| *** ociuhandu has joined #openstack-neutron | 00:07 | |
| *** tbachman has joined #openstack-neutron | 00:07 | |
| *** ccamposr__ has quit IRC | 00:11 | |
| *** ccamposr__ has joined #openstack-neutron | 00:12 | |
| *** ociuhandu has quit IRC | 00:14 | |
| openstackgerrit | Merged openstack/neutron master: Check router interface not in use properly https://review.opendev.org/c/openstack/neutron/+/765129 | 00:26 |
|---|---|---|
| openstackgerrit | Merged openstack/neutron stable/ussuri: Fix migration from the HA to non-HA routers https://review.opendev.org/c/openstack/neutron/+/763345 | 00:26 |
| *** zzzeek has quit IRC | 00:30 | |
| *** zzzeek has joined #openstack-neutron | 00:32 | |
| *** yumiriam has joined #openstack-neutron | 00:54 | |
| openstackgerrit | Merged openstack/neutron stable/victoria: Fix calling of add_tunnel_port method from sanity checks module https://review.opendev.org/c/openstack/neutron/+/765283 | 01:10 |
| openstackgerrit | Merged openstack/neutron stable/victoria: Ensure ovsdb_probe_interval set before connect() https://review.opendev.org/c/openstack/neutron/+/765599 | 01:10 |
| openstackgerrit | Merged openstack/neutron master: Rely on worker count for HashRing caching https://review.opendev.org/c/openstack/neutron/+/765874 | 01:15 |
| *** rcernin has joined #openstack-neutron | 01:17 | |
| *** yumiriam has quit IRC | 01:39 | |
| *** macz_ has quit IRC | 01:49 | |
| *** ociuhandu has joined #openstack-neutron | 01:55 | |
| openstackgerrit | Brian Haley proposed openstack/networking-ovn stable/train: Fix gate failure https://review.opendev.org/c/openstack/networking-ovn/+/767424 | 01:57 |
| *** hamalq_ has quit IRC | 01:59 | |
| *** ociuhandu has quit IRC | 02:08 | |
| *** rcernin has quit IRC | 02:26 | |
| *** zzzeek has quit IRC | 02:30 | |
| *** zzzeek has joined #openstack-neutron | 02:31 | |
| *** spatel has joined #openstack-neutron | 02:44 | |
| *** armax has quit IRC | 02:47 | |
| *** tbarron has quit IRC | 02:52 | |
| *** rcernin has joined #openstack-neutron | 02:58 | |
| *** rcernin has quit IRC | 03:10 | |
| openstackgerrit | liuyulong proposed openstack/neutron-specs master: Add spec for Distributed DHCP https://review.opendev.org/c/openstack/neutron-specs/+/768588 | 03:17 |
| openstackgerrit | zhanghao proposed openstack/neutron master: [OVN] Implement router gateway IP QoS https://review.opendev.org/c/openstack/neutron/+/749012 | 03:22 |
| *** zzzeek has quit IRC | 03:43 | |
| *** zzzeek has joined #openstack-neutron | 03:45 | |
| *** macz_ has joined #openstack-neutron | 03:49 | |
| *** ccamposr__ has quit IRC | 03:53 | |
| *** ccamposr__ has joined #openstack-neutron | 03:53 | |
| *** zzzeek has quit IRC | 03:53 | |
| *** macz_ has quit IRC | 03:54 | |
| *** zzzeek has joined #openstack-neutron | 03:55 | |
| *** sapd1 has joined #openstack-neutron | 03:57 | |
| *** zzzeek has quit IRC | 04:23 | |
| *** zzzeek has joined #openstack-neutron | 04:25 | |
| *** ircuser-1 has joined #openstack-neutron | 04:43 | |
| *** ratailor has joined #openstack-neutron | 04:57 | |
| *** ociuhandu has joined #openstack-neutron | 05:32 | |
| *** ociuhandu has quit IRC | 05:38 | |
| *** macz_ has joined #openstack-neutron | 05:51 | |
| *** macz_ has quit IRC | 05:55 | |
| *** lajoskatona has joined #openstack-neutron | 06:02 | |
| *** hemanth_n has joined #openstack-neutron | 06:04 | |
| *** gcheresh_ has joined #openstack-neutron | 06:16 | |
| *** spatel has quit IRC | 06:17 | |
| *** zzzeek has quit IRC | 06:31 | |
| *** zzzeek has joined #openstack-neutron | 06:33 | |
| *** zzzeek has quit IRC | 06:46 | |
| *** zzzeek has joined #openstack-neutron | 06:47 | |
| openstackgerrit | zhanghao proposed openstack/neutron master: [OVN] Implement router gateway IP QoS https://review.opendev.org/c/openstack/neutron/+/749012 | 06:51 |
| openstackgerrit | wdd proposed openstack/ovsdbapp master: update py38 and py36 test https://review.opendev.org/c/openstack/ovsdbapp/+/757991 | 07:08 |
| openstackgerrit | liuyulong proposed openstack/neutron-specs master: Neutron detector resource and agent https://review.opendev.org/c/openstack/neutron-specs/+/662541 | 07:10 |
| *** lpetrut has joined #openstack-neutron | 07:12 | |
| openstackgerrit | liuyulong proposed openstack/neutron-specs master: L3 agent self-service metering https://review.opendev.org/c/openstack/neutron-specs/+/658511 | 07:13 |
| openstackgerrit | liuyulong proposed openstack/neutron-specs master: L3 agent capacity and scheduling https://review.opendev.org/c/openstack/neutron-specs/+/658451 | 07:16 |
