30:09.000000000","message":"\u003e Small rodent eats the external link on a controller hosting the active router. Router stays active and starts dropping external traffic. It /should/ fail over to another controller, but it doesn\u0027t unless this setting is on to monitor the external gateway? \n\nAh, so that is the case, a separate external network connection.\n\nI am working with sites where there is a single bond for all the traffic, so if we loose the external connection, generally we also loose HA network that uses tunnel net, and VRRP correctly elects a new primary (regardless of our config change). So in a way, maybe this depends on your network setup?\n\nTo be clear, the issue we have is that until ovs plugs the VIF nicely, we get traffic failures, which triggers the failover to another node, which hits the same problem, etc.\n\n\u003e It is trivial to enable / disable this via config overrides. I don\u0027t think we need to \u0027own\u0027 the setting in Kolla, rather it is a question of what is the best behaviour to provide by default.\n\nSo, out the box, the neutron default is 0. The docs say, if you use HA you should add this config we add here. But we have also found it to cause large instability problems if you set this to anything but 0. Hence my suggestion to stop adding this by default.","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"},{"author":{"_account_id":17669,"name":"Doug Szumski","email":"doug@stackhpc.com","username":"DougSzumski"},"change_message_id":"d10c02ac16917878833bc9282613fa88ad97b5f0","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"083c77f9_d2835f3b","in_reply_to":"2ef0a876_be689468","updated":"2026-02-09 14:12:57.000000000","message":"Small rodent eats the external link on a controller hosting the active router. Router stays active and starts dropping external traffic. It /should/ fail over to another controller, but it doesn\u0027t unless this setting is on to monitor the external gateway? \n\nIt is trivial to enable / disable this via config overrides. I don\u0027t think we need to \u0027own\u0027 the setting in Kolla, rather it is a question of what is the best behaviour to provide by default.","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"},{"author":{"_account_id":15197,"name":"Pierre Riteau","email":"pierre@stackhpc.com","username":"priteau","status":"StackHPC"},"change_message_id":"c5c0564aa43bd8247239e6bd925f13268a0390e4","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"6437b75e_b83ebc8a","in_reply_to":"43a3329a_49131915","updated":"2026-02-09 10:30:48.000000000","message":"I wonder if this should still be opt-in? Or allow the old behaviour to be opt-in? Small deployments with few L3 routers could keep using the old setting.","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"c119112122bf19dc5c982dd5b851fc36263ada93","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"2ef0a876_be689468","in_reply_to":"6437b75e_b83ebc8a","updated":"2026-02-09 12:23:02.000000000","message":"Do we know in what specific cases it helps the small deployments? I couldn\u0027t think of a practical example where it helps.","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"},{"author":{"_account_id":15197,"name":"Pierre Riteau","email":"pierre@stackhpc.com","username":"priteau","status":"StackHPC"},"change_message_id":"10628833f82bef4a8af9f1772bbe64864c5ca393","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"ad27b20b_c2f927ff","in_reply_to":"9fbd1e87_26522172","updated":"2026-02-06 20:45:37.000000000","message":"I don\u0027t think this is a new issue at all. At a deployment with hundreds of Neutron HA routers, we bumped the check interval from 5 to 30 seconds in 2021. More recently (in 2024), we disabled the health check completely, as proposed here.","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"},{"author":{"_account_id":17669,"name":"Doug Szumski","email":"doug@stackhpc.com","username":"DougSzumski"},"change_message_id":"2db7daafb5572392c45b8f3e21424f7bd251eb0e","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"43a3329a_49131915","in_reply_to":"ad27b20b_c2f927ff","updated":"2026-02-09 09:27:42.000000000","message":"Thanks, that\u0027s helpful. It seems like a tradeoff worth making.","commit_id":"a590a4048f7d69ca4833cd8cf52c178bc44281c0"}],"releasenotes/notes/diable_ha_vrrp_health_check_interval-c211e930802bd8dd.yaml":[{"author":{"_account_id":15197,"name":"Pierre Riteau","email":"pierre@stackhpc.com","username":"priteau","status":"StackHPC"},"change_message_id":"44ec8040264add66d33892dfe51a4d62ad68b187","unresolved":true,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"upgrade:"},{"line_number":3,"context_line":"  - |"},{"line_number":4,"context_line":"    By default, ml2 ovs l3 routers no longer set"},{"line_number":5,"context_line":"    ``ha_vrrp_health_check_interval`` when"},{"line_number":6,"context_line":"    ``enable_neutron_agent_ha`` is true."},{"line_number":7,"context_line":"    This is to prevent stability issues when"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"ba8068db_8f964a95","line":4,"range":{"start_line":4,"start_character":16,"end_line":4,"end_character":26},"updated":"2026-02-06 15:19:37.000000000","message":"ML2/OVS L3","commit_id":"28f74b5557d70829d16ed5edd8398072fc11da01"},{"author":{"_account_id":15197,"name":"Pierre Riteau","email":"pierre@stackhpc.com","username":"priteau","status":"StackHPC"},"change_message_id":"44ec8040264add66d33892dfe51a4d62ad68b187","unresolved":true,"context_lines":[{"line_number":5,"context_line":"    ``ha_vrrp_health_check_interval`` when"},{"line_number":6,"context_line":"    ``enable_neutron_agent_ha`` is true."},{"line_number":7,"context_line":"    This is to prevent stability issues when"},{"line_number":8,"context_line":"    neutron is unable to keep up with l3 router state changes."}],"source_content_type":"text/x-yaml","patch_set":1,"id":"442c5f24_f26bc35a","line":8,"range":{"start_line":8,"start_character":38,"end_line":8,"end_character":40},"updated":"2026-02-06 15:19:37.000000000","message":"L3","commit_id":"28f74b5557d70829d16ed5edd8398072fc11da01"},{"author":{"_account_id":15197,"name":"Pierre Riteau","email":"pierre@stackhpc.com","username":"priteau","status":"StackHPC"},"change_message_id":"44ec8040264add66d33892dfe51a4d62ad68b187","unresolved":true,"context_lines":[{"line_number":5,"context_line":"    ``ha_vrrp_health_check_interval`` when"},{"line_number":6,"context_line":"    ``enable_neutron_agent_ha`` is true."},{"line_number":7,"context_line":"    This is to prevent stability issues when"},{"line_number":8,"context_line":"    neutron is unable to keep up with l3 router state changes."}],"source_content_type":"text/x-yaml","patch_set":1,"id":"4c7e9d47_7df996e8","line":8,"range":{"start_line":8,"start_character":4,"end_line":8,"end_character":11},"updated":"2026-02-06 15:19:37.000000000","message":"Neutron","commit_id":"28f74b5557d70829d16ed5edd8398072fc11da01"}]}