| *** changzhi has joined #openstack-neutron | 00:00 | |
| *** carl_baldwin has joined #openstack-neutron | 00:01 | |
| *** saksham has joined #openstack-neutron | 00:01 | |
| *** thumpba has quit IRC | 00:03 | |
| *** saksham has quit IRC | 00:05 | |
| *** aranjan has quit IRC | 00:08 | |
| *** Marga_ has joined #openstack-neutron | 00:11 | |
| *** iwamoto has joined #openstack-neutron | 00:12 | |
| *** piyanai has quit IRC | 00:13 | |
| *** armax has quit IRC | 00:14 | |
| *** Marga_ has quit IRC | 00:16 | |
| *** kbyrne has quit IRC | 00:17 | |
| *** shoutm has quit IRC | 00:17 | |
| openstackgerrit | Takashi NATSUME proposed openstack/neutron-specs: Fix typos in ipv6-router.rst/qos-api-extension.rst https://review.openstack.org/208847 | 00:20 |
|---|---|---|
| *** thumpba has joined #openstack-neutron | 00:20 | |
| *** kbyrne has joined #openstack-neutron | 00:20 | |
| *** shoutm has joined #openstack-neutron | 00:20 | |
| *** sridhar_ram has joined #openstack-neutron | 00:23 | |
| *** shadower has quit IRC | 00:23 | |
| *** shadower has joined #openstack-neutron | 00:23 | |
| *** changzhi has quit IRC | 00:31 | |
| *** thedodd has quit IRC | 00:32 | |
| *** longguang_ has joined #openstack-neutron | 00:34 | |
| *** longguang has quit IRC | 00:36 | |
| *** longguang_ is now known as longguang | 00:36 | |
| *** thumpba has quit IRC | 00:38 | |
| *** tiswanso has joined #openstack-neutron | 00:43 | |
| *** vivek-ebay has quit IRC | 00:44 | |
| *** shoutm has quit IRC | 00:44 | |
| *** tiswanso has quit IRC | 00:56 | |
| *** shoutm has joined #openstack-neutron | 01:02 | |
| *** yalie has joined #openstack-neutron | 01:03 | |
| *** zhenguo has joined #openstack-neutron | 01:04 | |
| *** aranjan has joined #openstack-neutron | 01:05 | |
| *** moshele has quit IRC | 01:06 | |
| *** haigang has joined #openstack-neutron | 01:09 | |
| *** johnbelamaric has joined #openstack-neutron | 01:12 | |
| *** Marga_ has joined #openstack-neutron | 01:12 | |
| *** shoutm has quit IRC | 01:16 | |
| *** Marga_ has quit IRC | 01:17 | |
| *** zhhuabj has joined #openstack-neutron | 01:18 | |
| *** shoutm has joined #openstack-neutron | 01:19 | |
| *** yamamoto has joined #openstack-neutron | 01:21 | |
| *** daneyon_ has quit IRC | 01:24 | |
| *** yamamoto has quit IRC | 01:26 | |
| *** haigang has quit IRC | 01:26 | |
| *** gildub has joined #openstack-neutron | 01:28 | |
| *** shoutm has quit IRC | 01:28 | |
| *** shoutm has joined #openstack-neutron | 01:30 | |
| *** erkules has joined #openstack-neutron | 01:35 | |
| *** erkules_ has quit IRC | 01:37 | |
| *** aranjan has quit IRC | 01:38 | |
| *** markvoelker has joined #openstack-neutron | 01:39 | |
| *** changzhi has joined #openstack-neutron | 01:39 | |
| *** markvoelker has quit IRC | 01:44 | |
| openstackgerrit | Henry Gessau proposed openstack/neutron: Consistent layout and headings for devref https://review.openstack.org/210769 | 01:48 |
| *** erkules has quit IRC | 01:52 | |
| *** erkules has joined #openstack-neutron | 01:54 | |
| openstackgerrit | ZhaoBo proposed openstack/neutron: Fixed pass wrong cidr check during create subnet https://review.openstack.org/201942 | 01:55 |
| *** alex_xu has joined #openstack-neutron | 01:56 | |
| *** vivek-ebay has joined #openstack-neutron | 01:56 | |
| *** johnbelamaric has quit IRC | 01:59 | |
| *** alejandrito has joined #openstack-neutron | 02:08 | |
| *** thumpba has joined #openstack-neutron | 02:08 | |
| *** banix has joined #openstack-neutron | 02:09 | |
| *** jyuso1 has joined #openstack-neutron | 02:09 | |
| *** minwang2 has joined #openstack-neutron | 02:12 | |
| *** baoli has quit IRC | 02:12 | |
| *** Marga_ has joined #openstack-neutron | 02:13 | |
| *** baoli has joined #openstack-neutron | 02:15 | |
| *** thumpba has quit IRC | 02:15 | |
| *** Marga_ has quit IRC | 02:17 | |
| *** vivek-ebay has quit IRC | 02:18 | |
| *** shwetaap has joined #openstack-neutron | 02:20 | |
| *** yamamoto has joined #openstack-neutron | 02:33 | |
| *** aranjan has joined #openstack-neutron | 02:35 | |
| *** alejandrito has quit IRC | 02:37 | |
| openstackgerrit | yalei wang proposed openstack/neutron: Add support for unaddressed port https://review.openstack.org/200809 | 02:39 |
| *** patrickeast has joined #openstack-neutron | 02:45 | |
| *** agireud has joined #openstack-neutron | 02:45 | |
| *** armax has joined #openstack-neutron | 02:50 | |
| *** hakimo has joined #openstack-neutron | 02:52 | |
| *** tiswanso has joined #openstack-neutron | 02:52 | |
| openstackgerrit | Zhiyuan Cai proposed openstack/neutron-specs: Define timestamp as extension attribute https://review.openstack.org/210955 | 02:54 |
| *** patrickeast has quit IRC | 02:54 | |
| *** hakimo_ has quit IRC | 02:55 | |
| *** vivek-ebay has joined #openstack-neutron | 02:59 | |
| openstackgerrit | YAMAMOTO Takashi proposed openstack/neutron: db_base_plugin_v2: Avoid creating another session https://review.openstack.org/198420 | 03:00 |
| *** tiswanso has quit IRC | 03:01 | |
| *** baoli has quit IRC | 03:01 | |
| openstackgerrit | Ryan Moats proposed openstack/neutron: Distributed router can not add routes https://review.openstack.org/144205 | 03:05 |
| *** aranjan has quit IRC | 03:09 | |
| *** thumpba has joined #openstack-neutron | 03:12 | |
| *** Marga_ has joined #openstack-neutron | 03:14 | |
| openstackgerrit | Ryan Moats proposed openstack/neutron: Distributed router can not add routes https://review.openstack.org/144205 | 03:15 |
| *** thumpba has quit IRC | 03:17 | |
| *** gangil has quit IRC | 03:18 | |
| *** Marga_ has quit IRC | 03:18 | |
| *** yushiro has quit IRC | 03:25 | |
| *** sridhar_ram has quit IRC | 03:25 | |
| openstackgerrit | Ryan Moats proposed openstack/neutron: Distributed router can not add routes https://review.openstack.org/144205 | 03:26 |
| *** regXboi has quit IRC | 03:27 | |
| *** shwetaap has quit IRC | 03:27 | |
| *** markvoelker has joined #openstack-neutron | 03:40 | |
| *** markvoelker has quit IRC | 03:45 | |
| *** norman has joined #openstack-neutron | 03:47 | |
| norman | Hi ,all, support that I have two openstack cloud based on vlan, is there a way (or possible?) to connect the 2 routers in each cloud by GRE tunnal, then VMs can communicate with each other with the it's fix ip ? | 03:50 |
| *** shz has quit IRC | 03:50 | |
| norman | 'connect the 2 routers in *different* cloud'... | 03:52 |
| *** vivek-ebay has quit IRC | 03:54 | |
| *** yamahata has joined #openstack-neutron | 03:59 | |
| *** yushiro has joined #openstack-neutron | 04:00 | |
| *** fedexo has joined #openstack-neutron | 04:00 | |
| *** aranjan has joined #openstack-neutron | 04:06 | |
| *** BhavyaM has joined #openstack-neutron | 04:08 | |
| *** yamamoto has quit IRC | 04:11 | |
| openstackgerrit | Yushiro FURUKAWA proposed openstack/neutron-specs: Logging API for security group and firewall rules https://review.openstack.org/203509 | 04:12 |
| *** sthillma has joined #openstack-neutron | 04:12 | |
| *** Marga_ has joined #openstack-neutron | 04:15 | |
| *** armax has quit IRC | 04:15 | |
| *** trinaths has joined #openstack-neutron | 04:15 | |
| *** sthillma_ has joined #openstack-neutron | 04:15 | |
| *** vthapar has joined #openstack-neutron | 04:16 | |
| *** yamamoto has joined #openstack-neutron | 04:18 | |
| *** sthillma has quit IRC | 04:18 | |
| *** sthillma_ is now known as sthillma | 04:18 | |
| *** Marga_ has quit IRC | 04:19 | |
| yushiro | haleyb, Hi. I've updated my spec : https://review.openstack.org/203509 Would you please review it again? | 04:20 |
| yushiro | and HenryG :) | 04:22 |
| *** agireud has quit IRC | 04:27 | |
| *** zhhuabj has quit IRC | 04:28 | |
| *** agireud has joined #openstack-neutron | 04:31 | |
| *** yamahata has quit IRC | 04:31 | |
| *** oomichi has joined #openstack-neutron | 04:32 | |
| *** banix has quit IRC | 04:34 | |
| *** vishwanathj has quit IRC | 04:36 | |
| *** aranjan has quit IRC | 04:39 | |
| *** banix has joined #openstack-neutron | 04:41 | |
| *** ajmiller has quit IRC | 04:49 | |
| *** achanda has joined #openstack-neutron | 04:53 | |
| *** banix has quit IRC | 04:55 | |
| *** itzikb_ has joined #openstack-neutron | 04:57 | |
| *** tiswanso has joined #openstack-neutron | 04:57 | |
| *** sdake has quit IRC | 05:01 | |
| *** yfried_ has quit IRC | 05:02 | |
| *** mkolesni__ has joined #openstack-neutron | 05:03 | |
| *** ekuris has joined #openstack-neutron | 05:04 | |
| yushiro | ping mestery | 05:07 |
| *** tiswanso has quit IRC | 05:08 | |
| *** irenab has quit IRC | 05:10 | |
| *** jprovazn has joined #openstack-neutron | 05:12 | |
| *** Guest41459 is now known as atmark | 05:13 | |
| *** atmark has quit IRC | 05:13 | |
| *** atmark has joined #openstack-neutron | 05:13 | |
| *** thumpba has joined #openstack-neutron | 05:14 | |
| *** Marga_ has joined #openstack-neutron | 05:15 | |
| *** slaweq has joined #openstack-neutron | 05:18 | |
| *** thumpba has quit IRC | 05:18 | |
| *** fedexo has quit IRC | 05:18 | |
| *** Marga_ has quit IRC | 05:20 | |
| *** jasondotstar has joined #openstack-neutron | 05:21 | |
| *** anilvenkata has joined #openstack-neutron | 05:21 | |
| *** lajos-katona has joined #openstack-neutron | 05:22 | |
| *** Sukhdev has joined #openstack-neutron | 05:24 | |
| *** jasondotstar has quit IRC | 05:26 | |
| *** agireud has quit IRC | 05:29 | |
| openstackgerrit | Hirofumi Ichihara proposed openstack/neutron: Add availability_zone support for API and DB https://review.openstack.org/183369 | 05:30 |
| *** SridharG has joined #openstack-neutron | 05:31 | |
| *** slaweq has quit IRC | 05:31 | |
| *** gangil has joined #openstack-neutron | 05:32 | |
| *** longguang_ has joined #openstack-neutron | 05:32 | |
| *** longguang has quit IRC | 05:32 | |
| *** longguang_ is now known as longguang | 05:33 | |
| *** aranjan has joined #openstack-neutron | 05:34 | |
| *** jianyong has quit IRC | 05:34 | |
| *** jianyong has joined #openstack-neutron | 05:36 | |
| openstackgerrit | Hirofumi Ichihara proposed openstack/neutron: Add availability_zone support for Scheduler https://review.openstack.org/204436 | 05:39 |
| *** salv-orlando has joined #openstack-neutron | 05:41 | |
| *** markvoelker has joined #openstack-neutron | 05:41 | |
| *** sudipto has joined #openstack-neutron | 05:43 | |
| *** markvoelker has quit IRC | 05:46 | |
| *** ihrachyshka has joined #openstack-neutron | 05:47 | |
| *** irenab has joined #openstack-neutron | 05:47 | |
| *** gampel has joined #openstack-neutron | 05:47 | |
| *** salv-orlando has quit IRC | 05:48 | |
| openstackgerrit | Hirofumi Ichihara proposed openstack/neutron: Add functional test for availability_zone support https://review.openstack.org/210977 | 05:48 |
| *** yfried_ has joined #openstack-neutron | 05:55 | |
| *** yfried__ has joined #openstack-neutron | 05:56 | |
| *** hanzhang has joined #openstack-neutron | 05:57 | |
| *** norman has quit IRC | 05:57 | |
| *** yfried__ is now known as yfried | 05:59 | |
| *** yfried_ has quit IRC | 05:59 | |
| *** salv-orlando has joined #openstack-neutron | 06:01 | |
| *** jyuso1 has quit IRC | 06:02 | |
| *** josecastroleon has joined #openstack-neutron | 06:07 | |
| *** atmark has quit IRC | 06:08 | |
| openstackgerrit | Takashi NATSUME proposed openstack/neutron-vpnaas: Rename a test method in test_ipsec.py https://review.openstack.org/210989 | 06:09 |
| *** aranjan has quit IRC | 06:09 | |
| *** Murali_ has joined #openstack-neutron | 06:10 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/neutron: Imported Translations from Transifex https://review.openstack.org/206885 | 06:11 |
| *** aranjan has joined #openstack-neutron | 06:12 | |
| *** aranjan_ has joined #openstack-neutron | 06:14 | |
| *** aranjan has quit IRC | 06:14 | |
| *** thumpba has joined #openstack-neutron | 06:15 | |
| *** minwang2 has quit IRC | 06:15 | |
| *** Marga_ has joined #openstack-neutron | 06:16 | |
| *** thumpba has quit IRC | 06:19 | |
| *** Marga_ has quit IRC | 06:21 | |
| *** afazekas_ has joined #openstack-neutron | 06:24 | |
| *** apuimedo has quit IRC | 06:24 | |
| openstackgerrit | Sudhakar Babu Gariganti proposed openstack/neutron: Setup firewall filters only for required ports https://review.openstack.org/210321 | 06:28 |
| *** scheuran has joined #openstack-neutron | 06:29 | |
| *** hichihara has quit IRC | 06:29 | |
| *** ildikov has quit IRC | 06:30 | |
| *** moshele has joined #openstack-neutron | 06:31 | |
| *** vikram has joined #openstack-neutron | 06:31 | |
| *** aranjan_ has quit IRC | 06:31 | |
| openstackgerrit | Yushiro FURUKAWA proposed openstack/neutron-specs: Logging API for security group and firewall rules https://review.openstack.org/203509 | 06:32 |
| *** rubasov has joined #openstack-neutron | 06:33 | |
| moshele | ihrachyshka: hi | 06:35 |
| ihrachyshka | moshele, hi | 06:35 |
| moshele | ihrachyshka: did you see my mail regarding the singleton ? | 06:36 |
| ihrachyshka | moshele, not yet. let me check | 06:36 |
| ihrachyshka | moshele, while I do it, https://review.openstack.org/210998 | 06:36 |
| moshele | ihrachyshka: cool | 06:37 |
| *** jkraj has joined #openstack-neutron | 06:43 | |
| openstackgerrit | venkata anil proposed openstack/neutron: Validate local_ip for OVS tunnel https://review.openstack.org/154043 | 06:49 |
| ihrachyshka | moshele, replied to email | 06:51 |
| *** abregman has joined #openstack-neutron | 06:52 | |
| moshele | ihrachyshka: so I will just fix the SR-IOV like this http://pastebin.com/4tWe3cwb part and we can revisit the singleton | 06:54 |
| ihrachyshka | yeah | 06:56 |
| ihrachyshka | moshele, what about resources_rpc? | 06:56 |
| *** jyuso1 has joined #openstack-neutron | 06:57 | |
| moshele | ihrachyshka: do you want me to fix it or we can just add TODO to revisit it ? | 06:58 |
| ihrachyshka | moshele, meh, I think I'll push a patch in 3 mins | 06:58 |
| moshele | ihrachyshka: your fast :) | 06:59 |
| *** nyechiel has joined #openstack-neutron | 07:01 | |
| *** hakimo has quit IRC | 07:02 | |
| *** diga has quit IRC | 07:02 | |
| *** hakimo has joined #openstack-neutron | 07:03 | |
| *** kiran-r has joined #openstack-neutron | 07:03 | |
| *** ildikov has joined #openstack-neutron | 07:03 | |
| *** diga has joined #openstack-neutron | 07:03 | |
| *** tiswanso has joined #openstack-neutron | 07:04 | |
| *** jpena has joined #openstack-neutron | 07:07 | |
| *** devvesa has joined #openstack-neutron | 07:07 | |
| *** Samos123 has quit IRC | 07:09 | |
| *** sthillma has quit IRC | 07:11 | |
| *** enikanorov2 has joined #openstack-neutron | 07:12 | |
| *** Samos123 has joined #openstack-neutron | 07:12 | |
| *** zhhuabj has joined #openstack-neutron | 07:12 | |
| vikram | Unit tests not working with the latest master.. | 07:13 |
| *** tiswanso has quit IRC | 07:13 | |
| vikram | ValueError: Expecting ',' delimiter: line 197 column 5 (char 8784) | 07:14 |
| vikram | Does anyone else facing similar issue? | 07:14 |
| *** thumpba has joined #openstack-neutron | 07:15 | |
| *** achanda has quit IRC | 07:16 | |
| *** abehl has joined #openstack-neutron | 07:16 | |
| *** abehl has quit IRC | 07:16 | |
| *** abehl has joined #openstack-neutron | 07:17 | |
| *** Marga_ has joined #openstack-neutron | 07:17 | |
| *** jlibosva has joined #openstack-neutron | 07:17 | |
| *** achanda has joined #openstack-neutron | 07:18 | |
| *** ronis has joined #openstack-neutron | 07:19 | |
| *** thumpba has quit IRC | 07:20 | |
| moshele | ihrachyshka: I implemented the delete and I move more logic to the qos_drive base for avoid dup I will send a patch on ovs soon | 07:20 |
| *** Marga_ has quit IRC | 07:21 | |
| *** Sukhdev has quit IRC | 07:23 | |
| *** kawa2014 has joined #openstack-neutron | 07:24 | |
| *** Marga_ has joined #openstack-neutron | 07:25 | |
| *** shoutm has quit IRC | 07:27 | |
| *** achanda has quit IRC | 07:28 | |
| *** shoutm has joined #openstack-neutron | 07:30 | |
| *** gangil has quit IRC | 07:31 | |
| *** achanda has joined #openstack-neutron | 07:31 | |
| ihrachyshka | moshele++ | 07:33 |