| openstackgerrit | Mamduh proposed openstack/os-vif stable/train: Refactor code of linux_net to more cleaner and increase performace https://review.opendev.org/c/openstack/os-vif/+/765912 | 07:17 |
| *** zzzeek has quit IRC | 07:17 | |
| openstackgerrit | Kamil Sambor proposed openstack/neutron master: Enable querier for multicast (IGMP) in OVN https://review.opendev.org/c/openstack/neutron/+/764427 | 07:18 |
| *** zzzeek has joined #openstack-neutron | 07:19 | |
| openstackgerrit | Mamduh proposed openstack/os-vif stable/train: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765970 | 07:25 |
| *** abdysn has joined #openstack-neutron | 07:26 | |
| *** rpittau|afk is now known as rpittau | 07:33 | |
| *** Abdallahyas has joined #openstack-neutron | 07:36 | |
| *** abdysn has quit IRC | 07:40 | |
| openstackgerrit | Mamduh proposed openstack/os-vif stable/stein: Refactor code of linux_net to more cleaner and increase performace https://review.opendev.org/c/openstack/os-vif/+/765914 | 07:42 |
| *** slaweq has joined #openstack-neutron | 07:42 | |
| *** zzzeek has quit IRC | 07:45 | |
| *** zzzeek has joined #openstack-neutron | 07:45 | |
| openstackgerrit | Mamduh proposed openstack/os-vif stable/stein: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765974 | 07:46 |
| *** zzzeek has quit IRC | 07:50 | |
| *** macz_ has joined #openstack-neutron | 07:51 | |
| *** ralonsoh has joined #openstack-neutron | 07:52 | |
| *** zzzeek has joined #openstack-neutron | 07:53 | |
| *** gary_perkins has quit IRC | 07:54 | |
| *** macz_ has quit IRC | 07:56 | |
| *** gary_perkins has joined #openstack-neutron | 07:57 | |
| *** tmorin has joined #openstack-neutron | 07:59 | |
| *** bengates has joined #openstack-neutron | 08:00 | |
| *** ccamposr has joined #openstack-neutron | 08:01 | |
| *** abdysn has joined #openstack-neutron | 08:03 | |
| *** ccamposr__ has quit IRC | 08:04 | |
| *** Abdallahyas has quit IRC | 08:06 | |
| *** bengates has quit IRC | 08:08 | |
| *** bengates has joined #openstack-neutron | 08:09 | |
| *** zzzeek has quit IRC | 08:11 | |
| *** zzzeek has joined #openstack-neutron | 08:11 | |
| *** zzzeek has quit IRC | 08:16 | |
| *** zzzeek has joined #openstack-neutron | 08:18 | |
| *** tesseract has joined #openstack-neutron | 08:18 | |
| *** elvira has joined #openstack-neutron | 08:19 | |
| *** bengates has quit IRC | 08:20 | |
| *** bengates has joined #openstack-neutron | 08:21 | |
| *** Abdallahyas has joined #openstack-neutron | 08:35 | |
| openstackgerrit | Bernard Cafarelli proposed openstack/neutron stable/stein: [stein] Fix tests with new pip resolver https://review.opendev.org/c/openstack/neutron/+/767291 | 08:38 |
| *** abdysn has quit IRC | 08:39 | |
| *** zzzeek has quit IRC | 08:39 | |
| *** zzzeek has joined #openstack-neutron | 08:41 | |
| *** jlibosva has joined #openstack-neutron | 08:48 | |
| *** ociuhandu has joined #openstack-neutron | 08:49 | |
| *** bengates has quit IRC | 08:54 | |
| openstackgerrit | Mamduh proposed openstack/os-vif stable/rocky: Refactor code of linux_net to more cleaner and increase performace https://review.opendev.org/c/openstack/os-vif/+/765962 | 08:55 |
| openstackgerrit | Mamduh proposed openstack/os-vif stable/rocky: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765977 | 08:55 |
| *** bengates has joined #openstack-neutron | 08:55 | |
| *** lucasagomes has joined #openstack-neutron | 09:09 | |
| *** ociuhandu has quit IRC | 09:09 | |
| *** ociuhandu has joined #openstack-neutron | 09:10 | |
| *** zzzeek has quit IRC | 09:11 | |
| *** gcheresh_ has quit IRC | 09:11 | |
| *** gcheresh has joined #openstack-neutron | 09:12 | |
| *** zzzeek has joined #openstack-neutron | 09:13 | |
| *** ociuhandu has quit IRC | 09:16 | |
| *** zzzeek has quit IRC | 09:26 | |
| openstackgerrit | zhangtongjian proposed openstack/neutron-specs master: remove unicode from code https://review.opendev.org/c/openstack/neutron-specs/+/769287 | 09:26 |
| *** zzzeek has joined #openstack-neutron | 09:27 | |
| *** ociuhandu has joined #openstack-neutron | 09:28 | |
| *** ociuhandu has quit IRC | 09:33 | |
| *** ociuhandu has joined #openstack-neutron | 09:36 | |
| *** ociuhandu has quit IRC | 09:42 | |
| *** tkajinam has quit IRC | 09:42 | |
| *** zzzeek has quit IRC | 09:45 | |
| *** zzzeek has joined #openstack-neutron | 09:47 | |
| *** ratailor_ has joined #openstack-neutron | 09:54 | |
| *** ratailor_ has quit IRC | 09:55 | |
| *** ratailor_ has joined #openstack-neutron | 09:57 | |