| *** yamamoto has quit IRC | 07:33 | |
| *** jschwarz has joined #openstack-neutron | 07:33 | |
| openstackgerrit | vikram.choudhary proposed openstack/python-neutronclient: [WIP] BGP Dynamic Routing: neutronclient changes https://review.openstack.org/205756 | 07:34 |
| *** Marga_ has quit IRC | 07:35 | |
| *** apuimedo has joined #openstack-neutron | 07:36 | |
| *** Marga_ has joined #openstack-neutron | 07:36 | |
| *** zhhuabj has quit IRC | 07:36 | |
| moshele | ihrachyshka: why the subscribe is in the init and not little before the pull? why it is not after https://github.com/openstack/neutron/blob/feature/qos/neutron/agent/l2/extensions/qos.py#L131 ? | 07:37 |
| *** itzikb_ has quit IRC | 07:39 | |
| *** markvoelker has joined #openstack-neutron | 07:42 | |
| *** yamamoto has joined #openstack-neutron | 07:42 | |
| *** zhhuabj has joined #openstack-neutron | 07:43 | |
| ihrachyshka | moshele, because it's too late (rpc consumers already started) | 07:44 |
| ihrachyshka | moshele, we should subscribe before agent calls consume_in_threads on the connection object | 07:44 |
| *** jprovazn_ has joined #openstack-neutron | 07:45 | |
| *** jprovazn_ has quit IRC | 07:45 | |
| *** jlanoux has joined #openstack-neutron | 07:46 | |
| moshele | ihrachyshka: so we can have update call before create | 07:46 |
| ihrachyshka | moshele, yes, that's the race I refer in TODO there | 07:46 |
| *** markvoelker has quit IRC | 07:46 | |
| *** scheuran has quit IRC | 07:46 | |
| *** achanda has quit IRC | 07:46 | |
| ihrachyshka | moshele, agent side solution is to lock on create | 07:47 |
| ihrachyshka | moshele, there is another race on server side where update can be sent before create. for that one, there is no easy solution since neutron-servers have workers and support HA | 07:47 |
| *** jasondotstar has joined #openstack-neutron | 07:48 | |
| ihrachyshka | moshele, solution could include e.g. timestamping update notifications so that agent can drop those that are older than the latest update seen for the policy. | 07:48 |
| *** yfried is now known as yfried|afk | 07:49 | |
| ihrachyshka | moshele, though that relies on correct NTP setup | 07:49 |
| moshele | ihrachyshka: I am guessing the in not for liberty | 07:50 |
| ihrachyshka | moshele, I suspect yes. we'll refine rpc api in M I guess | 07:50 |
| *** jlanoux has quit IRC | 07:50 | |
| *** jasondotstar has quit IRC | 07:52 | |
| *** yfried|afk is now known as yfried | 07:53 | |
| *** salv-orlando has quit IRC | 07:54 | |
| *** rossella_s has joined #openstack-neutron | 07:54 | |
| *** jlanoux has joined #openstack-neutron | 07:56 | |
| ihrachyshka | ajo, hi | 07:59 |
| ajo | hi ihrachyshka, morning | 08:00 |
| ihrachyshka | ajo, we need some reviews :) | 08:01 |
| ihrachyshka | ajo, here are small ones | 08:01 |
| ihrachyshka | sriov singleton: https://review.openstack.org/211005 | 08:01 |
| ihrachyshka | resources_rpc singleton: https://review.openstack.org/211003 | 08:01 |
| * ajo reads | 08:01 | |
| ihrachyshka | kill config value for agent qos driver: https://review.openstack.org/210998 | 08:01 |
| ajo | ihrachyshka, I disagree about that last one ;) | 08:01 |
| ajo | moshele ^ | 08:01 |
| ihrachyshka | tell me | 08:02 |
| ajo | The plan is that OVS-hybrid could use a "linuxbridge" implementation too ;) | 08:02 |
| ajo | by the fact that we are using LB | 08:02 |
| ajo | if lb implementation were more capable, more performant, or whatever, why locking the user into a specific one? | 08:02 |
| ajo | I only believe the agent should default to the common one | 08:03 |
| ajo | also | 08:03 |
| ajo | in OVS, | 08:03 |
| *** itzikb_ has joined #openstack-neutron | 08:03 | |
| *** jprovazn has quit IRC | 08:03 | |
| ajo | somebody could come, and create a "openflow + metering" based one, and just refer it from config | 08:03 |
| ihrachyshka | ajo, so then agent will pass its own config option as driver_type. | 08:03 |
| *** gildub has quit IRC | 08:03 | |
| moshele | ajo: even so this change cause problem with the duplicate config, I think we can push it later | 08:04 |
| ihrachyshka | we should not expose that on the extension manager itself | 08:04 |
| ajo | (as I POC'd for the ovs/ct firewall) | 08:04 |
| ajo | aha | 08:04 |
| ihrachyshka | ajo, that's fine, as long as its agent that decides whether it's configurable | 08:04 |
| ajo | ok, in that case in OVS agent we would have that specific "qos_driver" config entry | 08:04 |
| ajo | but not in other agents | 08:04 |
| ihrachyshka | for sr-iov case, it does not make sense. while you still enforce it to define the option | 08:04 |
| ajo | that'd make sense | 08:04 |
| *** yog_ has joined #openstack-neutron | 08:04 | |
| ajo | ok | 08:05 |
| ihrachyshka | yeap. and since we don't have lb anyway... it's YAGNI | 08:05 |
| ajo | ihrachyshka, yeah, but we can't introduce new config options later, | 08:05 |
| ajo | I'd like to leave that open for OVS (because of openflow, and because of hybrid) | 08:05 |
| ajo | if I make sense | 08:05 |
| ihrachyshka | ajo, hm? why can't we? | 08:05 |
| ihrachyshka | ajo, you define a new option, you pass it thru driver_type. What's so hard? | 08:06 |
| ajo | ihrachyshka, later in cycle, can we? (doc impacts?) sorry, probably what we can't do is backporting without a good reason | 08:06 |
| * ajo pushes more coffee inside | 08:06 | |
| ihrachyshka | ajo, well... do you mean you are going to work on lb in this cycle? | 08:06 |
| ihrachyshka | I haven't seen anyone up to the job | 08:06 |
| ihrachyshka | I doubt it will be in | 08:06 |
| ajo | ok | 08:06 |
| ajo | let's leave my yagni for M then ;) | 08:07 |
| ihrachyshka | I think if a point in the cycle allows to introduce new drivers for qos, it should allow to introduce a knob for enabling it. | 08:07 |
| ajo | yes, you're right guys :) | 08:08 |
| jschwarz | morning guys :() | 08:10 |
| jschwarz | :) | 08:10 |
| jschwarz | when are we sending the mergeback patch? | 08:10 |
| *** jistr has joined #openstack-neutron | 08:12 | |
| ihrachyshka | jschwarz, I planned that today, but I wanted to see a bit more functional tests before that. ajo, do you plan to get more scenarios today? | 08:15 |
| *** abregman is now known as abregman|brb | 08:16 | |
| ajo | ihrachyshka, yes, looking into delete & update | 08:16 |
| moshele | ajo: I have something almost ready | 08:16 |
| ajo | moshele: functional tests? :) | 08:17 |
| *** zhhuabj has quit IRC | 08:17 | |
| moshele | ajo: no, code for the delete I mean :) | 08:17 |
| ajo | ah, ok, moshele , then I will look into update first, I planned to fix the delete part as I was writing the test | 08:17 |
| ajo | moshele, but it makes sense to split it, I can take your patch, and add the functional test on top | 08:18 |
| *** lajos-katona has left #openstack-neutron | 08:18 | |
| moshele | ajo: cool | 08:18 |
| ajo | ping me as it's ready | 08:18 |
| moshele | ajo: sure | 08:18 |
| *** iwamoto has quit IRC | 08:18 | |
| ihrachyshka | moshele, hm, isn't it working? have you tried to rebase? | 08:19 |
| ihrachyshka | it worked for me (I made some late changes in Kuba's patch for agent integration before merge) | 08:19 |
| jschwarz | ihrachyshka, functional? | 08:19 |
| jschwarz | ihrachyshka, where we're going, we don't need... functional... | 08:19 |
| jschwarz | (movie reference) | 08:19 |
| ihrachyshka | jschwarz, ? what do you mean? we need both. | 08:19 |
| jschwarz | lol | 08:19 |
| moshele | ihrachyshka: the delete is working? | 08:20 |
| ihrachyshka | moshele, it did for me. | 08:20 |
| jschwarz | ihrachyshka, https://www.youtube.com/watch?v=flge_rw6RG0 | 08:20 |
| ihrachyshka | I don't know, maybe I was sleeping | 08:20 |
| ihrachyshka | thought in that case, who uploaded a new patchset? | 08:20 |
| jlibosva | ihrachyshka: moshele hi, thanks for handling the notifications patch! | 08:21 |
| *** changzhi has quit IRC | 08:21 | |
| jlibosva | ajo: jschwarz how can I help with testing? | 08:22 |
| jschwarz | jlibosva, working on fullstack so I've got that under control atm | 08:22 |
| jschwarz | rumor has it that more functional tests are needed, though | 08:23 |
| *** lajos-katona has joined #openstack-neutron | 08:23 | |
| *** scheuran has joined #openstack-neutron | 08:24 | |
| *** changzhi has joined #openstack-neutron | 08:24 | |
| ihrachyshka | jschwarz, I assumed ajo will handle those, but we'll see what he thinks about it | 08:24 |
| *** john-davidge has joined #openstack-neutron | 08:24 | |
| moshele | ihrachyshka: I see how this could have work with ovs, it because you assume one rule only | 08:25 |
| moshele | ihrachyshka: I made it more generic to support future rules | 08:26 |
| *** matrohon has joined #openstack-neutron | 08:26 | |
| *** claudiub has joined #openstack-neutron | 08:27 | |
| *** john-davidge_ has joined #openstack-neutron | 08:27 | |
| *** john-davidge has quit IRC | 08:29 | |
| *** john-davidge_ is now known as john-davidge | 08:29 | |
| jlibosva | ihrachyshka: Is there a patch that removes resource manager's constructor? | 08:29 |
| *** abregman|brb is now known as abregman | 08:29 | |
| ihrachyshka | jlibosva, it was included in you patch and merged | 08:29 |
| jlibosva | ihrachyshka: cool! thanks | 08:30 |
| ihrachyshka | jlibosva, notifications in, it seems to work, we have first functional tests in review. | 08:30 |
| ihrachyshka | I believe we'll start merge process today in the evening if all goes well | 08:30 |
| ajo | jschwarz, https://review.openstack.org/210012 | 08:34 |
| *** shoutm has quit IRC | 08:34 | |
| ajo | I was reviewing and then starting update cases, unless you want to take on them quickly, we can divide work (update & delete) | 08:34 |
| ihrachyshka | jlibosva, if you have spare cycles, can you check why https://review.openstack.org/#/c/211003/ fails? I may not have time for that in next 2 hours. | 08:34 |
| ajo | I can make delete ready for moshele | 08:34 |
| ajo | and you can work on update, | 08:34 |
| ajo | anyway, it should be easy once the framework is in place | 08:35 |
| jlibosva | ihrachyshka: ack, I'm gonna also send the patch that lock on agent side | 08:35 |
| ihrachyshka | jlibosva, cool | 08:35 |
| ihrachyshka | jlibosva, please rebase on top of ajo's work so that it's actually tested in some reasonable way :) | 08:35 |
| jlibosva | ihrachyshka: oki doki | 08:35 |
| moshele | ajo: like ihrachyshka | 08:36 |
| moshele | said delete is already working I am just moving some part to the qos driver interface to avoid dup | 08:36 |
| ajo | moshele, is the patch up? | 08:36 |
| moshele | ajo: it merged | 08:36 |
| ajo | ahhh | 08:36 |
| openstackgerrit | Sridhar Gaddam proposed openstack/neutron: Avoid dhcp_release for ipv6 addresses https://review.openstack.org/209456 | 08:36 |
| ajo | ack, sorry | 08:36 |
| ajo | I missunderstood | 08:36 |
| jschwarz | ajo, what about it? | 08:36 |
| ajo | jschwarz, so, what do you want to functionally test? delete or update? | 08:37 |
| ajo | I'll take the other one ;) | 08:37 |
| jschwarz | ajo, I don't want to do anything ;-) | 08:37 |
| jschwarz | ajo, I'm busy with the fullstack tests atm | 08:37 |
| jschwarz | so take one and I'll do the other one if it's not too late when I get to it | 08:37 |
| *** haigang has joined #openstack-neutron | 08:40 | |
| ajo | jschwarz, ack, I thought you ran out of fullstacks and needed something else to look at ;) | 08:40 |
| jschwarz | ajo, was under the impression that jlibosva was looking for something to do ;)? | 08:40 |
| jlibosva | I still am | 08:41 |
| *** shihanzhang has joined #openstack-neutron | 08:41 | |
| jschwarz | ajo, say I have a neutron port and an ovs port | 08:41 |
| jschwarz | and I want to manually plug them to one another | 08:42 |
| jschwarz | how do I go about it? | 08:42 |
| haigang | hi there | 08:43 |
| haigang | I got a problem when using neutron with provider network without l3 agent | 08:43 |
| haigang | only controller and compute node, no network node. when boot an instance, nova will inject public key, vm can't access http://169.254.169.254:80 | 08:44 |
| *** hanzhang has quit IRC | 08:44 | |
| haigang | is there anyone have any idea on this? | 08:45 |
| jschwarz | haigang, do you have a DHCP agent attached to the network? | 08:45 |
| jschwarz | ihrachyshka, jlibosva, perhaps you'll know re: what I asked a bit above? | 08:45 |
| haigang | yes, dhcp service is in controller node | 08:45 |
| jschwarz | haigang, OK. so you need to configure an option so that it will know to service metadata | 08:46 |
| haigang | as I know, neutron-s-metadata-proxy will handle the http://169.254.169.254 and redirect it to nova-api | 08:46 |
| jschwarz | haigang, in the dhcp agent configuration. look for 'enable_isolated_metadata' - it should be set to True | 08:46 |
| *** tfukushima has joined #openstack-neutron | 08:47 | |
| ihrachyshka | ok guys, I leave you alone for now, please behave ;) | 08:47 |
| haigang | ok, I'll check on it. But I tried to capture the packet instance send out, vm send http request without http header such as instance-id/ tenant-id | 08:48 |
| haigang | neutron-ns-metadata-proxy will add addition informal and proxy request to nova-api, by now, without proxy, how to addition these information in http request? | 08:48 |
| jschwarz | haigang, I'm not very familiar with the request itself - but I think the problem is that the packet doesn't reach the metadata proxy at all because the option is not set | 08:50 |
| haigang | jschwarz, I've tried use iptables on compute and make a dnat rule, to make the http://169.254.169.254:80 request redirect to nova-api, it works. But nova-api response 400 because of less http headers such as instance-id/tenant id | 08:52 |
| jschwarz | haigang, like I said, I don't know exactly how the packets are supposed to look. I do know that if the option is set, the dhcp agent is supposed to inject the routing information correctly on its own | 08:53 |
| *** ihrachyshka has quit IRC | 08:53 | |
| jschwarz | then it should work properly :) | 08:53 |
| openstackgerrit | Pavel Bondar proposed openstack/neutron: [DO NOT MERGE] Enable pluggable IPAM implementation by default https://review.openstack.org/181023 | 08:53 |
| haigang | jschwarz, ok, I'll test first, thx! | 08:53 |
| jschwarz | :) | 08:53 |
| openstackgerrit | Lajos Katona proposed openstack/python-neutronclient: Support CLI changes for managing trunk ports https://review.openstack.org/210301 | 08:54 |
| haigang | btw, I'm a freshman in IRC. I don't know how to add ur name in our chat :( | 08:54 |
| jlibosva | guys, I need to push an amended patch but I get this - http://paste.openstack.org/ | 08:54 |
| jlibosva | jschwarz: ajo moshele ^^ | 08:54 |
| jschwarz | jlibosva, that paste is empty | 08:54 |
| jschwarz | ^_^ | 08:54 |
| jlibosva | do you know if it's safe to push? I don't get why it does that - as you can see the patch is based on HEAD | 08:55 |
| jlibosva | jschwarz: yes, I made it empty so we don't have it that easy :) http://paste.openstack.org/show/412324/ | 08:55 |
| jschwarz | jlibosva, I won't go ahead with the submission if I were you | 08:56 |
| jlibosva | based on master HEAD* | 08:56 |
| jschwarz | it looks like you're merging to master | 08:56 |
| jlibosva | jschwarz: I didn't but I'd like to update Ihar's patch | 08:56 |
| jlibosva | jschwarz: indeed, but .gitreview says it's qos | 08:56 |
| *** neophy has joined #openstack-neutron | 08:57 | |
| jschwarz | jlibosva, so you're trying to --amend a patch of Ihar's on feature/qos? | 08:57 |
| jschwarz | and that's what you get after running git-review? | 08:57 |
| *** shausy has joined #openstack-neutron | 08:57 | |
| jlibosva | jschwarz: hm, weird, when I explicitly said to push to feature/qos it passed ... | 08:58 |
| jschwarz | jlibosva, maybe it's trying to rebase on feature/qos | 08:58 |
| jschwarz | erm | 08:58 |
| jschwarz | I was about to suggest rebasing on latest feature/qos but OK :) | 08:59 |
| jlibosva | ajo: I fixed https://review.openstack.org/#/c/211003/ | 08:59 |
| jschwarz | jlibosva, anyway do you know how do I go about manually attaching a neutron port to an ovs port? | 08:59 |
| jschwarz | moshele, ^ | 08:59 |
| jlibosva | jschwarz: well, it is rebased on feature/qos but git-review treid to push it to master .. | 08:59 |
| ajo | thanks jlibosva :) | 08:59 |
| *** yushiro has quit IRC | 08:59 | |
| jschwarz | lol | 08:59 |
| jlibosva | jschwarz: sec | 09:00 |
| jlibosva | jschwarz: maybe this will help? https://review.openstack.org/#/c/191147/7/neutron/tests/common/conn_testers.py L276 | 09:01 |
| jschwarz | jlibosva, maybe :) thanks | 09:02 |
| jlibosva | jschwarz: you meant on agent side, right? | 09:02 |
| jschwarz | jlibosva, I meant on fullstack side | 09:02 |
| jschwarz | but yes :) | 09:02 |
| jlibosva | jschwarz: hm, fullstack likely requires db change as well | 09:02 |
| jschwarz | jlibosva, erm... what kind of db change? | 09:03 |
| jlibosva | jschwarz: port binding table update ... I guess | 09:03 |
| jlibosva | jschwarz: Assaf talked about some "real vm" for fullstack | 09:03 |
| jlibosva | maybe Cedric would know | 09:03 |
| jschwarz | jlibosva, yeah, they have that class all written but they missed that part | 09:03 |
| *** afazekas_ is now known as afazekas | 09:04 | |
| *** shausy has quit IRC | 09:05 | |
| *** shausy has joined #openstack-neutron | 09:05 | |
| *** shoutm has joined #openstack-neutron | 09:09 | |
| *** tiswanso has joined #openstack-neutron | 09:09 | |
| jlibosva | jschwarz: aaah | 09:12 |
| *** mohankumar has joined #openstack-neutron | 09:12 | |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: Service plugin for trunk ports https://review.openstack.org/210311 | 09:14 |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: DB model for trunk ports https://review.openstack.org/210310 | 09:14 |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: API extension https://review.openstack.org/210309 | 09:14 |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: ML2 extension driver for subports https://review.openstack.org/210313 | 09:14 |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: DB model for subports https://review.openstack.org/210312 | 09:14 |
| *** mkoderer has quit IRC | 09:14 | |
| *** klindgren__ has joined #openstack-neutron | 09:15 | |
| *** trinaths has quit IRC | 09:16 | |
| *** mkoderer has joined #openstack-neutron | 09:17 | |
| *** thumpba has joined #openstack-neutron | 09:17 | |
| *** klindgren_ has quit IRC | 09:17 | |
| *** trinaths has joined #openstack-neutron | 09:18 | |
| *** tiswanso has quit IRC | 09:20 | |
| *** thumpba has quit IRC | 09:22 | |
| *** abregman is now known as abregman|lunch | 09:23 | |
| *** alexpilotti has joined #openstack-neutron | 09:25 | |
| jschwarz | jlibosva, AHAH! OVSInterfaceDriver._ovs_add_port! :) | 09:27 |
| *** jyuso1 has quit IRC | 09:27 | |
| *** jyuso1 has joined #openstack-neutron | 09:28 | |
| *** _kiran_ has joined #openstack-neutron | 09:28 | |
| *** salv-orlando has joined #openstack-neutron | 09:29 | |
| *** kiran-r has quit IRC | 09:31 | |
| *** yamamoto has quit IRC | 09:32 | |
| *** anilvenkata has quit IRC | 09:35 | |
| haigang | jschwarz, yes, the dhcp agent enable_isolated_metadata works, vm's http://169.254.169.254 reach the nova-api metadata service | 09:37 |
| *** kawa2014 has quit IRC | 09:38 | |
| *** kawa2014 has joined #openstack-neutron | 09:39 | |
| *** yalie has quit IRC | 09:39 | |
| *** yalie has joined #openstack-neutron | 09:39 | |
| *** markvoelker has joined #openstack-neutron | 09:43 | |
| *** markvoelker has quit IRC | 09:49 | |
| *** zhhuabj has joined #openstack-neutron | 09:49 | |
| *** sambetts has quit IRC | 09:49 | |
| *** sambetts has joined #openstack-neutron | 09:54 | |
| *** fkobzik has quit IRC | 09:55 | |
| *** itzikb_ has quit IRC | 09:56 | |
| *** zhhuabj has quit IRC | 09:56 | |
| *** gampel has quit IRC | 09:58 | |
| *** salv-orlando has quit IRC | 09:58 | |
| *** shoutm has quit IRC | 09:58 | |
| *** salv-orlando has joined #openstack-neutron | 09:58 | |
| *** BhavyaM has quit IRC | 10:00 | |
| *** yfried is now known as yfried|afk | 10:00 | |
| moshele | ajo: do we handle the case when port which belong to policy is deleted? | 10:02 |
| *** yamamoto has joined #openstack-neutron | 10:03 | |
| ajo | moshele I guess that may be implementation specific, for example, ovs doesn't need to do anything, but probably SR-IOV will need to clear out the port? | 10:03 |
| ajo | (I'm just guessing the VF will remain, with the limit set) | 10:03 |
| *** sbelous has quit IRC | 10:04 | |
| *** eezhova has quit IRC | 10:04 | |
| *** eezhova has joined #openstack-neutron | 10:06 | |
| moshele | ajo: ok I will reset it on delete | 10:07 |
| moshele | ajo: for SR-IOV | 10:07 |
| ajo | moshele, now I'm unsure if may be ovs needs to cleanup the queues, or if they will go away with the port | 10:08 |
| ajo | gsagie ^ | 10:08 |
| ajo | can you check the las 6-7 lines of conversation between moshele and me ? :) | 10:08 |
| irenab | apuimedo: hey | 10:08 |
| *** abregman|lunch is now known as abregman | 10:09 | |
| *** gampel has joined #openstack-neutron | 10:09 | |
| jschwarz | haigang, glad to be of service :) | 10:09 |
| openstackgerrit | Ilya Shakhat proposed openstack/python-neutronclient: Remove newlines from request and response log https://review.openstack.org/211085 | 10:10 |
| *** sambetts has quit IRC | 10:10 | |
| haigang | jschwarz, but still got the 500 error, due to the http header problem~~ | 10:11 |
| *** sbelous has joined #openstack-neutron | 10:11 | |
| apuimedo | irenab: hey | 10:11 |
| *** alex_klimov has joined #openstack-neutron | 10:11 | |
| jschwarz | haigang, :( | 10:11 |
| jschwarz | haigang, how are you trying to access the metadata server? curl? | 10:11 |
| irenab | apuimedo: is there kuryr meeting today? | 10:12 |
| haigang | no~, just boot a vm~ and capture the packet, find the http response is 500 | 10:12 |
| *** zhhuabj has joined #openstack-neutron | 10:12 | |
| *** gongysh has quit IRC | 10:13 | |
| jschwarz | haigang, try accessing it manually using curl | 10:13 |
| haigang | the same result because of less http head x-instance-id | 10:14 |
| *** yushiro has joined #openstack-neutron | 10:14 | |
| apuimedo | irenab: there should be | 10:14 |
| apuimedo | I'll try to keep alive until then | 10:14 |
| irenab | apuimedo: :-) | 10:14 |
| *** vthapar has quit IRC | 10:14 | |
| *** sambetts has joined #openstack-neutron | 10:16 | |
| openstackgerrit | Sean M. Collins proposed openstack/neutron: [WIP] Linux Bridge: Create a configurable network manager class https://review.openstack.org/209385 | 10:16 |
| openstackgerrit | Sean M. Collins proposed openstack/neutron: [WIP] Linux Bridge Agent Restructuring https://review.openstack.org/208666 | 10:16 |
| gsagie | apuimedo: you feeling bad? | 10:16 |
| jschwarz | moshele, ping | 10:16 |
| moshele | jschwarz: hi | 10:16 |
| jschwarz | moshele, good morning :) | 10:17 |
| apuimedo | gsagie: I've been ill the whole weekend, still a bit crushed | 10:17 |
| apuimedo | but working | 10:17 |
| jschwarz | moshele, so I've created a neutron port attached to an ovs port and updated the port to have a qos policy | 10:17 |
| moshele | jschwarz: is it still morning :) | 10:17 |
| jschwarz | moshele, how do I check if the bandwidth is applied to the ovs port? | 10:17 |
| jschwarz | moshele, I woke up not too long ago ;-) | 10:18 |
| gsagie | apuimedo : hope it improves soon :) wanted to talk with you about the vif binding, but lets do it in the meeting | 10:18 |
| moshele | jschwarz: this question is for gsagie | 10:18 |
| jschwarz | gsagie, ^^ :) | 10:18 |
| gsagie | ajo: the QoS profile and Queue must be deleted when the port is deleted its not done automatically at least from looking at the schema | 10:19 |
| gsagie | jschwarz: you can dump Open_vSwitch db and see that the QoS profile is set for that port | 10:20 |
| moshele | ajo: we should add delete api here https://github.com/openstack/neutron/blob/feature/qos/neutron/agent/l2/extensions/manager.py#L64-L65 to be called from agent on remove | 10:20 |
| *** zhhuabj has quit IRC | 10:20 | |
| *** yfried|afk is now known as yfried | 10:20 | |
| jschwarz | gsagie, do you have a quick command to do that? | 10:20 |
| ajo | ok, moshele , to delete a port, right? | 10:21 |
| gsagie | jschwarz: ovsdb-client dump Open_vSwitch | 10:21 |
| moshele | ajo: yes | 10:21 |
| jschwarz | thanks | 10:21 |
| ajo | gsagie, do you have time to coordinate with moshele and get the port_delete for OVS&SR-IOV fast? | 10:21 |
| moshele | ajo: and as gsagie | 10:22 |
| moshele | it should be used in the ovs as well | 10:22 |
| ajo | you're always fast, but I'm asking due to merge-back | 10:22 |
| ajo | it's a resource-deallocation bug which could be handled as a bugfix I guess, but, better if we merge with that cleared up I guess | 10:23 |
| gsagie | ajo: this change is only in the agent right? | 10:23 |
| ajo | correct | 10:23 |
| ajo | gsagie: only agent side | 10:23 |
| gsagie | we dont need specific code for SR-IOV or OVS agent | 10:23 |
| ajo | gsagie, only in the extension and extension manager | 10:23 |
| ajo | the ovs-agent / sr-iov agents may need to pass down the port delete to the extension manager I guess | 10:24 |
| moshele | gsagie: I tand plug it in the correct location of each agent | 10:24 |
| ajo | I asume moshele Is looking at that | 10:24 |
| moshele | ajo: yep | 10:24 |
| gsagie | ok, i am a bit busy with something else but lets see how it goes tommorow and speak | 10:25 |
| *** itzikb_ has joined #openstack-neutron | 10:25 | |
| gsagie | the question is, if the extension has all the rules configured on that port so it can direct the delete to it | 10:26 |
| *** gongysh has joined #openstack-neutron | 10:26 | |
| gsagie | or maybe a simpler solution would be to extend the API to support port deletion | 10:26 |
| ajo | yeah, I'd handle port deletion | 10:26 |
| ajo | then we just delete any queue related to the deleted port | 10:26 |
| gsagie | (i mean extend the driver API) | 10:26 |
| ajo | gsagie, If you're ok with it, I'll look at OVS part today with moshele if I have enough time, and sync with you tomorrow | 10:27 |
| gsagie | ajo: sure np | 10:27 |
| ajo | thanks gsagie :) | 10:28 |
| gsagie | its still not clear to me if you are going to do what i suggested, meaning the driver itself will need to support port deletion OR the extension manager will support port deletion and calls all the extension drivers delete method | 10:28 |
| *** salv-orl_ has joined #openstack-neutron | 10:28 | |
| gsagie | so we will have port_delete for OVS and SRIOV | 10:28 |
| gsagie | each in its own driver | 10:29 |
| ajo | gsagie, I think port_delete is the way to go | 10:29 |
| ajo | each on it's own driver | 10:29 |
| *** jasondotstar has joined #openstack-neutron | 10:29 | |
| ajo | so they can cleanup any port-related resources when the port is going to be deleted | 10:29 |
| gsagie | ajo: ok, thats probably the simpler solution right now | 10:29 |
| apuimedo | gsagie: diga and tfukushima are here, they all want to talk about the vif binding | 10:29 |
| ajo | gsagie: and could cover the case for other different extensions, I guess | 10:29 |
| ajo | moshele ^ | 10:30 |
| apuimedo | I don't know if diga will be able to join the meeting (it may be quite late in his tz) | 10:30 |
| ajo | so adding port_delete all the way down to the drivers ? | 10:30 |
| ajo | (just to check we're all in the same page) | 10:30 |
| gsagie | ajo: then all you need to do, is just to call the OVS delete method with the port-name | 10:30 |
| gsagie | in OVS part | 10:30 |
| gsagie | ajo: yes, thats what i think | 10:30 |
| *** salv-orlando has quit IRC | 10:30 | |
| ajo | gsagie: amazing, thanks for the pointer :) | 10:31 |
| irenab | gsagie: apuimedo |s there any etherpad started for vif binding? | 10:31 |
| gsagie | ajo: del_qos_bw_limit_for_port(port_name), please mind that there is a patch working on roll back for this: https://review.openstack.org/#/c/209200/13 | 10:31 |
| irenab | gsagie: apuimedo :maybe worth to catch up before the meeting to make it more productive | 10:31 |
| ajo | gsagie, ack, I saw it yes, but it should be independent, right? | 10:32 |
| gsagie | ajo: just direct the delete API to this method, it is also converted to be "safe" in case no QoS is actually defined | 10:32 |
| gsagie | ajo: yeah | 10:32 |
| ajo | ahh, perfect | 10:32 |
| moshele | ajo: can you review https://review.openstack.org/#/c/211090/ ? | 10:33 |
| ajo | gsagie: that's nice, since we can statelessly clean any port, even if the agent restarted and does not know about previously applied qos policies, we clean up | 10:33 |
| *** zhhuabj has joined #openstack-neutron | 10:33 | |
| gsagie | irenab: sure we can do it | 10:33 |
| apuimedo | irenab: I do not remember if we did have it | 10:33 |
| irenab | gsagie: just to put ideas and make meeting more focused | 10:34 |
| gsagie | ajo: yes, i keep the port_id as an external_id of the QoS/Queue entries in OVSDB so they are easly deleted by port id even if the port is already deleted | 10:34 |
| *** vikram has quit IRC | 10:35 | |
| *** zhhuabj has quit IRC | 10:36 | |
| gsagie | irenab, apuimedo: sure, want me to open the etherpad? | 10:36 |
| *** pc_m has joined #openstack-neutron | 10:36 | |
| tfukushima | apuimedo, gsagie, irenab: I talked to diga and he will work on the entire logics for /NetworkDriver.Join and /NetworkDriver.Leave including the vif, an endpoint of the veth pair, binding. | 10:36 |
| ajo | gsagie: have a quick look to L64 here: https://review.openstack.org/#/c/211090/1/neutron/agent/l2/extensions/qos.py I guess that works | 10:36 |
| *** yushiro has quit IRC | 10:36 | |
| ajo | delete with policy = None ? :) | 10:36 |
| tfukushima | So I leave them for him. | 10:37 |
| ajo | and we got called to delete any kind of rule type on the port... | 10:37 |
| irenab | gsagie: tfukushima: Cool. Then probably we need diga to share details | 10:37 |
| diga | thanks tfukushima | 10:38 |
| tfukushima | NP. | 10:38 |
| gsagie | ajo: looks good to me, then just call this from the OVS agent on port delete | 10:40 |
| *** josecastroleon has quit IRC | 10:41 | |
| ajo | thanks for the feedback gsagie , I'll mask other interruptions for today as you told me you'll be busy :) | 10:41 |
| gsagie | diga: then it will be good if you can please share the TODO's and plans on an etherpad and maybe we can help out, unless you think its all covered by you | 10:41 |
| gsagie | which is also good :) | 10:41 |
| gsagie | ajo: gampel is keeping me busy :) | 10:42 |
| diga | gsagie: Sure, I will create a etherpad for the tasks I am working on | 10:42 |
| apuimedo | diga: thanks! | 10:42 |
| diga | I will absolutely appreciate your help | 10:43 |
| diga | hey apuimedo | 10:43 |
| diga | welcome! | 10:43 |
| ajo | hehe gsagie , gampel , that's life ;) | 10:43 |
| gsagie | ajo: but let me know if you encounter any problems and i will let you know, i am going to review Nate's patch, found the problem with his code | 10:43 |
| gsagie | and why its failing the tests | 10:43 |
| ajo | thanks gsagie :) | 10:43 |
| *** scheuran has quit IRC | 10:47 | |
| *** abregman is now known as abregman|brb | 10:48 | |
| *** longguang has quit IRC | 10:49 | |
| apuimedo | diga: hey ;-) | 10:50 |
| *** baoli has joined #openstack-neutron | 10:51 | |
| *** kiran-r has joined #openstack-neutron | 10:51 | |
| *** yalie has quit IRC | 10:52 | |
| *** jaypipes has joined #openstack-neutron | 10:53 | |
| *** carl_baldwin has quit IRC | 10:54 | |
| *** baoli_ has joined #openstack-neutron | 10:54 | |
| *** _kiran_ has quit IRC | 10:55 | |
| *** yamamoto has quit IRC | 10:55 | |
| *** mwagner_lap has quit IRC | 10:57 | |
| *** baoli has quit IRC | 10:57 | |
| diga | apuimedo: I am working on binding/unbinding VIF as discussed earlier | 11:02 |
| apuimedo | nice | 11:03 |
| openstackgerrit | vikram.choudhary proposed openstack/neutron: [WIP] BGP Dynamic Routing: introduce entities and model https://review.openstack.org/201621 | 11:03 |