| *** ratailor has quit IRC | 09:57 | |
| *** ratailor_ has quit IRC | 09:57 | |
| openstackgerrit | Mamduh proposed openstack/os-vif stable/rocky: Refactor code of linux_net to more cleaner and increase performace https://review.opendev.org/c/openstack/os-vif/+/765962 | 09:58 |
| openstackgerrit | Mamduh proposed openstack/os-vif stable/rocky: Fix - os-vif fails to get the correct UpLink Representor https://review.opendev.org/c/openstack/os-vif/+/765977 | 09:58 |
| *** ratailor has joined #openstack-neutron | 09:58 | |
| *** ratailor has quit IRC | 09:59 | |
| *** ratailor has joined #openstack-neutron | 10:02 | |
| *** abdysn has joined #openstack-neutron | 10:06 | |
| *** ociuhandu has joined #openstack-neutron | 10:09 | |
| *** Abdallahyas has quit IRC | 10:09 | |
| openstackgerrit | Jakub Libosvar proposed openstack/neutron master: ovn: Support live migration to DPDK nodes https://review.opendev.org/c/openstack/neutron/+/769306 | 10:15 |
| jlibosva | flaviof: ^^ | 10:16 |
| *** ociuhandu has quit IRC | 10:31 | |
| *** rpittau is now known as rpittau|bbl | 10:32 | |
| openstackgerrit | Lajos Katona proposed openstack/neutron-dynamic-routing master: Add doc/requirements https://review.opendev.org/c/openstack/neutron-dynamic-routing/+/769315 | 10:39 |
| *** ociuhandu has joined #openstack-neutron | 10:40 | |
| *** ratailor has quit IRC | 10:54 | |
| openstackgerrit | Merged openstack/neutron master: Refactor the Neutron OSKenApp class https://review.opendev.org/c/openstack/neutron/+/768052 | 10:55 |
| *** yumiriam has joined #openstack-neutron | 11:00 | |
| ralonsoh | slaweq, https://github.com/openstack/neutron/blob/master/neutron/services/revisions/revision_plugin.py#L249 | 11:03 |
| ralonsoh | I'm wondering, as we talk yesterday, if this is needed | 11:04 |
| ralonsoh | 1) if we are updating an object, this should be in the session | 11:06 |
| *** ccamposr has quit IRC | 11:06 | |
| ralonsoh | 2) if we are just executing a GET, this doesn't need to be in the session (this is when we should retrieve it) | 11:07 |
| ralonsoh | (sorry, I realized you are in a training today) | 11:08 |
| slaweq | ralonsoh: from quick look I think this may be needed e.g. for security groups | 11:10 |
| slaweq | if e.g. You will update SG rule, SG revision should be bumped probably | 11:10 |
| slaweq | and that's why it can be here | 11:10 |
| slaweq | but maybe I'm wrong | 11:11 |
| *** ociuhandu has quit IRC | 11:13 | |
| *** ociuhandu has joined #openstack-neutron | 11:16 | |
| *** ratailor has joined #openstack-neutron | 11:16 | |
| openstackgerrit | Bernard Cafarelli proposed openstack/neutron stable/train: [train] Fix tests with new pip resolver https://review.opendev.org/c/openstack/neutron/+/767287 | 11:21 |
| openstackgerrit | Bernard Cafarelli proposed openstack/neutron stable/train: Re-enable networking-ovn job https://review.opendev.org/c/openstack/neutron/+/769323 | 11:23 |
| bcafarel | 2021 new year resolution: get these pip resolver issues finally sorted out... | 11:24 |
| *** zzzeek has quit IRC | 11:26 | |
| *** zzzeek has joined #openstack-neutron | 11:28 | |
| *** sapd1 has quit IRC | 11:30 | |
| ralonsoh | bcafarel++ | 11:31 |
| *** zzzeek has quit IRC | 11:43 | |
| *** zzzeek has joined #openstack-neutron | 11:43 | |
| *** ociuhandu has quit IRC | 11:50 | |
| openstackgerrit | Lajos Katona proposed openstack/neutron-tempest-plugin master: Add doc/requirements https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/769330 | 11:53 |
| *** yumiriam has quit IRC | 11:58 | |
| *** abdysn has quit IRC | 12:03 | |
| *** Abdallahyas has joined #openstack-neutron | 12:03 | |
| openstackgerrit | Merged openstack/neutron stable/ussuri: Fix calling of add_tunnel_port method from sanity checks module https://review.opendev.org/c/openstack/neutron/+/765284 | 12:05 |
| openstackgerrit | Merged openstack/neutron stable/ussuri: Ensure ovsdb_probe_interval set before connect() https://review.opendev.org/c/openstack/neutron/+/765628 | 12:06 |
| *** ratailor has quit IRC | 12:15 | |
| *** zzzeek has quit IRC | 12:20 | |
| *** ociuhandu has joined #openstack-neutron | 12:20 | |
| *** abdysn has joined #openstack-neutron | 12:24 | |
| *** zzzeek has joined #openstack-neutron | 12:24 | |
| *** Abdallahyas has quit IRC | 12:25 | |
| *** ociuhandu has quit IRC | 12:30 | |
| *** tbarron has joined #openstack-neutron | 12:32 | |
| *** ociuhandu has joined #openstack-neutron | 12:37 | |
| *** ccamposr has joined #openstack-neutron | 12:38 | |
| *** ociuhandu has quit IRC | 12:41 | |
| *** ociuhandu has joined #openstack-neutron | 12:41 | |
| *** ociuhandu has quit IRC | 12:42 | |
| *** ociuhandu has joined #openstack-neutron | 12:44 | |
| *** rpittau|bbl is now known as rpittau | 12:44 | |
| *** Luzi has joined #openstack-neutron | 12:55 | |
| *** ociuhandu has quit IRC | 13:02 | |
| *** ociuhandu has joined #openstack-neutron | 13:05 | |
| *** hemanth_n has quit IRC | 13:10 | |
| *** ociuhandu has quit IRC | 13:10 | |
| *** ociuhandu has joined #openstack-neutron | 13:13 | |
| *** bengates has quit IRC | 13:19 | |
| *** bengates has joined #openstack-neutron | 13:20 | |
| *** ociuhandu has quit IRC | 13:20 | |
| *** ociuhandu has joined #openstack-neutron | 13:31 | |
| *** ociuhandu has quit IRC | 13:31 | |
| *** ociuhandu has joined #openstack-neutron | 13:32 | |
| *** ociuhandu has quit IRC | 13:38 | |
| *** ociuhandu has joined #openstack-neutron | 13:42 | |
| *** Luzi has quit IRC | 13:43 | |
| openstackgerrit | Merged openstack/neutron master: [ovn]: fix usage of lrp_set_options api to use **options https://review.opendev.org/c/openstack/neutron/+/768277 | 13:47 |
| *** macz_ has joined #openstack-neutron | 13:54 | |
| *** macz_ has quit IRC | 13:59 | |
| *** bengates has quit IRC | 14:06 | |
| *** bengates has joined #openstack-neutron | 14:11 | |
| *** sapd1 has joined #openstack-neutron | 14:25 | |
| *** bengates has quit IRC | 14:30 | |
| *** bengates has joined #openstack-neutron | 14:30 | |
| *** abdysn has quit IRC | 14:41 | |
| openstackgerrit | Flavio Fernandes proposed openstack/ovsdbapp master: Add northbound API support for meters https://review.opendev.org/c/openstack/ovsdbapp/+/767574 | 14:46 |
| *** ociuhandu has quit IRC | 14:50 | |
| *** ociuhandu has joined #openstack-neutron | 14:50 | |
| *** ociuhandu has quit IRC | 14:54 | |
| *** armax has joined #openstack-neutron | 14:57 | |
| *** nweinber has joined #openstack-neutron | 15:02 | |
| *** ociuhandu has joined #openstack-neutron | 15:03 | |
| slaweq | lajoskatona: hi, thx a lot for taking care of the docs requirements issue in neutron projects | 15:11 |
| openstackgerrit | Slawek Kaplonski proposed openstack/neutron-vpnaas master: Add doc/requirements https://review.opendev.org/c/openstack/neutron-vpnaas/+/769380 | 15:11 |
| *** sapd1 has quit IRC | 15:15 | |
| *** jangutter_ has joined #openstack-neutron | 15:19 | |
| *** jangutter has quit IRC | 15:22 | |
| lajoskatona | slaweq: no problem, as I see other networking projects are safe from this issue | 15:26 |
| openstackgerrit | Bernard Cafarelli proposed openstack/neutron master: Update major release checklist https://review.opendev.org/c/openstack/neutron/+/765959 | 15:28 |
| *** sapd1 has joined #openstack-neutron | 15:28 | |
| slaweq | lajoskatona: I just sent same patch to neutron-vpnaas | 15:28 |
| slaweq | and those were only 3 on the list | 15:29 |
| *** sapd1 has quit IRC | 15:35 | |
| *** jangutter has joined #openstack-neutron | 15:40 | |
| *** jangutter has quit IRC | 15:41 | |
| *** jangutter_ has quit IRC | 15:41 | |
| *** jangutter has joined #openstack-neutron | 15:43 | |
| *** yoctozepto has quit IRC | 15:43 | |
| *** yoctozepto has joined #openstack-neutron | 15:44 | |
| *** ramishra has quit IRC | 15:45 | |
| *** bengates has quit IRC | 15:46 | |
| *** bengates_ has joined #openstack-neutron | 15:46 | |
| *** jangutter_ has joined #openstack-neutron | 15:48 | |
| *** sapd1 has joined #openstack-neutron | 15:50 | |
| *** jangutter has quit IRC | 15:51 | |
| *** ramishra has joined #openstack-neutron | 15:52 | |
| openstackgerrit | Merged openstack/neutron stable/ussuri: Fix imports order in neutron.services.ovn_l3_plugin module https://review.opendev.org/c/openstack/neutron/+/767821 | 15:54 |
| *** macz_ has joined #openstack-neutron | 15:57 | |
| *** macz_ has quit IRC | 16:01 | |
| openstackgerrit | Rodolfo Alonso proposed openstack/neutron master: Check the revision number only once https://review.opendev.org/c/openstack/neutron/+/769390 | 16:01 |
| zigo | slaweq: Hi. Any chance you give the final +w ? Rodlfo and Brian already gave a +2 ... https://review.opendev.org/c/openstack/neutron/+/669395 | 16:01 |
| zigo | It's working well, though I still have some related issues. | 16:02 |
| zigo | When I set the external gateway of a virtual router, the port is not using the BGP routed type. | 16:02 |
| zigo | That would have to be fixed somewhere else, though I have no idea how... | 16:03 |
| zigo | Pointers would be welcome. | 16:03 |
| zigo | Also, directly attaching a public IP to a VM doesn't work, for some reasons... | 16:03 |