| haigang | if there's no neutron-ns-metadata-proxy , how to handle the vm metadata service request? | 11:03 |
| *** scheuran has joined #openstack-neutron | 11:05 | |
| *** changzhi has quit IRC | 11:06 | |
| *** tfukushima has quit IRC | 11:08 | |
| *** salv-o___ has joined #openstack-neutron | 11:11 | |
| *** fawadkhaliq has joined #openstack-neutron | 11:11 | |
| *** salv-orl_ has quit IRC | 11:11 | |
| *** josecastroleon has joined #openstack-neutron | 11:11 | |
| *** Marga_ has quit IRC | 11:12 | |
| *** dtynan has quit IRC | 11:13 | |
| jschwarz | gsagie, ping | 11:14 |
| *** tiswanso has joined #openstack-neutron | 11:16 | |
| *** Marga_ has joined #openstack-neutron | 11:16 | |
| *** jasondotstar has quit IRC | 11:18 | |
| *** rossella_s has quit IRC | 11:25 | |
| *** tiswanso has quit IRC | 11:27 | |
| *** gcossu has joined #openstack-neutron | 11:27 | |
| *** sdake has joined #openstack-neutron | 11:28 | |
| *** markvoelker has joined #openstack-neutron | 11:29 | |
| *** shoutm has joined #openstack-neutron | 11:30 | |
| *** Marga_ has quit IRC | 11:31 | |
| *** ihrachyshka has joined #openstack-neutron | 11:32 | |
| *** sdake_ has joined #openstack-neutron | 11:32 | |
| *** kiran-r has quit IRC | 11:32 | |
| *** haigang has quit IRC | 11:33 | |
| *** Marga_ has joined #openstack-neutron | 11:34 | |
| *** markvoelker has quit IRC | 11:34 | |
| *** haigang has joined #openstack-neutron | 11:35 | |
| *** yfried is now known as yfried|afk | 11:35 | |
| *** sdake has quit IRC | 11:36 | |
| *** sudipto has quit IRC | 11:37 | |
| *** ddepaoli has joined #openstack-neutron | 11:38 | |
| gcossu | hello ihrachyshka ajo | 11:39 |
| ajo | hi gcossu | 11:40 |
| *** haigang has quit IRC | 11:40 | |
| *** thorst has joined #openstack-neutron | 11:40 | |
| gcossu | how is going the ovs agent development? | 11:41 |
| ajo | gcossu, moshele is doing some final refactors and fixups to the qos extension | 11:42 |
| ajo | to properly cleanup resources at port deletion | 11:42 |
| *** haigang has joined #openstack-neutron | 11:42 | |
| *** amuller has joined #openstack-neutron | 11:47 | |
| moshele | ajo: the delete port in the ovs https://review.openstack.org/#/c/211117/ | 11:48 |
| *** julim has joined #openstack-neutron | 11:48 | |
| *** sdake has joined #openstack-neutron | 11:49 | |
| moshele | ajo: I am not sure I plug it in the right place I am not expert in OVS, also I have not test it except the unit test | 11:49 |
| *** scheuran has quit IRC | 11:51 | |
| *** sdake_ has quit IRC | 11:52 | |
| ajo | moshele, looks like the right place to put the hook | 11:52 |
| *** vikram has joined #openstack-neutron | 11:53 | |
| amuller | haleyb: ping https://review.openstack.org/#/c/154043/ | 11:57 |
| ajo | gsagie, ping, I found something weird | 11:58 |
| ajo | ping me if you have a moment | 11:58 |
| ajo | gsagie: shouldn't test_simple_port_policy_update(native) not go through the vsctl path ? | 11:59 |
| *** yfried|afk is now known as yfried | 11:59 | |
| ajo | amuller: can you take a look at https://review.openstack.org/#/c/210012/ when you have a few spare minutes? :) | 11:59 |
| ajo | it's functional testing for the port creation | 12:00 |
| *** mohankumar has quit IRC | 12:00 | |
| amuller | ajo: ok just entering a meeting I'll try to take a look after | 12:00 |
| ajo | I extracted OVSAgentTestFramework to neutron/tests/functional/agent/l2/base.py | 12:00 |
| ajo | ack amuller , thanks | 12:00 |
| ajo | gsagie: http://fpaste.org/253378/20793614/ | 12:01 |
| *** dave-mccowan has joined #openstack-neutron | 12:01 | |
| *** abregman|brb is now known as abregman | 12:02 | |
| *** baoli_ has quit IRC | 12:03 | |
| *** lyanchih has joined #openstack-neutron | 12:03 | |
| openstackgerrit | Sean M. Collins proposed openstack/neutron: [WIP] Linux Bridge: Create a configurable network manager class https://review.openstack.org/209385 | 12:07 |
| openstackgerrit | Sean M. Collins proposed openstack/neutron: [WIP] Linux Bridge Agent Restructuring https://review.openstack.org/208666 | 12:07 |
| *** scheuran has joined #openstack-neutron | 12:07 | |
| *** markvoelker has joined #openstack-neutron | 12:08 | |
| *** Miouge has joined #openstack-neutron | 12:08 | |
| sc68cal | scheuran: ping | 12:08 |
| *** tfukushima has joined #openstack-neutron | 12:09 | |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: Service plugin for trunk ports https://review.openstack.org/210311 | 12:09 |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: DB model for trunk ports https://review.openstack.org/210310 | 12:09 |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: API extension https://review.openstack.org/210309 | 12:09 |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: ML2 extension driver for subports https://review.openstack.org/210313 | 12:09 |
| openstackgerrit | Bence Romsics proposed openstack/neutron: Trunk port: DB model for subports https://review.openstack.org/210312 | 12:09 |
| *** rossella_s has joined #openstack-neutron | 12:11 | |
| *** mkolesni__ has quit IRC | 12:12 | |
| openstackgerrit | Madhu Mohan proposed openstack/neutron: Redundant tests removed from ovs-lib unit tests: https://review.openstack.org/208860 | 12:12 |
| *** mkolesni__ has joined #openstack-neutron | 12:12 | |
| openstackgerrit | Ramanjaneya Reddy Palleti proposed openstack/python-neutronclient: Support RBAC neutron-client changes. https://review.openstack.org/205482 | 12:13 |
| *** tfukushima has quit IRC | 12:14 | |
| *** annegentle has joined #openstack-neutron | 12:20 | |
| *** txomon|fon has quit IRC | 12:22 | |
| *** itzikb_ has quit IRC | 12:23 | |
| *** tiswanso has joined #openstack-neutron | 12:23 | |
| *** busterswt has joined #openstack-neutron | 12:23 | |
| *** salv-o___ has quit IRC | 12:26 | |
| *** johnbelamaric has joined #openstack-neutron | 12:26 | |
| *** salv-orlando has joined #openstack-neutron | 12:26 | |
| *** itzikb_ has joined #openstack-neutron | 12:29 | |
| *** trinaths has quit IRC | 12:31 | |
| *** oomichi has quit IRC | 12:31 | |
| openstackgerrit | Sean M. Collins proposed openstack/neutron: [WIP] Linux Bridge: Create a configurable network manager class https://review.openstack.org/209385 | 12:31 |
| *** moshele has quit IRC | 12:32 | |
| *** yfried is now known as yfried|afk | 12:32 | |
| gsagie | ajo: the problem is that the delete method fails when the port doesnt exists (you need Nate patch and another fix in it to make it safe) | 12:32 |
| ajo | gsagie I think I found the issue | 12:32 |
| ajo | when we ovsctl and native separatelly, they both works | 12:32 |
| gsagie | ajo: if you want you could call "get" inside the delete and return elsewhere, and the OVS tests always runs both for OVSDB and the CLI | 12:32 |
| *** johnbelamaric has quit IRC | 12:32 | |
| ajo | if we test them together the 2nd always fails, because it still remembers the old port from the previous test | 12:33 |
| ajo | and tries to delete policy | 12:33 |
| ajo | somehow the manager is not stateless... | 12:33 |
| ajo | gsagie | 12:34 |
| *** yfried|afk is now known as yfried | 12:34 | |
| *** tiswanso has quit IRC | 12:34 | |
| ajo | I inverted the tests, and now 2nd fails | 12:34 |
| ajo | but look: http://fpaste.org/253385/43921006/ | 12:34 |
| ajo | again, this one is now reporting as "vsctl" , but it looks like native | 12:34 |
| ajo | File "neutron/agent/ovsdb/impl_idl.py", line 70, in commit | 12:34 |
| ajo | raise result.ex | 12:34 |
| ajo | neutron.agent.ovsdb.native.idlutils.RowNotFound: Cannot find Port with name=tap24896927653 | 12:34 |
| ajo | gsagie, are they reporting inverted? | 12:35 |
| ajo | or am I getting something wrong? | 12:35 |
| *** txomon|fon has joined #openstack-neutron | 12:36 | |
| *** bapalm has joined #openstack-neutron | 12:36 | |
| ihrachyshka | ajo, how's the progress on more functional scenarios? | 12:36 |
| *** tfukushima has joined #openstack-neutron | 12:37 | |
| ajo | ihrachyshka, I'm finding some testing issues in update | 12:37 |
| ajo | ihrachyshka update is working itself | 12:37 |
| ihrachyshka | bad; good! | 12:37 |
| ajo | ihrachyshka, but when we test the vsctl / native scenarios, the 2nd always fail | 12:37 |
| ihrachyshka | ajo, so maybe cleanup is not solid? | 12:37 |
| ajo | ihrachyshka, yes, may be cleanup related, I'm just looking at it | 12:37 |
| ajo | ihrachyshka: or singleton related | 12:37 |
| ihrachyshka | execute one, check for QoS stuff in ovsdb-client | 12:38 |
| ajo | ihrachyshka: qos stuff is cleared out | 12:38 |
| ajo | sorry, no, that's what moshele is working on | 12:38 |
| ihrachyshka | maybe some singleton caches something | 12:38 |
| ajo | but the error is that the agent is somehow remembering an old port, that it's not there yer anymore | 12:38 |
| ajo | yes | 12:38 |
| ajo | ihrachyshka, no worries, I'm on it | 12:38 |
| *** edmondsw has joined #openstack-neutron | 12:38 | |
| ihrachyshka | thanks :) | 12:39 |
| ajo | just understood what the issue was | 12:39 |
| ajo | ihrachyshka: thank you! ;) | 12:39 |
| *** dtynan has joined #openstack-neutron | 12:41 | |
| *** Ramanjaneya has joined #openstack-neutron | 12:41 | |
| ajo | ha! | 12:41 |
| ajo | ihrachyshka, I think I got it | 12:41 |
| ihrachyshka | tell me | 12:41 |
| ajo | ihrachyshka, old agent keeps subscribed to the update notifications | 12:41 |
| ajo | hence not cleanup from memory | 12:41 |
| ajo | when I send the push to update | 12:42 |
| ajo | it will try to fiddle with the old port, not anymore there | 12:42 |
| ajo | ihrachyshka, I'll add some cleanup to the base test class to cleanup the consumers | 12:42 |
| ajo | that's what the clear was for! | 12:42 |
| ihrachyshka | woohoo. awesome journey. | 12:42 |
| ajo | '':) | 12:42 |
| * ihrachyshka dances more (>'-')> <('-'<) ^(' - ')^ <('-'<) (>'-')> | 12:43 | |
| ihrachyshka | ajo, what do you think about https://review.openstack.org/#/c/207096/ | 12:43 |
| ajo | LOL, ihrachyshka is loosing his head :D | 12:43 |
| ihrachyshka | I am in great mood today | 12:43 |
| *** Ramanjaneya has quit IRC | 12:43 | |
| ajo | ihrachyshka++ | 12:43 |
| ajo | let me look | 12:43 |
| *** lyanchih has quit IRC | 12:43 | |
| *** annegentle has quit IRC | 12:44 | |
| jschwarz | ihrachyshka, ajo, jlibosva, https://review.openstack.org/#/c/202492/ | 12:44 |
| jschwarz | XD | 12:44 |
| ihrachyshka | I will refrain from dancing | 12:44 |
| ajo | ihrachyshka, thanks, I agree 207096 is clearing a yagni :) | 12:44 |
| ajo | ihrachyshka, dancing is good :D | 12:44 |
| *** annegentle has joined #openstack-neutron | 12:44 | |
| ihrachyshka | ok then (>'-')> <('-'<) ^(' - ')^ <('-'<) (>'-')> | 12:44 |
| ajo | woow jschwarz !!! I'm dancing too ;D | 12:45 |
| jschwarz | (>'-')> <('-'<) ^(' - ')^ <('-'<) (>'-')> | 12:45 |
| *** mwagner_lap has joined #openstack-neutron | 12:45 | |
| *** baoli has joined #openstack-neutron | 12:46 | |
| gsagie | (>'-')> <('-'<) ^(' - ')^ <('-'<) (>'-')> | 12:46 |
| jschwarz | so that's that | 12:46 |
| *** zhhuabj has joined #openstack-neutron | 12:46 | |
| jschwarz | am I needed somewhere else? :) | 12:46 |
| jschwarz | ihrachyshka, ajo, ^ | 12:46 |
| *** yamamoto has joined #openstack-neutron | 12:47 | |
| *** piyanai has joined #openstack-neutron | 12:47 | |
| ajo | jschwarz, did you test all ops? :) (port policy change, port policy removal, rule update ? ) :) | 12:47 |
| *** kawa2014 has quit IRC | 12:48 | |
| jschwarz | ajo, nope | 12:48 |
| openstackgerrit | vikram.choudhary proposed openstack/neutron: i[WIP] BGP Dynamic Routing: introduce dragent scheduler model https://review.openstack.org/207607 | 12:48 |
| jschwarz | wysiwyg | 12:48 |
| jschwarz | ajo, ihrachyshka only asked me for a simple scenario... | 12:49 |
| ajo | jschwarz, if you still have time, and we can go for those It'd be awesome. | 12:49 |
| ajo | ihrachyshka, what do you think? | 12:49 |
| jschwarz | also I didn't check to see that the actual transfer rate is limited | 12:49 |
| ajo | may be I'm wrong | 12:49 |
| ihrachyshka | jschwarz, you can follow up | 12:49 |
| jschwarz | ajo, on what side? | 12:49 |
| jschwarz | functional/fullstack? | 12:49 |
| ihrachyshka | it won't go now anyway, so you can continue on this one | 12:49 |
| ajo | jschwarz fullstack: I'm tackling functional :) | 12:49 |
| ihrachyshka | jschwarz, I don't believe we should test rate | 12:50 |
| *** yamamoto has quit IRC | 12:50 | |
| jschwarz | ihrachyshka, why not? I actually think we do | 12:50 |
| *** kawa2014 has joined #openstack-neutron | 12:50 | |
| ihrachyshka | jschwarz, well, it's more like testing ovs | 12:50 |
| jschwarz | erm | 12:50 |
| ajo | may be that's more for tempest | 12:50 |
| ihrachyshka | tempest could do though | 12:50 |
| ihrachyshka | yeah.. | 12:50 |
| ajo | like, validating that limiting happens regardless of the plugin ? | 12:50 |
| ajo | we can test manually now | 12:51 |
| ajo | and then, let tempest do it | 12:51 |
| jschwarz | aye, itzikb_ is working on the tempest test | 12:51 |
| ihrachyshka | I don't care about actual quality of service, I'm fine if I see magic numbers in ovsdb-client ;) | 12:51 |
| *** jpena is now known as jpena|lunch | 12:51 | |
| ihrachyshka | jschwarz, anyway, I think more scenarios is more important than actual rate measurement | 12:52 |
| *** jamie_h has joined #openstack-neutron | 12:52 | |
| *** neophy has quit IRC | 12:52 | |
| jschwarz | ihrachyshka, your wish is my command | 12:52 |
| ajo | jschwarz, I agree with ihrachyshka :) | 12:52 |
| jschwarz | so, which scenarios? create, update, delete...? | 12:52 |
| ajo | jschwarz: | 12:53 |
| ajo | 1) create, you did | 12:53 |
| ajo | 2) update port policy id | 12:53 |
| ajo | 3) update policy rule | 12:53 |
| ajo | 4) delete port qos_policy_id | 12:53 |
| ajo | 5) delete port with a policy and make sure heave doesn't explode? | 12:53 |
| ajo | heaven | 12:53 |
| jschwarz | erm | 12:54 |
| jschwarz | how does one check that heaven doesn't explode? XD\ | 12:54 |
| *** zhhuabj has quit IRC | 12:55 | |
| jschwarz | either way this isn't urgent (won't merge for a few weeks at least because it has multiple dependencies that aren't met) | 12:55 |
| ajo | jschwarz, not sure :D how does one know if heaven exist? | 12:55 |
| jschwarz | I'll write those scenarios but first I want to work a bit on the fullstack infra | 12:55 |
| ajo | jschwarz, yes, I agree on that, but could help us identify if there's something wrong with the service :) | 12:55 |
| jschwarz | yeps :) | 12:55 |
| *** julim has quit IRC | 12:56 | |
| ihrachyshka | jschwarz, from oslo_utils import heaven; self.assertFalse(heaven.exploded) | 12:56 |
| ajo | lol | 12:56 |
| jschwarz | lol! | 12:56 |
| ajo | may be it deserves it's own module | 12:56 |
| ajo | oslo_heaven | 12:56 |
| jschwarz | ihrachyshka, so now you can tell mestery that we're the first ones to also include fullstack tests | 12:56 |
| ajo | "" with all the functions you ever will need, implemented and debugged """ | 12:56 |
| ihrachyshka | jschwarz, yeah, we rock | 12:56 |
| jschwarz | let him say now that qos isn't ready for merge-back :) | 12:56 |
| ihrachyshka | ajo, yeah, oslo_heaven is like mock, but with actual implementation | 12:57 |
| ajo | LOL | 12:57 |
| ihrachyshka | ajo, you just call random attributes and it gives you what you want | 12:57 |
| ajo | ihrachyshka, and an ambiguous description of what you want, it may come back with more questions, of course.. | 12:57 |
| *** tongli has joined #openstack-neutron | 12:58 | |
| ihrachyshka | jschwarz, it's not just mestery to decide I suppose. I need to start preparing for communicating the merge. though looking into the current patch queue, I think we may need to postpone till tomorrow | 12:58 |
| ajo | my_func = oslo_heaven.implement('sum(a,b)', description='I want a function that adds two numbers') | 12:58 |
| ajo | "do you want me to store it somewhere, or may I return it?" | 12:58 |
| ajo | XD | 12:58 |
| *** bapalm_ has joined #openstack-neutron | 12:58 | |
| jschwarz | ihrachyshka, well all the patches that I'm an owner of can be posponed after the mergeback, and also kuba's fullstack patch | 12:59 |