| *** dave-mccowan has quit IRC | 16:03 | |
| zigo | Back to the router external gateway port, here's what I get: | 16:05 |
| zigo | # openstack port show c151d9a0-5e80-4efc-935c-d2d8e2f81907 -c binding_vif_details -c binding_vif_type | 16:05 |
| zigo | +---------------------+-------------------------------------------------------------------------------------------------------------+ | 16:05 |
| zigo | | Field | Value | | 16:05 |
| zigo | +---------------------+-------------------------------------------------------------------------------------------------------------+ | 16:05 |
| zigo | | binding_vif_details | bridge_name='br-int', connectivity='l2', datapath_type='system', ovs_hybrid_plug='True', port_filter='True' | | 16:05 |
| zigo | | binding_vif_type | ovs | | 16:05 |
| zigo | +---------------------+-------------------------------------------------------------------------------------------------------------+ | 16:05 |
| *** dave-mccowan has joined #openstack-neutron | 16:07 | |
| zigo | Floating IPs have this: | 16:09 |
| zigo | +---------------------+---------+ | 16:09 |
| zigo | | Field | Value | | 16:09 |
| zigo | +---------------------+---------+ | 16:09 |
| zigo | | binding_vif_details | | | 16:09 |
| zigo | | binding_vif_type | unbound | | 16:09 |
| zigo | +---------------------+---------+ | 16:09 |
| zigo | Am I missing something? | 16:09 |
| openstackgerrit | Merged openstack/neutron stable/ussuri: Flush ebtables arp protect chains before deleting them https://review.opendev.org/c/openstack/neutron/+/765408 | 16:10 |
| openstackgerrit | Merged openstack/neutron-dynamic-routing master: Add doc/requirements https://review.opendev.org/c/openstack/neutron-dynamic-routing/+/769315 | 16:20 |
| *** lpetrut has quit IRC | 16:21 | |
| openstackgerrit | Lajos Katona proposed openstack/neutron-specs master: BFD support for Neutron https://review.opendev.org/c/openstack/neutron-specs/+/767337 | 16:29 |
| openstackgerrit | Rodolfo Alonso proposed openstack/neutron master: [WIP] Add debug information for TestMinBwQoSOvs https://review.opendev.org/c/openstack/neutron/+/768476 | 16:40 |
| *** lucasagomes has quit IRC | 16:51 | |
| openstackgerrit | Merged openstack/neutron-tempest-plugin master: Add doc/requirements https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/769330 | 16:52 |
| openstackgerrit | Lajos Katona proposed openstack/neutron-specs master: BFD support for Neutron https://review.opendev.org/c/openstack/neutron-specs/+/767337 | 16:54 |
| *** jlibosva has quit IRC | 16:58 | |
| *** sapd1 has quit IRC | 16:59 | |
| *** bengates_ has quit IRC | 17:06 | |
| *** bengates has joined #openstack-neutron | 17:06 | |
| *** rpittau is now known as rpittau|afk | 17:08 | |
| *** ociuhandu has quit IRC | 17:09 | |
| *** vulkanz has joined #openstack-neutron | 17:11 | |
| vulkanz | hey people | 17:12 |
| vulkanz | I'm deploying a openstack test environment and would appreciate if someone could clear up some information regarding neutron service | 17:12 |
| vulkanz | how does one deploy multiple neutron servers to avoid network congestion since all vm traffic goes through it? | 17:12 |
| vulkanz | would appreciate any insights | 17:12 |
| *** lajoskatona has quit IRC | 17:14 | |
| vulkanz | [neutron] anyone? :) | 17:16 |
| *** ociuhandu has joined #openstack-neutron | 17:21 | |
| *** ociuhandu has quit IRC | 17:26 | |
| *** jangutter_ has quit IRC | 17:41 | |
| *** jangutter has joined #openstack-neutron | 17:43 | |
| *** tesseract has quit IRC | 17:52 | |
| *** bengates has quit IRC | 18:01 | |
| *** bengates has joined #openstack-neutron | 18:02 | |
| *** bengates has quit IRC | 18:08 | |
| *** bengates has joined #openstack-neutron | 18:09 | |
| *** gcheresh has quit IRC | 18:11 | |
| *** ccamposr__ has joined #openstack-neutron | 18:13 | |
| vulkanz | admin0 would love your input here | 18:14 |
| *** ccamposr has quit IRC | 18:15 | |
| *** vulkanz has quit IRC | 18:16 | |
| *** vulkanz has joined #openstack-neutron | 18:17 | |
| *** elvira has quit IRC | 18:18 | |
| admin0 | vulkanz, define congestion :) | 18:20 |
| admin0 | you can have multiple networks | 18:20 |
| vulkanz | okay :) | 18:20 |
| admin0 | unless you have 10 routers .. it still hits the same router | 18:20 |
| vulkanz | I'm totally new to learning openstack but the neutron has my brain boiled, maybe because of minimal installation documentation? | 18:20 |
| admin0 | ok .. i think you are confused about neutron agents and neutron server | 18:21 |
| admin0 | neutron server = API stuff .. | 18:21 |
| admin0 | neutron agents = the actual stuff from where the traffic passes | 18:21 |
| *** ralonsoh has quit IRC | 18:21 | |
| admin0 | one way that I do is make all compute nodes also a network node | 18:21 |
| admin0 | that way, you don't have a single point of failure and less hot zones | 18:21 |
| vulkanz | oh woah | 18:22 |
| *** bengates_ has joined #openstack-neutron | 18:22 | |
| admin0 | so as you scale up compute, you get more nodes that can also act as network . the benefit is one node going down does not affect your whole customers | 18:22 |
| vulkanz | I thought all the vm traffic was passing through the neutron server via l3 agents | 18:22 |
| admin0 | but if you are small ( and not running public clouds) , you can have one dedicated server or a few just for neutron | 18:22 |
| admin0 | you know those old servers that have less ram that cannot act as compute .. or some old cpu that is not fast enough .. make those neutron servers | 18:23 |
| vulkanz | actually I'm trying to deploy a public cloud with 2 compute nodes | 18:23 |
| admin0 | ok .. then make both network nodes as well :) | 18:23 |
| vulkanz | so make both of the compute nodes network as well? | 18:24 |
| vulkanz | on the backend we will have only one neutron server running on a separate dedi | 18:25 |
| vulkanz | correct? | 18:25 |
| *** bengates has quit IRC | 18:25 | |
| *** bengates_ has quit IRC | 18:26 | |
| vulkanz | one more thing: do we need to setup both provider and self service network in order to have self service network? | 18:27 |
| vulkanz | admin0 do let me know whenever you have some free time :) | 18:38 |
| *** mgoddard has quit IRC | 18:45 | |
| admin0 | the neutron server runs on a controller node | 19:01 |
| admin0 | which also runs all other api services | 19:01 |
| admin0 | so you have apis ( that run on controllers) .. hypervisors and then network agents .. you are free to run the network agents on its own nodes, on controllers or on compute nodes | 19:02 |
| vulkanz | oh I see | 19:04 |
| admin0 | there is also storage node you need for cinder if you are going to use volumes as well | 19:04 |
| admin0 | or ceph | 19:04 |
| admin0 | normally for a small setup, i will use one server for deployment + cinder + glance ( via nfs) and then 1 controller and rest of them computes/network | 19:05 |
| vulkanz | theres is one more thing which I'm confused about, if I run the ceph on a 10gig network, can we make it the default storage for vms? | 19:06 |
| vulkanz | I've seen by default host storage is used for vms to boot | 19:06 |
| vulkanz | whats the best practice? because I need HA for storage and ceph is perfect for that purpose | 19:06 |
| *** vulkanz has quit IRC | 19:30 | |
| *** vulkanz has joined #openstack-neutron | 19:32 | |
| *** vulkanz has left #openstack-neutron | 19:34 | |
| *** gcheresh has joined #openstack-neutron | 19:34 | |
| *** vulkanz has joined #openstack-neutron | 19:34 | |
| *** _mlavalle3 has quit IRC | 19:38 | |
| *** gcheresh has quit IRC | 19:52 | |
| openstackgerrit | Brian Haley proposed openstack/ovn-octavia-provider master: Add a Kuryr Kubernetes co-gating job https://review.opendev.org/c/openstack/ovn-octavia-provider/+/760225 | 19:57 |
| openstackgerrit | Brian Haley proposed openstack/ovn-octavia-provider master: Add SCTP support https://review.opendev.org/c/openstack/ovn-octavia-provider/+/754429 | 20:03 |
| openstackgerrit | Merged openstack/ovsdbapp master: Fix gate failure https://review.opendev.org/c/openstack/ovsdbapp/+/767956 | 20:11 |
| admin0 | vulkanz, you can start with the 10g network and as you monitor the bandwidth/usage, bond it or upgrade to 40 or 100 | 20:27 |
| *** tmorin has quit IRC | 20:30 | |
| vulkanz | admin0, got it! will bond a few ports as I have a 10gig switch which 16 ports should be enough for initial setup I believe. | 20:32 |
| vulkanz | admin0, did you read my question about storage? :) | 20:32 |
| *** slaweq has quit IRC | 20:33 | |
| *** dave-mccowan has quit IRC | 20:37 | |
| admin0 | you can use ceph for cinder, glance and also nova ( vms ) | 20:39 |
| vulkanz | do most public providers use central storage for booting vms? or host based | 20:41 |
| vulkanz | that's the confusion I have | 20:41 |
| vulkanz | because if we use host-based storage then there can't be HA | 20:41 |
| *** dave-mccowan has joined #openstack-neutron | 20:41 | |
| admin0 | vulkanz, most do both :) | 20:55 |
| admin0 | based on your use case | 20:56 |