| jschwarz | that doesn't leave a lot of patches | 12:59 |
| jschwarz | ihrachyshka, is there a patch I can help with? | 13:00 |
| openstackgerrit | Sergey Vilgelm proposed openstack/neutron-vpnaas: Use oslo.log library instead of system logging module https://review.openstack.org/211136 | 13:00 |
| *** rushil has joined #openstack-neutron | 13:00 | |
| ihrachyshka | jschwarz, I'm looking thru the queue... | 13:00 |
| *** Marga_ has quit IRC | 13:00 | |
| *** bapalm has quit IRC | 13:02 | |
| *** Kennan has quit IRC | 13:03 | |
| tfukushima | irenab: Regarding the validation in Kuryr, I was thinking about to add it in the latter patch. Do you think if I need to add it now or I can do it later. | 13:03 |
| *** park__ has quit IRC | 13:03 | |
| ihrachyshka | jschwarz, https://etherpad.openstack.org/p/qos-sync see above | 13:04 |
| irenab | tfukushima: to move faster, you can add #TODO with note where validation should be applied and then resolve it in follow-up patch | 13:04 |
| ihrachyshka | it seems most critical stuff is on moshele, hm. | 13:05 |
| ihrachyshka | can we somehow reach to him? gsagie, do you have access to moshele? :) | 13:05 |
| jschwarz | ihrachyshka, probably irenab can reach moshele | 13:05 |
| irenab | ihrachyshka: will ping him | 13:06 |
| ihrachyshka | irenab, thanks. :) | 13:06 |
| jschwarz | there ya' go | 13:06 |
| tfukushima | @irenab: Ok, thanks. | 13:06 |
| *** xek_ has joined #openstack-neutron | 13:06 | |
| *** claudiub has quit IRC | 13:06 | |
| jschwarz | ihrachyshka, I'll start reviewing | 13:06 |
| *** mkolesni__ has quit IRC | 13:06 | |
| gsagie | i have limited access to moshele, he has rate limit enabled | 13:07 |
| gsagie | and it works ;) | 13:07 |
| ihrachyshka | jschwarz, that's a good idea | 13:07 |
| jschwarz | lol | 13:07 |
| ihrachyshka | heh | 13:07 |
| jschwarz | how come we don't have an #openstack-qos channel or something? ;-) | 13:07 |
| *** erlarese has joined #openstack-neutron | 13:07 | |
| *** moshele has joined #openstack-neutron | 13:08 | |
| sc68cal | jschwarz: there is neutron-qos | 13:08 |
| jschwarz | :O | 13:08 |
| sc68cal | oops | 13:08 |
| sc68cal | openstack-neutron-qos | 13:08 |
| moshele | hi | 13:08 |
| irenab | ihrachyshka: done | 13:08 |
| amuller | jschwarz: that will give the rest of the community less exposure to what it is you're doing | 13:08 |
| amuller | I would discourage the usage of another channel | 13:08 |
| *** haigang has quit IRC | 13:09 | |
| jschwarz | amuller, ! | 13:09 |
| jschwarz | amuller, how was your flight? | 13:09 |
| ihrachyshka | moshele, hey! so I came up with a list of pieces in review and critical: https://etherpad.openstack.org/p/qos-sync | 13:09 |
| amuller | jschwarz: nearly infinite | 13:09 |
| sc68cal | amuller: it's pretty dead | 13:09 |
| ihrachyshka | moshele, as you can see, most critical pieces that require work are on you | 13:09 |
| jschwarz | amuller, so re: fullstack - i've reordered and submitted the lot of them | 13:09 |
| ihrachyshka | moshele, do you think you can share some with me and jschwarz and others so that we get it done today? | 13:09 |
| amuller | jschwarz: I saw, good work :) | 13:10 |
| moshele | ihrachyshka: yes I need help I guess | 13:10 |
| jschwarz | also i've split the EnvironmentDescription patch away, AND wrote a qos fullstack test here: https://review.openstack.org/202492 | 13:10 |
| jschwarz | amuller, I do have some ideas for a review and some changes that might be a good idea after working with the infra a bit | 13:10 |
| moshele | ihrachyshka: I would like to focus on the SR-IOV part | 13:10 |
| ihrachyshka | moshele, what do you think we should take over? I guess not sr-iov :) | 13:10 |
| jschwarz | so I'll comment them on the review :) | 13:10 |
| ihrachyshka | ok, then let us take two others :) | 13:11 |
| amuller | jschwarz: the split of the environmentdescription patch is because you need it for the qos branch I assume | 13:11 |
| *** xek_ has quit IRC | 13:11 | |
| ihrachyshka | jschwarz, https://review.openstack.org/211117 or https://review.openstack.org/211090 you choose | 13:11 |
| moshele | ihrachyshka: https://review.openstack.org/#/c/211117/ | 13:11 |
| *** apuimedo has quit IRC | 13:11 | |
| jschwarz | amuller, yes. Also, it doesn't make a lot of sense to put it in the tunneling-support patch | 13:11 |
| ihrachyshka | moshele, and 211090, right? or not? | 13:11 |
| jschwarz | amuller, it's a bit independent | 13:11 |
| jschwarz | ihrachyshka, looking | 13:12 |
| moshele | ihrachyshka: I did some of the comments on https://review.openstack.org/#/c/211090/ I will push what I have in few min | 13:12 |
| openstackgerrit | Sergey Vilgelm proposed openstack/neutron-vpnaas: Use oslo.log library instead of system logging module https://review.openstack.org/211136 | 13:12 |
| jschwarz | ihrachyshka, looks like 211117 is easier to get into so I'll take that I guess | 13:12 |
| *** lindycoder has joined #openstack-neutron | 13:13 | |
| ihrachyshka | jschwarz, ok cool :) | 13:13 |
| ihrachyshka | moshele, ok, ping me when you are done with 211090 and I'll take it from there | 13:13 |
| moshele | ihrachyshka: ok thanks | 13:13 |
| ihrachyshka | mestery, are you avail for some final qos reviews? | 13:13 |
| *** shwetaap has joined #openstack-neutron | 13:14 | |
| mestery | ihrachyshka: Aye aye, sir, I'm here now. | 13:16 |
| *** park__ has joined #openstack-neutron | 13:16 | |
| openstackgerrit | vikram.choudhary proposed openstack/neutron: [WIP] BGP Dynamic Routing: introduce dragent scheduler model https://review.openstack.org/207607 | 13:16 |
| *** scheuran has quit IRC | 13:16 | |
| ihrachyshka | mestery, that's cool :) so let's start from simple ones: https://review.openstack.org/211003 | 13:17 |
| *** rawl has joined #openstack-neutron | 13:17 | |
| *** scheuran has joined #openstack-neutron | 13:18 | |
| ihrachyshka | mestery, first functional test: https://review.openstack.org/210012 | 13:18 |
| *** prad_ has joined #openstack-neutron | 13:18 | |
| openstackgerrit | Madhu Mohan proposed openstack/neutron: Redundant tests removed from ovs-lib unit tests: https://review.openstack.org/208860 | 13:18 |
| ihrachyshka | mestery, race condition safety in qos agent extension: https://review.openstack.org/211044 | 13:18 |
| ihrachyshka | mestery, and finally, removal of redundant config option before merge-back: https://review.openstack.org/210998 | 13:18 |
| *** thumpba has joined #openstack-neutron | 13:19 | |
| mestery | ihrachyshka: Looking | 13:19 |
| *** regXboi has joined #openstack-neutron | 13:19 | |
| openstackgerrit | Sergey Vilgelm proposed openstack/neutron-vpnaas: Use oslo.log library instead of system logging module https://review.openstack.org/211136 | 13:19 |
| moshele | ihrachyshka: you can take https://review.openstack.org/#/c/211090/ enjoy :) | 13:20 |
| *** Kennan has joined #openstack-neutron | 13:20 | |
| ihrachyshka | mestery, we have three more patches that we want in before merge-back, but those are in-work now, but we expect them to be ready later today (specifically, 1) sr-iov qos driver; 2) and some agent extension API refining) | 13:20 |
| *** jecarey has quit IRC | 13:20 | |
| ihrachyshka | moshele, thanks | 13:20 |
| Sam-I-Am | mestery: i'm trying to find information on the status of arp spoofing/ebtables in liberty | 13:20 |
| *** Marga_ has joined #openstack-neutron | 13:20 | |
| Sam-I-Am | mestery: particularly with linuxbridge | 13:21 |
| *** merooney has joined #openstack-neutron | 13:21 | |
| *** dane_leblanc has joined #openstack-neutron | 13:22 | |
| * sc68cal perks up | 13:22 | |
| ihrachyshka | moshele, I need to understand why you needed to support flushing port with qos_policy=None. | 13:22 |
| *** julim has joined #openstack-neutron | 13:22 | |
| sc68cal | Sam-I-Am: I think those patches are landing in liberty - I think the ovs side landed already, have to check on LB | 13:23 |
| *** zhenguo has quit IRC | 13:23 | |
| sc68cal | Sam-I-Am: yep - 04197bc4bbf2bc611371060db839028c2686f87a | 13:23 |
| Sam-I-Am | sc68cal: ok, i'm hearing reports that its broken with linuxbridge | 13:24 |
| sc68cal | Sam-I-Am: https://github.com/openstack/neutron/commit/04197bc4bbf2bc611371060db839028c2686f87a | 13:24 |
| jschwarz | moshele, ihrachyshka, re https://review.openstack.org/#/c/211117/ - done :) | 13:24 |
| moshele | ihrachyshka: on update you have the old policy and the new one so you want to remove the old rules and apply the new ones | 13:24 |
| *** xek has quit IRC | 13:24 | |
| ihrachyshka | moshele, but why not flushing all rules and then just apply the new one? | 13:24 |
| sc68cal | Sam-I-Am: OK - well I see that the testing is setting it to false - https://github.com/openstack/neutron/commit/04197bc4bbf2bc611371060db839028c2686f87a#diff-12f1a91f563a8c1711e560a06520ef45R92 - so maybe we need to enable and see what breaks | 13:25 |
| ihrachyshka | at least now, when we have a single rule type | 13:25 |
| *** xek has joined #openstack-neutron | 13:25 | |
| *** erlarese_ has joined #openstack-neutron | 13:25 | |
| Sam-I-Am | sc68cal: yeah, something about floating ips not working | 13:25 |
| Sam-I-Am | sc68cal: waiting for the person who tested it with osad to wake up and get me some details | 13:25 |
| *** xek has quit IRC | 13:26 | |
| sc68cal | Sam-I-Am: ok - when he does - link me to the launchpad bug and we'll start digging | 13:26 |
| moshele | ihrachyshka: for now that what we are doing because I didn't implement the diff, it is more for future | 13:26 |
| Sam-I-Am | sc68cal: thx, will do | 13:26 |
| *** xek has joined #openstack-neutron | 13:26 | |
| *** erlarese has quit IRC | 13:26 | |
| *** erlarese_ is now known as erlarese | 13:26 | |
| moshele | ihrachyshka: see note on https://review.openstack.org/#/c/211090/2/neutron/agent/l2/extensions/qos.py Line 171 | 13:27 |
| mestery | Sam-I-Am sc68cal: Looks like you folks are covered now? | 13:27 |
| Sam-I-Am | mestery: yarp, thx | 13:27 |
| sc68cal | mestery: ++ | 13:28 |
| mestery | Sam-I-Am: awesome, thanks! | 13:28 |
| *** tiswanso has joined #openstack-neutron | 13:28 | |
| Sam-I-Am | progress | 13:28 |
| *** salv-orl_ has joined #openstack-neutron | 13:28 | |
| Sam-I-Am | osad... living on the edge of neutron features with linuxbridge | 13:28 |
| moshele | jschwarz: can you verify that it is working? I mean that we clean to rules when deleting the port | 13:28 |
| ihrachyshka | moshele, nah, I meant, why do we have special casing in line 70? why not just always flush all supported rules? | 13:28 |
| *** zzzeek has joined #openstack-neutron | 13:29 | |
| *** jasondotstar has joined #openstack-neutron | 13:29 | |
| mestery | ihrachyshka: See my concerns on https://review.openstack.org/#/c/211044/ | 13:30 |
| *** haigang has joined #openstack-neutron | 13:31 | |
| moshele | ihrachyshka: we can do it you way it was just optimization not to delete rule which is not change or just update | 13:31 |
| *** salv-orlando has quit IRC | 13:31 | |
| jschwarz | moshele, sure | 13:32 |
| *** banix has joined #openstack-neutron | 13:33 | |
| ihrachyshka | mestery, those locks are agent side | 13:34 |
| mestery | ihrachyshka: Thanks for the clarification, I wasn't sure :) | 13:34 |
| jlibosva | ihrachyshka: mestery I just replied at the very same second on review :)0 | 13:34 |
| *** jasondotstar has quit IRC | 13:34 | |
| *** changbl has quit IRC | 13:34 | |
| ihrachyshka | mestery, it's in neutron/agent/l2/extensions :) | 13:34 |
| mestery | ihrachyshka: I'm doing too many things this morning apparently | 13:34 |
| ihrachyshka | mestery, that's fine to ask | 13:35 |
| mestery | ihrachyshka: Literally juggling 5 conversations a mere 10 minutes after coming online :D | 13:35 |
| ihrachyshka | better safe than sorry :) | 13:35 |
| mestery | ihrachyshka: And failing at at least 3 of them :( | 13:35 |
| ihrachyshka | mestery, you're popular | 13:35 |
| mestery | lol | 13:35 |
| mestery | ihrachyshka: The -3 milestone rush I guess | 13:35 |
| *** changzhi has joined #openstack-neutron | 13:36 | |
| *** lyanchih has joined #openstack-neutron | 13:36 | |
| *** prad_ is now known as pradk | 13:36 | |
| *** banix has quit IRC | 13:36 | |
| *** banix has joined #openstack-neutron | 13:37 | |
| jschwarz | moshele, so I created a neutron port and attached it to an ovs port, then deleted the neutron port | 13:37 |
| jschwarz | moshele, ovsdb-client dump Open_vSwitch still shows the rate settings in the ovs after the deletion | 13:37 |
| mestery | ihrachyshka: Around for a question on https://review.openstack.org/#/c/210998/3? | 13:37 |
| jschwarz | is that what you meant for me to do? | 13:37 |
| ihrachyshka | mestery, of course! | 13:37 |
| mestery | ihrachyshka: So, with that one, by removing the driver config and hard-coding into ML2, how does a non-ML2 plugin implement a QoS driver? :) | 13:38 |
| *** dtynan has quit IRC | 13:38 | |
| moshele | jschwarz: it should be create a port that attached to policy and the delete it | 13:38 |
| jschwarz | mestery, why would a non-ML2 plugin implement a QoS driver...? ;-) | 13:38 |
| mestery | lol | 13:39 |
| jschwarz | moshele, that's what I did | 13:39 |
| jschwarz | forgot to write the policy part | 13:39 |
| moshele | jschwarz: so it not working? | 13:39 |
| jschwarz | I guess not | 13:39 |
| ihrachyshka | mestery, not sure how ml2 is relevant here. it's agent side qos driver, so what agent does is it initializes agent extension manager and passes driver_type to it, that is then propagated to each extension, including qos. and then qos extension will use driver_type to choose the needed backend driver (by pulling driver_type alias from corresponding setup.cfg namespace) | 13:39 |
| ihrachyshka | jschwarz, that's funny, but they will :) | 13:40 |
| *** mkolesni__ has joined #openstack-neutron | 13:40 | |
| mestery | ihrachyshka: So in essence, this patch is hard-coding the agent side driver to OVS? Seems reasonable to me. | 13:40 |
| ihrachyshka | mestery, for OVS? yes. | 13:40 |
| moshele | jschwarz: do you get to the delete_port part ? | 13:40 |
| mestery | ihrachyshka: Yes, for OVS. | 13:40 |
| ihrachyshka | at least until we have a case for pluggable drivers | 13:40 |
| mestery | Cool | 13:40 |
| * jschwarz is checking | 13:40 | |
| ihrachyshka | then we'll get a config option for OVS agent specifically | 13:40 |
| ihrachyshka | not for extension manager | 13:40 |
| ihrachyshka | because most agents will have a single driver | 13:41 |
| ihrachyshka | it's just that for OVS it may be both LB and OVS | 13:41 |
| ihrachyshka | due to hybrid nature | 13:41 |
| *** dtynan has joined #openstack-neutron | 13:41 | |
| ihrachyshka | in any case, we don't have LB delivered, so it's YAGNI | 13:42 |
| jschwarz | moshele, yes, it enters the function | 13:42 |
| *** ferest has joined #openstack-neutron | 13:43 | |
| jschwarz | and it crashes :P | 13:43 |
| amuller | ihrachyshka: ping https://review.openstack.org/#/c/210769/ | 13:43 |
| jschwarz | looking into it | 13:43 |
| *** banix_ has joined #openstack-neutron | 13:43 | |
| vichoward | i'm a bit lost on what is needed for the remove policy todo for neutron/agent/l2/extensions/qos_agent.py, anyone have a few minutes to discuss with me when convenient? | 13:43 |
| *** shoutm has quit IRC | 13:44 | |
| *** ferest has quit IRC | 13:46 | |
| *** banix has quit IRC | 13:46 | |
| *** banix_ is now known as banix | 13:46 | |
| *** haigang has quit IRC | 13:47 | |
| *** gsagie_ has joined #openstack-neutron | 13:48 | |
| moshele | ihrachyshka, jschwarz | 13:48 |
| moshele | : I have to leave now I will be back in few hours | 13:48 |
| jschwarz | ok | 13:49 |
| *** mriedem has joined #openstack-neutron | 13:49 | |
| *** jasondotstar has joined #openstack-neutron | 13:49 | |
| mriedem | anyone know what's going on with the vpnaas issues blocking kilo jobs that use neutron? https://bugs.launchpad.net/neutron/+bug/1483266 | 13:49 |