| admin0 | if customers pay for HA, then HA .. and sometimes even if you have HA, you want to provide faster disks ( like local nvme storage ) for latency sensitive databases | 20:56 |
| admin0 | just one thing to note is that one compute node can do either HA or local disk | 20:56 |
| admin0 | and its controlled via flavors on where you want the instance to boot | 20:57 |
| vulkanz | I see, so in case of proper HA everything goes to ceph | 20:57 |
| *** nplanel_ has quit IRC | 20:58 | |
| *** nplanel has joined #openstack-neutron | 20:58 | |
| *** jmlowe has joined #openstack-neutron | 21:00 | |
| admin0 | if you have say raid10 on local disks, and also do not oversell on ram or disk , and only slightly on cpu, then its stable as well | 21:05 |
| admin0 | local disk works just fine too | 21:05 |
| vulkanz | I see, but I'm aiming for proper HA | 21:09 |
| vulkanz | just in case a customer starts a vm and we boot it up from local raid10, drives or array fails and he loses everything | 21:10 |
| vulkanz | that's why I figured it might be a better idea to run ssd raid on ceph and use that as main storage for vms etc and here asking you for advice :) | 21:10 |
| admin0 | you can do that | 21:16 |
| admin0 | you need 3x storage nodes for ceph, | 21:16 |
| admin0 | 2x disks on raid1 go to the OS itself, rest to ceph | 21:17 |
| vulkanz | perfect! | 21:22 |
| vulkanz | admin0 you have cleared so many questions in my mind that i couldn't find a proper answer to anywhere else, thank you very much for that! | 21:22 |
| admin0 | you are welcome | 21:23 |
| vulkanz | I will get on with futher reading and get back to you later if I have any more questions :D | 21:23 |
| admin0 | if you have more,feel free to ask as well | 21:23 |
| *** tbachman has quit IRC | 21:23 | |
| vulkanz | Okay :D | 21:25 |
| *** tbachman has joined #openstack-neutron | 21:26 | |
| vulkanz | somehwere in the docs I read that network nodes are not required any moore but in victoria docs I see that theres a proper way to setup network nodes for self-f serv networks (https://docs.openstack.org/neutron/victoria/admin/deploy-lb-selfservice.html) | 21:26 |
| vulkanz | all of this can be on host nodes instead of dedicated network nodes so each node handles its public internet through a dedicated port directly to the 10gig switch etc? | 21:27 |
| vulkanz | if* | 21:28 |
| vulkanz | and in the doc page above now theres a third interface (overlay), can't this be achieved on a normal provider network interface using bridge? for self service networks that is | 21:30 |
| vulkanz | networking part of openstack is sooo overwhelming tbh lol | 21:33 |
| admin0 | for openstack, you can have just one network interface, or many .. .. irrespective of the number of network interfaces you have, (minimum one) , you will be able to create your own self service networks . and also provider network | 21:36 |
| admin0 | provider network means a network which takes your traffic outside of openstack ( to the internet) for example | 21:36 |
| admin0 | so if you have 1 10g port, or you use a bond, its possible to do what you want to do | 21:37 |
| admin0 | in a single network card, all traffic ( like api, storage, vxlan ) are on fixed vlans .. and then other external networks you add, they get their own vlan | 21:37 |
| admin0 | so you might add ext-net as a network on vlan 101 and it goes to your public router towards internet .. and you can also add ext-lan towards your own internal router/vlan/vpn and its not exposed to internet | 21:38 |
| vulkanz | I see! | 21:38 |
| *** ircuser-1 has quit IRC | 21:38 | |
| admin0 | if you have multiple network cards, its good to separate the traffic domains and put storage, vxlan, api, vlan etc in their own separate cards | 21:39 |
| vulkanz | I currently have 4x1G and 2x10G sfp+ ports in each server | 21:39 |
| vulkanz | so for management network we could use the 1g, 1x10G for storage and 1x10G for vm network? | 21:39 |
| vulkanz | vm network as in for where the vms connect to internet | 21:40 |
| admin0 | you can also bond 4x1g into bond1 and use that for ext-net ( towards internet) and then bond 2x10g, create 3 vlans -- for api, storage and east-west and be done with it | 21:40 |
| admin0 | it also depends on what type of workload, what kind of connection the instances need with internet etc | 21:40 |
| admin0 | would you be shelving the vms, using snapshots, using glance a lot etc -- this determins how big you need your api network to be | 21:42 |
| vulkanz | yes certainly a lot of customers do that in our scenario and they require persisent disks hence the ceph choice | 21:43 |
| admin0 | if you plan to have 100 vms on 1 server and they are all cpanel server, then ofcourse 4x 1g bond for internet will not be enough .. you need to give maybe 10g | 21:43 |
| vulkanz | wouldn't it be better to run 10g cards for ceph for better io? | 21:44 |
| admin0 | ceph/storage = always on 10g | 21:44 |
| admin0 | 10g bond | 21:44 |
| vulkanz | so 20gig in total for each compute host | 21:44 |
| vulkanz | and another 10g card for pub net | 21:44 |
| vulkanz | yeah this makes sense to just ditch 1g at all and run 10g and make life easier than having to create bonds and use lot of port space lol | 21:45 |
| admin0 | if you can get 4x 10g, then what you can do is 1x 10g = vlan(north-south) , 1x10g = vxlan ( east-west) , 2x 10gb = bond0 = api and storage | 21:46 |
| vulkanz | correct me if I am wrong but isnt east-west for storage and internal traffic flow of the cloud? or are we considering a direct connection to storage? | 21:48 |
| vulkanz | internal I mean api + storage etc | 21:48 |
| admin0 | east-west is more like vm <-> vm traffic | 21:49 |
| admin0 | and its a good idea to separate it from storage | 21:49 |
| admin0 | coz east-west and north-south is what users can utilize to the full .. | 21:49 |
| vulkanz | oh wow that clears up a lot of things and I was thinking totally the opposite! thanks for clearing that up sir! | 21:49 |
| admin0 | so in a big cloud if multiple customers decide to run say iperf continously, your storage traffic will go slow and all vms will go slow as all are affected | 21:50 |
| admin0 | so its a good idea to separate these types of abuse domains | 21:50 |
| vulkanz | woot, that is indeed perfect, I'm saving this conversation for reference when building :D | 21:50 |
| vulkanz | do you also think its a good idea to have a totally separate switch for east-west as well? | 21:51 |
| admin0 | there might be 50+ ways to do these stuff and how to make it work optimially | 21:51 |
| admin0 | the best one depends on your workload | 21:51 |
| vulkanz | exactly that's what I'm looking to build from the start an optimal system that gives no troubles later | 21:52 |
| admin0 | what your customers are, hwo are they in the cloud, what are they expeceted to do .. what kind of vms will run etc | 21:52 |
| vulkanz | tbh most of the customers run normal day to day servers e.g cpanel hosting, and then we have some of our own threat intelligence feeds that run on elastic clusters which generate a lot of feeds for clients. Also, some clients run backup servers on our vms ( currently we have a custom kvm setup) | 21:54 |
| admin0 | how do you plan to deploy openstack ? | 21:55 |
| vulkanz | Well my plan is to buy all the required equipment and deploy it in a datacenter where we have colo space and run a public cloud | 21:55 |
| vulkanz | starting budget is around 60k | 21:56 |
| admin0 | ok | 21:56 |
| admin0 | there is a lot of stuff you need to prepare . ips, bgp, vlan tags, network gear, maybe look into m1000e blades etc | 21:57 |
| vulkanz | good thing is we currently have a /20 from arin | 21:57 |
| vulkanz | but do we really need our own bgp setup? currently all our ranges are being advertised on the dc core router | 21:58 |
| admin0 | it depends on how you want to add to openstack .. all floating or direct public etc | 21:58 |
| openstackgerrit | Brian Haley proposed openstack/networking-ovn stable/train: Fix gate failure https://review.opendev.org/c/openstack/networking-ovn/+/767424 | 22:13 |
| *** jamesdenton has quit IRC | 22:19 | |
| *** jamesden_ has joined #openstack-neutron | 22:19 | |
| *** tbachman_ has joined #openstack-neutron | 22:31 | |
| *** tbachman has quit IRC | 22:32 | |
| *** tbachman_ is now known as tbachman | 22:32 | |
| *** nweinber has quit IRC | 22:32 | |
| *** tkajinam has joined #openstack-neutron | 23:01 | |
| *** dave-mccowan has quit IRC | 23:14 | |
| *** dave-mccowan has joined #openstack-neutron | 23:15 | |
| *** zzzeek has quit IRC | 23:25 | |
| *** zzzeek has joined #openstack-neutron | 23:27 | |
| *** rcernin has joined #openstack-neutron | 23:33 | |
| openstackgerrit | Merged openstack/ovn-octavia-provider stable/ussuri: Fix gate failure https://review.opendev.org/c/openstack/ovn-octavia-provider/+/767224 | 23:38 |
| *** rcernin has quit IRC | 23:53 | |
| *** rcernin has joined #openstack-neutron | 23:56 | |
| *** rcernin has quit IRC | 23:59 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!