| openstack | Launchpad bug 1483266 in neutron "q-svc fails to start in kilo due to "ImportError: No module named neutron_vpnaas.services.vpn.service_drivers.ipsec"" [Undecided,New] | 13:49 |
| moshele | ihrachyshka: SR-IOV will be ready today I hope, but it depend on https://review.openstack.org/#/c/211090/ and https://review.openstack.org/#/c/211117/ anyway | 13:49 |
| *** devvesa has quit IRC | 13:49 | |
| *** moha_hunt has joined #openstack-neutron | 13:50 | |
| ihrachyshka | moshele, we'll make an effort to get those in | 13:50 |
| *** irenab has quit IRC | 13:50 | |
| *** kawa2014 has quit IRC | 13:50 | |
| ihrachyshka | moshele, ok, we'll probably wait for sr-iov then before moving with merge requests. | 13:50 |
| moshele | ihrachyshka: when you plan to do the merge to master? | 13:50 |
| *** yamamoto has joined #openstack-neutron | 13:50 | |
| ihrachyshka | moshele, I planned today, but I guess we may need to take it till tomorrow | 13:50 |
| moshele | ihrachyshka: ok thanks | 13:50 |
| *** kawa2014 has joined #openstack-neutron | 13:51 | |
| jschwarz | moshele, fixed the bug, was my fault :) | 13:51 |
| jschwarz | it now works | 13:51 |
| *** annegentle has quit IRC | 13:52 | |
| moshele | jschwarz: great thanks for the help :) | 13:52 |
| *** ijw has joined #openstack-neutron | 13:53 | |
| *** arif-ali has quit IRC | 13:53 | |
| jschwarz | moshele, sure :) | 13:53 |
| jschwarz | ihrachyshka, uploaded a new patchset to https://review.openstack.org/#/c/211117 - manually checked it and it now works | 13:54 |
| *** moshele has quit IRC | 13:54 | |
| *** alexpilotti has quit IRC | 13:55 | |
| *** yamamoto has quit IRC | 13:55 | |
| *** merooney has quit IRC | 13:57 | |
| *** salv-orl_ has quit IRC | 13:59 | |
| *** jasondotstar has quit IRC | 13:59 | |
| *** jasondotstar has joined #openstack-neutron | 14:00 | |
| *** jpena|lunch is now known as jpena | 14:00 | |
| ajo | ihrachyshka, mestery : https://review.openstack.org/#/c/211173/ | 14:01 |
| ajo | I'm adding more now in separate patches, the worst part was finding that I needed to cleanup the registry :) | 14:01 |
| ajo | since it's a singleton.. | 14:02 |
| *** annegentle has joined #openstack-neutron | 14:02 | |
| *** changzhi has quit IRC | 14:02 | |
| gsagie_ | irenab : you know where apuimedo is? | 14:02 |
| *** mkolesni__ has quit IRC | 14:02 | |
| *** boris-42 has joined #openstack-neutron | 14:03 | |
| jschwarz | ajo, boom, review :) | 14:04 |
| ihrachyshka | ajo, minor comments | 14:04 |
| *** Murali_ has quit IRC | 14:05 | |
| sc68cal | mriedem: I think pc_m is working on fixing | 14:06 |
| *** josecastroleon has quit IRC | 14:06 | |
| mriedem | and/or dougwig | 14:06 |
| *** park_ has joined #openstack-neutron | 14:07 | |
| *** johnbelamaric has joined #openstack-neutron | 14:07 | |
| *** changzhi has joined #openstack-neutron | 14:07 | |
| openstackgerrit | Federico Ressi proposed openstack/neutron: Support delegation of bind_port to networking-odl backend driver. https://review.openstack.org/205102 | 14:08 |
| *** kbringard has joined #openstack-neutron | 14:08 | |
| *** kbringard has joined #openstack-neutron | 14:08 | |
| *** merooney has joined #openstack-neutron | 14:09 | |
| jschwarz | ihrachyshka, btw I'll be off most day tomorrow except for a few hours in the afternoon | 14:09 |
| jschwarz | so if you need something from me now's the time ;-) | 14:09 |
| ihrachyshka | jschwarz, that's fine. but thanks for heads-up :) | 14:09 |
| *** Kennan2 has joined #openstack-neutron | 14:09 | |
| ihrachyshka | jschwarz, well, probably functional tests :) | 14:09 |
| ihrachyshka | jschwarz, sorry | 14:09 |
| ihrachyshka | jschwarz, fullstack | 14:09 |
| *** Kennan has quit IRC | 14:09 | |
| jschwarz | ihrachyshka, fullstack is blocked on numerous patches | 14:10 |
| amuller | jschwarz: which ones | 14:10 |
| ihrachyshka | jschwarz, sure, but more scenarios are desired anyway | 14:10 |
| amuller | jschwarz: something I can help with? | 14:10 |
| *** arif-ali has joined #openstack-neutron | 14:10 | |
| ajo | jschwarz + ihrachyshka thanks | 14:10 |
| jschwarz | amuller, 3 patches of the neutronclient, and the 2 fullstack patches I created yesterday | 14:10 |
| ihrachyshka | amuller, I think it's more other cores need to review the patches you have up | 14:10 |
| *** Murali_ has joined #openstack-neutron | 14:10 | |
| amuller | ihrachyshka: did you review the first patch in the series? the fake VM one | 14:10 |
| ihrachyshka | jschwarz, 3 for client? | 14:10 |
| jschwarz | amuller, https://review.openstack.org/191197 and | 14:10 |
| *** park__ has quit IRC | 14:11 | |
| jschwarz | https://review.openstack.org/210830 | 14:11 |
| ihrachyshka | amuller, no, and I feel guilty. but we always have other stuff | 14:11 |
| jschwarz | ihrachyshka, well actually it's 4 :) | 14:11 |
| amuller | ihrachyshka: there is no other stuff, only fullstack | 14:11 |
| amuller | neutron is an excuse to build testing infrastructures | 14:11 |
| amuller | not the other way around! | 14:11 |
| ihrachyshka | ah right, totally forgot | 14:11 |
| amuller | features... pfft | 14:11 |
| ihrachyshka | that's why everyone requires tests | 14:12 |
| jschwarz | ihrachyshka, let me output those neutronclient patches for you inorder: 209887, 210021, 189655, 198277 | 14:12 |
| jschwarz | ihrachyshka, those are the neutronclient ones :) | 14:12 |
| ihrachyshka | jschwarz, ah those fixing vpnaas? | 14:12 |
| jschwarz | yes | 14:12 |
| jschwarz | it's still in the (bloody) queue | 14:12 |
| ihrachyshka | ok, I don't care about those. I try not to | 14:12 |
| jschwarz | ihrachyshka, well you should | 14:12 |
| jschwarz | that's a total of 6 patches blocking fullstack atm ;-) | 14:13 |
| pc_m | jschwarz: The blocking commit is 209887, which needs a second infra +2/+A. | 14:13 |
| jschwarz | pc_m, it still blocks either way :( | 14:14 |
| pc_m | just pointing out it is not a neutronclient patch. | 14:14 |
| pc_m | it's a project-config patch. | 14:14 |
| jschwarz | pc_m, any way we can hasten the infra guys to +A it? | 14:15 |
| sc68cal | pc_m: jschwarz: reach out to anteaya ? | 14:15 |
| pc_m | yeah just saw her on and am asking... | 14:15 |
| sc68cal | although I forget if the ops meetup is this week, anteaya may be traveling | 14:15 |
| jschwarz | pc_m, thanks :) | 14:16 |
| *** yfried is now known as yfried|afk | 14:16 | |
| haleyb | amuller: thanks for the review on that one, it looks better now | 14:17 |
| *** ajmiller has joined #openstack-neutron | 14:17 | |
| amuller | haleyb: having the test target the right function is important :) | 14:17 |
| *** yfried|afk is now known as yfried | 14:17 | |
| *** breitz has joined #openstack-neutron | 14:18 | |
| openstackgerrit | Assaf Muller proposed openstack/neutron: TESTING.rst love https://review.openstack.org/210754 | 14:18 |
| amuller | HenryG: ihrachyshka: ^ | 14:19 |
| amuller | might have to wait until henry's patch merges first though | 14:19 |
| *** jckasper has quit IRC | 14:19 | |
| amuller | there will probably be conflicts | 14:19 |
| *** r-daneel has joined #openstack-neutron | 14:19 | |
| anteaya | sc68cal: ops is next week | 14:19 |
| *** tarun has joined #openstack-neutron | 14:19 | |
| anteaya | I am not, repeat not traveling this week | 14:19 |
| * anteaya does a happy dance | 14:19 | |
| anteaya | I get to pet the cat | 14:20 |
| anteaya | also I'm not a guy | 14:20 |
| tarun | hi, Have query on rpc mechanism in neutron .. | 14:20 |
| tarun | We are trying to create a dummy rpc server to capture rpc request coming from neutron... | 14:21 |
| *** superdan is now known as dansmith | 14:23 | |
| *** kbringard has quit IRC | 14:23 | |
| *** sdake_ has joined #openstack-neutron | 14:23 | |
| tarun | Do we need to register our dummy program in neutron agent list so as to get the event capture or will it be sufficient to register a topic in exchange? | 14:24 |
| openstackgerrit | Rossella Sblendido proposed openstack/neutron: OVS agent react to events instead of polling https://review.openstack.org/164880 | 14:25 |
| openstackgerrit | Rossella Sblendido proposed openstack/neutron: Don't sync all devices when some fail https://review.openstack.org/202098 | 14:25 |
| openstackgerrit | Rossella Sblendido proposed openstack/neutron: check_changed_vlans doesn't need registered_ports as param https://review.openstack.org/211191 | 14:25 |
| openstackgerrit | Rossella Sblendido proposed openstack/neutron: Remove cache of registered ports in the OVS agent https://review.openstack.org/211192 | 14:25 |
| *** Marga_ has quit IRC | 14:25 | |
| *** wverdugo500 has joined #openstack-neutron | 14:25 | |
| *** Marga_ has joined #openstack-neutron | 14:26 | |
| amuller | tarun: you just need to register to a topic like current agents do | 14:26 |
| *** banix has quit IRC | 14:26 | |
| *** sdake has quit IRC | 14:27 | |
| tarun | amuller : we are trying the same. We have registered the hardcoded topic as the one used in ovsvapp but still no luck.. | 14:28 |
| *** ekuris has quit IRC | 14:28 | |
| *** josecastroleon has joined #openstack-neutron | 14:29 | |
| *** Piet has quit IRC | 14:29 | |
| *** agireud has joined #openstack-neutron | 14:29 | |
| *** baoli has quit IRC | 14:29 | |
| openstackgerrit | Fawad Khaliq proposed openstack/neutron: PLUMgrid plugin decomposition part II https://review.openstack.org/211196 | 14:29 |
| *** kbringard has joined #openstack-neutron | 14:30 | |
| *** lajos-katona has left #openstack-neutron | 14:30 | |
| openstackgerrit | Ryu Ishimoto proposed openstack/neutron: midonet: add midonet type driver https://review.openstack.org/209367 | 14:30 |
| *** gmurphy has joined #openstack-neutron | 14:30 | |
| *** afazekas has quit IRC | 14:31 | |
| *** baoli_ has joined #openstack-neutron | 14:31 | |
| *** apuimedo has joined #openstack-neutron | 14:31 | |
| *** sakaYK has joined #openstack-neutron | 14:33 | |
| amuller | tarun: eh, I guess it's just a matter of playing with it until it works | 14:33 |
| amuller | tarun: it's probably something stupid | 14:33 |
| ajo | ihrachyshka, jschwarz : new version | 14:33 |
| ajo | https://review.openstack.org/#/c/211173/ | 14:34 |
| *** sdake has joined #openstack-neutron | 14:34 | |
| ajo | tarun, do you mean agent notifications? | 14:34 |
| *** claudiub has joined #openstack-neutron | 14:34 | |
| *** gmurphy has left #openstack-neutron | 14:34 | |
| ajo | CREATE/DELETE/UPDATE for ports/nets etc? | 14:35 |
| ajo | tarun, look on neutron/plugins/ml2/drivers/linuxbridge/agent/linuxbridge_neutron_agent.py around L861 | 14:36 |
| ajo | agent_rpc.create_consumers(....,..., | 14:36 |
| ajo | **consumers** | 14:36 |
| ajo | ) | 14:36 |
| tarun | yes | 14:36 |
| ajo | you need to add there all the combinations you want to listen too | 14:36 |
| ajo | too->to | 14:37 |
| tarun | aah i will try that.. | 14:37 |
| ajo | and, of course, provide endpoints to handle those | 14:37 |
| *** sdake_ has quit IRC | 14:37 | |
| tarun | for now i am trying with neutron create only. | 14:37 |
| ajo | tarun: then look around L667 for the "endpopint" callbacks | 14:38 |
| ajo | endpoint | 14:38 |
| *** ecrosso has joined #openstack-neutron | 14:38 | |
| tarun | ok | 14:39 |
| *** moshele has joined #openstack-neutron | 14:40 | |
| jschwarz | amuller, so looking at https://review.openstack.org/#/c/191197/12/neutron/tests/fullstack/resources/config.py | 14:40 |
| *** jecarey has joined #openstack-neutron | 14:41 | |
| jschwarz | amuller, I would like to make all config classes (and process classes) receive the EnvironmentDescription and the relevant HostDescription | 14:41 |
| jschwarz | it makes it really hard keeping track which constructor has what and it makes sense that each config/process will know where it's living in | 14:42 |
| *** piyanai has quit IRC | 14:42 | |
| amuller | jschwarz: ok, that can be a patch against HEAD, master branch | 14:42 |
| jschwarz | will do | 14:42 |
| amuller | jschwarz: yeah that should simplify all of our config/process inits | 14:42 |
| jschwarz | amuller, will work on a patch such as that | 14:43 |
| *** gampel has quit IRC | 14:45 | |
| *** crose has joined #openstack-neutron | 14:46 | |
| *** ushkalim has joined #openstack-neutron | 14:46 | |
| *** fifieldt_ has quit IRC | 14:47 | |
| *** banix has joined #openstack-neutron | 14:48 | |
| *** irenab has joined #openstack-neutron | 14:49 | |
| *** SridharG has left #openstack-neutron | 14:50 | |
| *** scheuran has quit IRC | 14:51 | |
| ihrachyshka | sc68cal, got +2 in fwaas? nice to see patches move in | 14:52 |
| *** itzikb_ has quit IRC | 14:53 | |
| *** changzhi has quit IRC | 14:53 | |
| sc68cal | ihrachyshka: yep, we'll be announcing some stuff on the ml fwaas related | 14:56 |
| apuimedo | banix: tfukushima: irenab: diga: gsagie: Kuryr meeting in 4 minutes | 14:56 |
| gsagie_ | k | 14:56 |
| banix | hi | 14:56 |
| *** dguitarbite has quit IRC | 14:56 | |
| *** jckasper has joined #openstack-neutron | 14:56 | |
| irenab | apuimedo: ack | 14:57 |
| banix | at openstack-meeting-4 i should add | 14:59 |
| *** dguitarbite has joined #openstack-neutron | 14:59 | |
| *** Murali__ has joined #openstack-neutron | 14:59 | |
| *** dave-mcc_ has joined #openstack-neutron | 15:00 | |
| ihrachyshka | jlibosva, plz take a look https://review.openstack.org/#/c/211090/ | 15:00 |
| *** piyanai has joined #openstack-neutron | 15:00 | |
| *** zhipeng has joined #openstack-neutron | 15:01 | |
| *** salv-orlando has joined #openstack-neutron | 15:01 | |
| gsagie_ | diga here? | 15:01 |
| jlibosva | ihrachyshka: ack | 15:01 |
| *** helenxu has joined #openstack-neutron | 15:02 | |
| *** yalie has joined #openstack-neutron | 15:02 | |
| *** scheuran has joined #openstack-neutron | 15:03 | |
| *** Murali_ has quit IRC | 15:03 | |
| *** scheuran has quit IRC | 15:03 | |
| *** dsanders has joined #openstack-neutron | 15:03 | |
| *** dave-mccowan has quit IRC | 15:04 | |
| *** scheuran has joined #openstack-neutron | 15:04 | |
| *** ecrosso has quit IRC | 15:05 | |
| *** carl_baldwin has joined #openstack-neutron | 15:05 | |
| *** daneyon has joined #openstack-neutron | 15:05 | |
| *** daneyon has quit IRC | 15:05 | |
| *** nlahouti has joined #openstack-neutron | 15:06 | |
| *** daneyon has joined #openstack-neutron | 15:06 | |
| *** tsg has joined #openstack-neutron | 15:06 | |
| *** fawadkhaliq has quit IRC | 15:06 | |
| scheuran | sc68cal, ping | 15:07 |
| *** dave-mccowan has joined #openstack-neutron | 15:07 | |
| regXboi | wonderful.... | 15:07 |
| sc68cal | scheuran: pong | 15:08 |
| *** saksham has joined #openstack-neutron | 15:08 | |
| *** Murali__ has quit IRC | 15:08 | |
| * regXboi now chases why particular unit tests pass under py27 but not py34 | 15:08 | |
| *** ecrosso has joined #openstack-neutron | 15:08 | |
| *** saksham has left #openstack-neutron | 15:08 | |
| anteaya | anyone able to share the currnt status of the neutron third party ci, voting permission revokation motion in -meeting? | 15:08 |
| anteaya | sc68cal: do you know anything about it? | 15:09 |
| *** carl_baldwin has quit IRC | 15:09 | |
| anteaya | mestery: ^^ | 15:09 |
| scheuran | sc68cal, I was spending some time on making the linuxbridge agent more modular, be I fear that this is going in a different direction than your approach bringing lb closer to ovs... | 15:09 |
| *** ddepaoli has quit IRC | 15:09 | |
| *** dave-mcc_ has quit IRC | 15:09 | |
| scheuran | sc68cal, s/be/but | 15:09 |
| mestery | anteaya: We're going to discuss at the meeting this week, I think it was discussed last week as well, the patch is out to review to iterate on: https://review.openstack.org/#/c/207198/ | 15:09 |
| anteaya | mestery: thank you | 15:10 |
| *** thedodd has joined #openstack-neutron | 15:10 | |
| scheuran | sc68cal, https://review.openstack.org/#/c/210250/ | 15:10 |
| sc68cal | anteaya: there is a review up that mestery put up | 15:10 |
| sc68cal | scheuran: thanks, I'll take a look | 15:11 |
| scheuran | sc68cal, another patchset on top of that is still outstanding moving all the rpc stuff into the manager class as well...but I have to fix the unittests firsts | 15:11 |
| *** carl_baldwin has joined #openstack-neutron | 15:11 | |
| anteaya | sc68cal: yup, thanks, kyle shared the link | 15:11 |
| anteaya | sc68cal: you multi-tasker, you :) | 15:11 |
| sc68cal | scheuran: ok, I'll take a look and see - I'm not dead set on anything really, so wherever things lead you, that's OK | 15:12 |
| ihrachyshka | jlibosva, one question in https://review.openstack.org/#/c/211117/ | 15:12 |
| jlibosva | ihrachyshka: I'm afraid I don't have that big overview how the extensions are planned to be used - I have limited knowledge to ovs extension | 15:13 |
| scheuran | sc68cal, another thing. I checked out your code with git review and commited my fixes on top. But then I couldn't make git review again as git review automatically wants to rebase with (master?) | 15:14 |
| ajo | ihrachyshka, working on the OVSAgentTestFramework to introduce port updates, which are not handled | 15:14 |
| *** jamie_h has quit IRC | 15:14 | |
| scheuran | sc68cal, so I used git review -R | 15:14 |
| scheuran | sc68cal, is there a better way? | 15:14 |
| sc68cal | scheuran: yeah there were some merge conflicts that I just resolved this morning, so rebase on top of the newest | 15:15 |
| sc68cal | scheuran: https://review.openstack.org/#/c/209385/ - patchset 4 should help | 15:15 |
| scheuran | sc68cal, ok, let me have a look | 15:15 |
| regXboi | carl_baldwin: good morning | 15:16 |
| carl_baldwin | regXboi: hi | 15:16 |
| regXboi | carl_baldwin: I've got a couple of backlog bug patchsets ready for review | 15:16 |
| *** armax has joined #openstack-neutron | 15:17 | |
| regXboi | I just added you to the one you weren't already on :) | 15:18 |
| carl_baldwin | regXboi: Thanks | 15:19 |
| regXboi | I can give you chapter and verse on the others when you have time | 15:19 |
| *** dannywilson has joined #openstack-neutron | 15:23 | |
| *** minwang2 has joined #openstack-neutron | 15:23 | |
| ihrachyshka | jschwarz, can I ask you to play with https://review.openstack.org/#/c/211090/4 against your fullstack to see whether it does not break smth? | 15:23 |
| jschwarz | ihrachyshka, you man ask | 15:23 |
| jschwarz | XD | 15:23 |
| ihrachyshka | jschwarz, ok, good. so I do | 15:23 |
| *** dannywilson has quit IRC | 15:24 | |
| *** dannywilson has joined #openstack-neutron | 15:24 | |
| *** rawl has quit IRC | 15:26 | |
| *** mlavalle has joined #openstack-neutron | 15:27 | |
| *** park_ has quit IRC | 15:27 | |
| *** odyssey4me has joined #openstack-neutron | 15:27 | |
| *** shakamunyi has joined #openstack-neutron | 15:27 | |
| Sam-I-Am | sc68cal: meet Apsu | 15:27 |
| Apsu | o/ | 15:27 |
| Sam-I-Am | Apsu: wakey wakey | 15:27 |
| Apsu | Yeah yeah, I'm here. Mostly | 15:27 |
| Sam-I-Am | sc68cal: apsu has the scoop on the secgroup problem | 15:27 |
| Sam-I-Am | moreso than i do | 15:27 |
| Apsu | re: the ARP filtering ebtables rules on linuxbridge | 15:28 |
| jschwarz | ihrachyshka, jenkins has nothing but good things to say about 211117 | 15:28 |
| *** josecastroleon has quit IRC | 15:28 | |
| Apsu | sc68cal: Ran into a problem testing them yesterday. Floating IP ports don't have a special case in the ebtables code, so they don't get an ARP reply exception rule added. | 15:28 |
| *** abehl has quit IRC | 15:28 | |
| *** park_ has joined #openstack-neutron | 15:28 | |
| Apsu | Was going to make a bug this morning about it | 15:28 |
| *** kiran-r has joined #openstack-neutron | 15:29 | |
| Apsu | The issue seems to be that floats don't have port security set on them, and I'm not quite sure how to set it ad-hoc. The binding:vif_details is read-only | 15:29 |
| *** thumpba has quit IRC | 15:30 | |
| Apsu | You can add floats to allowed-address-pairs on the router gateway port, which adds the ebtables rules, but that's ugly | 15:30 |
| Apsu | So presumably the simple fix is set port_filter: True for float ports | 15:30 |
| ihrachyshka | jschwarz, why so sure | 15:30 |
| ihrachyshka | jschwarz, there is always some silly test somewhere | 15:30 |
| jschwarz | ihrachyshka, because it does :) | 15:31 |
| jschwarz | ihrachyshka, your patch passes my fullstack test | 15:32 |
| *** tiswanso has quit IRC | 15:32 | |
| *** daneyon has quit IRC | 15:32 | |
| *** tiswanso has joined #openstack-neutron | 15:32 | |
| *** daneyon has joined #openstack-neutron | 15:32 | |
| *** yfried has quit IRC | 15:32 | |
| *** thumpba has joined #openstack-neutron | 15:32 | |
| sc68cal | Apsu: ack. I'm getting a bit up to speed but floating IPs are NAT'd at the L3 agent, so unless I'm mistaken port security and the ebtables shouldn't be involved, since the packet's dst ip is rewritten to the fixed IP of the vm at the L3 agent's namespace | 15:33 |
| Apsu | sc68cal: Yes, but that happens at the L3/iptables layer. The very first thing that happens is ARP, at L2/ebtables. And the qrouter netns ARP reply doesn't make it back through ebtables. | 15:34 |
| Apsu | Because there's no -j ACCEPT for ARP replies from the floating IP :) | 15:34 |
| Apsu | So L3 traffic never actually happens | 15:34 |
| sc68cal | Apsu: which device is sending the ARP for the floating IP | 15:35 |
| *** SourabhP has joined #openstack-neutron | 15:35 | |
| sc68cal | Apsu: or are you saying the arp spoofing/ebtables is inserting these rules on the l3 agent namespace? | 15:35 |
| Apsu | sc68cal: The qg-* in the qrouter netns the float lives on. | 15:35 |
| Apsu | No, it's inserting them on the root namespace. | 15:35 |
| *** tsg has quit IRC | 15:35 | |
| odyssey4me | Apsu if I'm not mistaken it may be useful to be clear that you're referencing the issue on the l3 agent, not the compute node | 15:35 |
| *** PabloPhoenix has joined #openstack-neutron | 15:35 | |
| sc68cal | Apsu: root namespace of the l3 node? | 15:35 |
| Apsu | The rules apply to the tap* devices that plug into the netns | 15:35 |
| Apsu | sc68cal: Yes | 15:35 |
| ihrachyshka | moshele, I think we may want to move sr-iov pieces and refactoring to after merge-back | 15:36 |
| sc68cal | Apsu: can you please create a launchpad bug in the neutron project - make a couple paste.openstack.org dumps of your network node, and the compute node where the VM is | 15:36 |
| *** thumpba has quit IRC | 15:37 | |
| Apsu | sc68cal: Sure, but the compute node isn't involved at all here. | 15:37 |
| ihrachyshka | moshele, especially refactoring since we don't have fullstack in the tree | 15:37 |
| sc68cal | Apsu: understood, however I would still like to get it so I am on the same page as you are | 15:37 |
| Apsu | This is way before traffic would ever make it to the compute/instance. | 15:37 |
| Apsu | Sure. | 15:37 |
| ihrachyshka | moshele, how about you going with sr-iov without refactoring, then we get fullstack in, then we modify the abstract class and friends? | 15:37 |
| *** thumpba has joined #openstack-neutron | 15:37 | |
| *** thumpba has quit IRC | 15:38 | |
| *** thumpba has joined #openstack-neutron | 15:38 | |
| ihrachyshka | moshele, jlibosva has concerns about the refactoring change and wants us to think better about data structures we use to maintain lists of policies/ports/... | 15:38 |
| ihrachyshka | and I need to agree that he is right we made a mess there. | 15:38 |
| ihrachyshka | ajo, jschwarz ^^ you may also be interested | 15:39 |
| *** mriedem has left #openstack-neutron | 15:39 | |
| *** changbl has joined #openstack-neutron | 15:39 | |
| *** rotbeard has joined #openstack-neutron | 15:40 | |
| *** thumpba has quit IRC | 15:40 | |
| jschwarz | ihrachyshka, do you mean on the versioned objects side? | 15:40 |
| jschwarz | we have a lot of lists ;-) | 15:40 |
| *** thumpba has joined #openstack-neutron | 15:41 | |
| ihrachyshka | jschwarz, nah, in the agent driver | 15:41 |
| ihrachyshka | ovs one | 15:41 |
| jlibosva | no, the policies-ports relations in qos extention | 15:41 |
| jschwarz | hmm | 15:41 |
| *** Sukhdev has joined #openstack-neutron | 15:41 | |
| jschwarz | jlanoux, ihrachyshka, this explains my opinion in this matter: http://i.imgur.com/xVyoSl.jpg | 15:42 |
| jschwarz | jlibosva, ^ | 15:42 |
| jschwarz | (sorry again) | 15:42 |
| jlibosva | we should at least encapsulate it into one type that is capable of handling it more pleasantly | 15:42 |
| *** vikram has quit IRC | 15:42 | |
| ihrachyshka | jschwarz, https://review.openstack.org/#/c/211090/4/neutron/agent/l2/extensions/qos.py line 121-124 and all that uses it | 15:42 |
| ihrachyshka | we now have three structures that handle the same 'cache' | 15:42 |
| jlibosva | jschwarz: :D I don't believe you | 15:43 |
| jschwarz | let me check that code out... | 15:43 |
| jlanoux | jschwarz: :D anyway, are we not all in the same situation :p ? | 15:43 |
| jschwarz | XD | 15:43 |
| jschwarz | ihrachyshka, can you point to the other 2 structures? | 15:44 |
| ihrachyshka | ajo, please review https://review.openstack.org/#/c/211117/ | 15:44 |
| * jschwarz is not familiar with the agent side to that level | 15:44 | |
| ihrachyshka | jschwarz, those are in lines 122-124 | 15:44 |
| jschwarz | ahh | 15:44 |
| jschwarz | so 121-124 are the problems | 15:44 |
| jlibosva | I left some comments in PS3 I guess | 15:44 |
| *** thumpba has quit IRC | 15:44 | |
| ihrachyshka | jschwarz, it's all to track relationships between ports we saw, their policies, etc. | 15:44 |
| jlibosva | There was an update meanwhile | 15:44 |
| *** thumpba has joined #openstack-neutron | 15:45 | |
| ajo | ihrachyshka: ack | 15:45 |
| ihrachyshka | amuller, could you please https://review.openstack.org/#/c/211173/ another functional for qos? | 15:45 |
| jschwarz | ihrachyshka, jlibosva, I agree | 15:45 |
| jschwarz | ihrachyshka, jlibosva, first of all we can ditch 'known_policies', because if it's not a key in self.qos_policy_ports then we probably don't know it | 15:46 |
| jschwarz | right? | 15:46 |
| jschwarz | then we have known_ports which is only used once - we can... hmm... flatten(self.qos_policy_ports.values()) to get all the ports we know about | 15:46 |
| ihrachyshka | jschwarz, I think we even did it before that way, so yes, I think that's one thing, but I realize I don't have brain capacity to reason about it now. | 15:47 |
| jschwarz | and that leaves us with qos_policy_ports | 15:47 |
| *** slaweq has joined #openstack-neutron | 15:47 | |
| ihrachyshka | jschwarz, we may want the structure behave better with cpu, so we may still want to maintain multiple structures. but if so, then isolated under some API | 15:47 |
| *** gangil has joined #openstack-neutron | 15:48 | |
| ihrachyshka | mestery, another functional test: https://review.openstack.org/#/c/211173/3 | 15:48 |
| openstackgerrit | Ryan Moats proposed openstack/neutron: l3_db, l3_dvr_db: separate _process_floating_{ips, interfaces}() https://review.openstack.org/124696 | 15:48 |
| jschwarz | ihrachyshka, ahh, I see that known_policies maps to a policy object (probably a dict of some sort) | 15:49 |
| ihrachyshka | ajo, I believe after your functionals go in + delete_port API, we should request a merge. What do you think? | 15:49 |
| *** dguitarbite has quit IRC | 15:49 | |
| ajo | ihrachyshka, I agree, there certainly will be some bugs here and there, but nothing hard to fix, | 15:49 |
| ajo | ihrachyshka, btw I'm stacking to manually test all I can | 15:49 |
| *** jistr has quit IRC | 15:50 | |
| ajo | ihrachyshka, since there are some cases I'm not able to easily add in the functional tests *yet* (WIP) | 15:50 |
| jlibosva | ajo: I can try to help if you want | 15:50 |
| ihrachyshka | ajo, good. I will wait for those to merge, then prepare merge patch for review and send an email. I hope those pieces will go in quick | 15:50 |
| jschwarz | ihrachyshka, jlibosva, I'm all for some class that holds a mapping between a policy_id to a qos_policy and its known ports. regarding known_ports.... | 15:50 |
| ajo | jlibosva, nice, if you can stack, may be we can divide the manual tests | 15:50 |
| ajo | my stacking failed :/ | 15:51 |
| jlibosva | ajo: I mean I can try to implement while you test manually :) | 15:51 |
| ajo | ahh jlibosva | 15:51 |
| ajo | ack | 15:51 |
| openstackgerrit | Merged openstack/neutron-fwaas: script.py.mako: update to latest standards https://review.openstack.org/206690 | 15:51 |
| openstackgerrit | Merged openstack/neutron-fwaas: Killed existing downgrade rules in migration scripts https://review.openstack.org/206984 | 15:51 |
| ajo | jlibosva, basically, I'm trying to implement a port update in the functional ovs agent test framework | 15:51 |
| jlibosva | ajo: or if you have any issue, but manually testing is also a way to go | 15:51 |
| jlibosva | ajo: the one on review is uptodate? | 15:51 |
| ajo | jlibosva, yes, | 15:51 |
| jlibosva | ajo++ | 15:52 |
| openstackgerrit | Merged openstack/neutron: Avoid dhcp_release for ipv6 addresses https://review.openstack.org/209456 | 15:52 |
| ajo | jlibosva, I'm trying to handle the case where you modify a port to change qos_policy_id or set it to None | 15:52 |
| * jlibosva fetching | 15:52 | |
| openstackgerrit | Merged openstack/neutron: Consistent layout and headings for devref https://review.openstack.org/210769 | 15:52 |
| ajo | jlibosva, you can add it as a follow up patch | 15:52 |
| ajo | jlibosva a separate one I mean | 15:52 |
| *** merooney has quit IRC | 15:52 | |
| jlibosva | ajo: ok, the approach would be to inject dehydrated dict to rpc layer? | 15:52 |
| ajo | jlibosva, that's done so far | 15:52 |
| ajo | I didn't dehidrate, that'd be more complete, let's leave that to fullstack | 15:53 |
| *** rushil has quit IRC | 15:53 | |
| ajo | jlibosva: https://review.openstack.org/#/c/211173/3/neutron/tests/functional/agent/l2/extensions/test_ovs_agent_qos_extension.py L149 | 15:53 |
| jlibosva | ajo: Makes sense to me to pass prepared dict | 15:53 |
| ajo | jlibosva, btw, | 15:53 |
| ajo | what we're missing now | 15:53 |
| jlibosva | ajo: that contains dehydrated object | 15:53 |
| ajo | is modifying the port itself, not the policy | 15:53 |
| jlibosva | ajo: basically simulation what comes on wire to agent | 15:54 |
| ihrachyshka | jlibosva, I think for port update, it's not thru pull/push but thru get_device_details | 15:54 |
| PabloPhoenix | I'm not sure if this is exactly the right place to ask, but is anyone here knowledgeable about what might cause neutron-db-manage to silently fail to apply any migrations? | 15:54 |
| ajo | jlibosva, yes, but let's cover the other cases first, then we can evolve this one into a bit up in the layer if you want | 15:54 |
| jlibosva | ajo: ihrachyshka ooh, I see | 15:54 |
| ajo | jlibosva, we're missing port['qos_policy_id'] change to other policy, or to None | 15:54 |
| ihrachyshka | jlibosva, we had a bug there btw. hopefully fixed now. | 15:54 |
| ihrachyshka | that 'to None' case failed before | 15:54 |
| jschwarz | ihrachyshka, so looking at the code - L159 has known_ports associated to some policy_id (so if we have qos_policy_ports like we said, we can flip the order of the if and that solves this). Also, L188 doesn't currently have the qos_policy_id but in that case we can have like 'orphaned' ports | 15:55 |
| ajo | jlibosva: you may need to modify neutron/tests/functional/agent/l2/base.py | 15:55 |
| ajo | to have some method to inject port updates to the agent | 15:55 |
| ajo | probably, you can send it directly to the port updates handler in the agent, and that will work | 15:55 |
| ihrachyshka | jschwarz, wanna play with refactor? | 15:55 |
| ajo | and also update self.ports | 15:55 |
| *** _cjones_ has joined #openstack-neutron | 15:55 | |
| *** sudipto has joined #openstack-neutron | 15:55 | |
| jschwarz | so in short we can have qos_policy_ports -> (qos_policy, known_ports) and also 'orphaned_ports' | 15:55 |
| Apsu | sc68cal: Sam-I-Am: https://bugs.launchpad.net/neutron/+bug/1483315 | 15:55 |
| openstack | Launchpad bug 1483315 in neutron "ebtables ARP rules don't account for floating IPs on LinuxBridge" [Undecided,New] | 15:55 |
| jschwarz | ihrachyshka, I could but rather not... I'll play with it a bit and pastebin something for you | 15:56 |
| jschwarz | then we can talk about it :) | 15:56 |
| *** TrevorV has joined #openstack-neutron | 15:56 | |
| ihrachyshka | ok :) or better jlibosva since he has better taste than me ;) | 15:56 |
| ajo | jlibosva, I believe you could do self.agent.port_update(context, port=....) | 15:56 |
| jschwarz | will ping you both | 15:56 |
| *** yamahata has joined #openstack-neutron | 15:57 | |
| *** tfukushima has quit IRC | 15:57 | |
| ajo | jlibosva, yes, that may work, the agent will come back with a get_devices_info for that port | 15:57 |
| ajo | so you need to patch the port in self.ports dict | 15:57 |
| jlibosva | ajo: oh, wait. | 15:58 |
| *** tfukushima has joined #openstack-neutron | 15:58 | |
| *** tfukushima has quit IRC | 15:58 | |
| *** tfukushima has joined #openstack-neutron | 15:58 | |
| jlibosva | ajo: do we test the extension in separation (extension manager api) or running agent with qos extension and interaction agent with "fake server"? | 15:58 |
| jlibosva | ajo: the latter would make more sense to me | 15:59 |
| jlibosva | ajo: question is based on the filename and place | 15:59 |
| *** crose has quit IRC | 16:00 | |
| openstackgerrit | Merged openstack/neutron: Imported Translations from Transifex https://review.openstack.org/206885 | 16:00 |
| ajo | jlibosva, we instantiate an agent, and test it as a whole | 16:01 |
| ajo | jlibosva, and we have tests for the lower level (ovs) too | 16:02 |
| *** helenxu has quit IRC | 16:02 | |
| jlibosva | ajo: ack, that sounds good. | 16:02 |
| openstackgerrit | Elena Ezhova proposed openstack/neutron: Handle SIGHUP: neutron-server (single process mode) https://review.openstack.org/170905 | 16:02 |
| *** yalie has quit IRC | 16:03 | |
| ajo | jlibosva, and what's missing is sending port updates to the agent | 16:03 |
| *** yalie has joined #openstack-neutron | 16:04 | |
| *** yalie has quit IRC | 16:04 | |
| *** jlanoux has quit IRC | 16:04 | |
| *** gcossu has quit IRC | 16:05 | |
| sc68cal | kevinbenton: ^ https://bugs.launchpad.net/neutron/+bug/1483315 | 16:05 |
| openstack | Launchpad bug 1483315 in neutron "ebtables ARP rules don't account for floating IPs on LinuxBridge" [Undecided,New] | 16:05 |
| *** SourabhP has quit IRC | 16:06 | |
| *** tiswanso has quit IRC | 16:06 | |
| *** alex_klimov has quit IRC | 16:06 | |
| *** amotoki has joined #openstack-neutron | 16:08 | |
| *** minwang2 has quit IRC | 16:09 | |
| ihrachyshka | ajo, mestery: there was some silly conflict on merge: https://review.openstack.org/#/c/210998/ | 16:09 |
| mestery | ihrachyshka: done | 16:10 |
| amuller | ihrachyshka: I wasn't able to review the first functional tests patch in time and now there's a second one =D | 16:10 |
| ajo | ihrachyshka: done | 16:10 |
| ihrachyshka | amuller, it's qos, we move quick here | 16:11 |
| ajo | amuller, you can do it afterwards then ;D we'll be happy to improve them ;) | 16:11 |
| amuller | ihrachyshka: no kidding I had 2 meetings this morning and the patch was already merged | 16:11 |
| HenryG | amuller: git was not smart enouigh. I am resolving the TESTING.rst conflict since it's my fault | 16:12 |
| ihrachyshka | mestery, I am actually writing an email to request merge-back. once pieces we nailed as critical land, I'll create a merge patch and send it. I hope it will be today but we'll see | 16:12 |
| amuller | HenryG: oh cool I was just about to do that | 16:12 |
| mestery | ihrachyshka: \o/ | 16:12 |
| mestery | ihrachyshka: Sweet! Lets do it! :) | 16:12 |
| mestery | ihrachyshka: I'd really like to do this as soon as we can. Next week I'm at Linuxcon in Seattle (with carl_baldwin among others), so will be tricker to coordinate things then. | 16:13 |
| ihrachyshka | mestery, you can start to hire the army of reviewers :) | 16:13 |
| mestery | lol | 16:13 |
| *** kiran-r has quit IRC | 16:14 | |
| ajo | mestery, ihrachyshka : what will be the strategy?, I'd suggest(if it makes sense) merge back where there's no fundamental flaw. And make a list of things that could be improved, to keep working on it over master. Of course, if something seems really broken, or could affect the general neutron core, then... fix in place. | 16:14 |
| openstackgerrit | Henry Gessau proposed openstack/neutron: TESTING.rst love https://review.openstack.org/210754 | 16:16 |
| daneyon | apuimedo follow-up regarding libnetwork store https://github.com/docker/libnetwork/blob/master/store.go | 16:16 |
| ajo | I guess structural improvements should be limited to <=L-3, only bugfixes for L-4, and then we can go-on for M-1 as for neutron in general | 16:16 |
| ihrachyshka | ajo, I think that's the plan, and I'll be explicit about it in the email. | 16:17 |
| ihrachyshka | we are happy to cover major stuff in feature/qos, but if it's manageable in master, we should do it postmerge | 16:17 |
| *** ushkalim has quit IRC | 16:17 | |
| apuimedo | daneyon: I was under the impression that this part of libnetwork is not open to the remote drivers | 16:17 |
| jschwarz | jlibosva, this is what I had in mind: http://pastebin.com/vvqTZ5kQ | 16:18 |
| apuimedo | but rather an internal part | 16:18 |
| apuimedo | daneyon: thanks for following up ;-) | 16:18 |
| *** gangil has quit IRC | 16:18 | |
| jschwarz | ihrachyshka, ^ | 16:18 |
| daneyon | apuimedo let me take a min to see how the overlay driver uses the kv store... i'll circle back with you. | 16:19 |
| apuimedo | daneyon: thanks! | 16:19 |
| *** tiswanso has joined #openstack-neutron | 16:19 | |
| apuimedo | afair it was read only | 16:19 |
| *** mitchjameson has joined #openstack-neutron | 16:20 | |
| jlibosva | jschwarz: yeah, me too. I would just adapt to dict() api | 16:20 |
| *** scheuran has quit IRC | 16:20 | |
| ihrachyshka | jschwarz, yeah, though you haven't patched all places of ugly dicts/sets | 16:20 |
| apuimedo | for those who missed the meeting http://eavesdrop.openstack.org/meetings/kuryr/2015/kuryr.2015-08-10-15.00.html | 16:21 |
| jschwarz | ihrachyshka, aye | 16:21 |
| *** azbiswas has joined #openstack-neutron | 16:21 | |
| apuimedo | the detailed log contains a pretty interesting discussion | 16:21 |
| *** gangil has joined #openstack-neutron | 16:21 | |
| jschwarz | ihrachyshka, left this to you guys since it's only a draft and I'm going home now ;-) | 16:21 |
| jschwarz | either way there's the thing with the orphaned which could be handled better IMO | 16:22 |
| amuller | ihrachyshka: I think one reasonable condition for merge back is devref | 16:22 |
| ihrachyshka | amuller, it's in | 16:22 |
| amuller | ihrachyshka: ovs agent devref too? | 16:22 |
| ihrachyshka | amuller, I'll link to it in the email | 16:22 |
| amuller | ihrachyshka: and something that describes the usage of versionedobjects, as well as the new rpc generalization layer? | 16:22 |
| ihrachyshka | amuller, hm, depends on what you mean. let me give a link to you | 16:22 |
| daneyon | aquimedo looks like you are correct. kv appears to be a requirement of libnetwork but not of libnetwork drivers. | 16:23 |
| ihrachyshka | for latter, yes and yes | 16:23 |
| ihrachyshka | amuller, http://git.openstack.org/cgit/openstack/neutron/tree/doc/source/devref/quality_of_service.rst?h=feature/qos | 16:23 |
| ihrachyshka | line 198+ | 16:23 |
| jschwarz | pffft | 16:24 |
| amuller | HenryG: thanks for rebasing :) | 16:24 |
| jschwarz | that ihrachyshka sure is serious with that devref :) | 16:24 |
| amuller | ihrachyshka: I don't think that's good enough | 16:24 |
| ihrachyshka | jschwarz, well, maybe it's not enough. let's see what amuller has to say | 16:25 |
| ihrachyshka | amuller, ok, tell me what you want to see there | 16:25 |
| *** shausy has quit IRC | 16:25 | |
| *** patrickeast has joined #openstack-neutron | 16:25 | |
| amuller | ihrachyshka: there's just no reference to how the OVS agent actually configures QoS, what OVS features it uses to do that | 16:25 |
| amuller | ihrachyshka: OVS queues and profiles | 16:25 |
| ihrachyshka | amuller, so you want ovs_lib layer covered | 16:25 |
| amuller | ihrachyshka: maybe a very thin explanation of the changes made to ovs_lib, a link to something that explains OVS profiles and queues, and how/when the OVS agent calls out to ovs_lib | 16:26 |
| ihrachyshka | amuller, ack on that one. any other stuff missing for your taste? | 16:27 |
| amuller | ihrachyshka: what happens once the OVS agent recevies a port update notification? | 16:27 |
| amuller | describe the flow | 16:27 |
| ihrachyshka | gsagie, we may need that description of low level ovs interactions ^^ | 16:27 |
| ihrachyshka | amuller, ok, that one, I can handle :) | 16:27 |
| openstackgerrit | Ryan Moats proposed openstack/neutron: L3 Agent support of manual dvr snat add/remove https://review.openstack.org/123815 | 16:28 |
| openstackgerrit | Ryan Moats proposed openstack/neutron: manual add/remove router for dvr_snat agent https://review.openstack.org/122024 | 16:28 |
| openstackgerrit | Ryan Moats proposed openstack/neutron: DVR: fix router rescheduling https://review.openstack.org/199514 | 16:28 |
| openstackgerrit | Ryan Moats proposed openstack/neutron: DVR: fix router scheduling https://review.openstack.org/199032 | 16:28 |
| *** kawa2014 has quit IRC | 16:28 | |
| *** tfukushima has quit IRC | 16:28 | |
| *** jkraj has quit IRC | 16:29 | |
| jschwarz | right guys, see you tomorrow. ping me tomorrow morning if you need me :) | 16:29 |
| amuller | ihrachyshka: one more thing I would mention in the devref is why oslo versionedobjects are called versioned. This could be a link to the library documentation that explains this, then just a short sentence about what is the plain in neutron/qos for RPC versioning using those objects | 16:31 |
| apuimedo | ;-) | 16:31 |
| amuller | what is the plan* | 16:31 |
| ihrachyshka | amuller, well, I am not sure we have complete RPC plan for all neutron. we integrated it with QoS and have type agnostic interface for those objects. But how it applies to other RPC, I don't know yet. | 16:32 |
| *** vivek-ebay has joined #openstack-neutron | 16:33 | |
| *** jschwarz has quit IRC | 16:33 | |
| amuller | ihrachyshka: I'll explain where I'm coming from: When I started reading QoS code and understood that oslo.versionedobjects was being used I started reading up about that, then I wondered how QoS specifically will use the versioning capability. Then after a conversation with you I found out that for the L release, QoS won't use versioning. | 16:33 |
| amuller | ihrachyshka: that was my thought process | 16:33 |
| amuller | ihrachyshka: your time could have been spared if this stuff was written in the devref | 16:34 |
| *** fawadkhaliq has joined #openstack-neutron | 16:34 | |
| *** skraynev has quit IRC | 16:34 | |
| ihrachyshka | amuller, what do you mean it won't use it? it's just that there are no versions yet since it's a new feature. but rpc_callbacks.rst defines our versioning strategy for the interface. | 16:34 |
| *** skraynev has joined #openstack-neutron | 16:34 | |
| amuller | ihrachyshka: yes I mean that for the initial release there will be no different versions | 16:35 |
| amuller | ihrachyshka: so the up/down functions of versionedobjects won't be used initially | 16:35 |
| ihrachyshka | OK, I'll add some notion of it just in case, but details belong to rpc_callbacks.rst, at least while it's QoS only | 16:35 |
| ihrachyshka | ok, I'll make a note on that. | 16:35 |
| *** tfukushima has joined #openstack-neutron | 16:37 | |
| *** tfukushima has quit IRC | 16:37 | |
| *** yfried has joined #openstack-neutron | 16:38 | |
| *** lyanchih has quit IRC | 16:38 | |
| amuller | ihrachyshka: the commit message will have to describe the testing strategy with respect to fullstack at least and why we have to merge before adding the fullstack test | 16:38 |
| amuller | cause of the client issue | 16:39 |
| amuller | ihrachyshka: and the important thing is getting the mindset of reviewers right... | 16:39 |
| amuller | -1's about comment spelling and variable names won't get us anywhere | 16:39 |
| amuller | we have to focus on big picture stuff and do the rest on master | 16:39 |
| ByPasS | Hi everyone, quick question is it possible to migrate from L3 HA router to a NON-HA setup ? I don't have millions of current networks but I was wondering if it was possible... | 16:40 |
| *** yog_ has quit IRC | 16:40 | |
| *** neophy has joined #openstack-neutron | 16:40 | |
| amuller | ByPasS: HA routers don't currently support migration (using router-update to update the HA flag) | 16:40 |
| amuller | ByPasS: you'll have to do something similar manually | 16:41 |
| amuller | ByPasS: or script it | 16:41 |
| *** rushil has joined #openstack-neutron | 16:41 | |
| *** dane_leblanc has quit IRC | 16:41 | |
| amuller | ByPasS: also I gotta ask if there's some issue with HA routers you're trying to get away from :) | 16:42 |
| *** bapalm has joined #openstack-neutron | 16:43 | |
| ByPasS | amuller : thanks for the info, I guess the first step is to test the manual proc in a staging environment and see how it goes :D | 16:43 |
| ByPasS | amuller : vpnaas support actually is the limitation | 16:43 |
| amuller | ByPasS: aye, it would have to list all of the floating IPs on the network the router is serving, detach them, get a list of the router ports, disconnect, clear gateway, create a non-HA router, then add it all back | 16:43 |
| *** odyssey4me has left #openstack-neutron | 16:43 | |
| ihrachyshka | mestery, ok, keeping in mind amuller's requests for devref, I think we'll take another day to handle it, so request tomorrow | 16:43 |
| *** jpena has quit IRC | 16:44 | |
| amuller | ByPasS: ahh ok | 16:44 |
| amuller | ByPasS: there's a patch for that up for review actually! | 16:44 |
| amuller | ByPasS: maybe you can cherry pick and test it | 16:44 |
| ByPasS | amuller : sure #/url ? | 16:44 |
| amuller | ByPasS: https://review.openstack.org/#/c/200636/ | 16:44 |
| amuller | ByPasS: any feedback on the patch itself would be awesome | 16:44 |
| amuller | ByPasS: if it works or not, any issues found etc | 16:45 |
| *** slaweq has quit IRC | 16:46 | |
| mestery | ihrachyshka: Ack sounds good | 16:46 |
| *** rotbeard has quit IRC | 16:46 | |
| ByPasS | amuller : noted we will test it soon you think I can have an issue trying to apply on kilo as it's from master ? | 16:48 |
| *** slaweq has joined #openstack-neutron | 16:48 | |
| amuller | ByPasS: I don't know how much that code changed since Kilo | 16:48 |
| amuller | pc_m: ^ | 16:48 |
| amuller | ByPasS: Maybe Paul will know | 16:48 |
| jlibosva | ajo: I'm almost done | 16:49 |
| jlibosva | ajo: I just need to somehow trigger the rpc_loop to be completed | 16:49 |
| jlibosva | ajo: I tried eventlet.sleep() but seems like it's not deterministic | 16:49 |
| *** merooney has joined #openstack-neutron | 16:50 | |
| *** dane_leblanc has joined #openstack-neutron | 16:50 | |
| *** lazy_prince has joined #openstack-neutron | 16:51 | |
| pc_m | amuller: I don't think there were drastic changes since kilo. Not sure how much would have to change, without looking at it much closer. | 16:52 |
| openstackgerrit | Ryan Moats proposed openstack/neutron: L3 Agent support of manual dvr snat add/remove https://review.openstack.org/123815 | 16:52 |
| *** ivar-lazzaro has joined #openstack-neutron | 16:52 | |
| *** minwang2 has joined #openstack-neutron | 16:53 | |
| ihrachyshka | amuller, ok, may I ask you to review https://review.openstack.org/211259 and say what else is missing? I want to tackle that tomorrow in the morning | 16:53 |
| ByPasS | amuller : Paul Lio ? | 16:53 |
| ByPasS | oups Liu | 16:53 |
| ihrachyshka | jlibosva, ajo btw we have time till tomorrow, no merge request today | 16:53 |
| amuller | ByPasS: pc_m, Paul Michali | 16:53 |
| amuller | ByPasS: he answered just a moment ago :) | 16:53 |
| ByPasS | amuller : oh I just sorry sorry about that :) thx pc_m | 16:54 |
| ajo | ihrachyshka, I found a regression in assigning a policy_id to a port that previously didn't have it | 16:54 |
| ihrachyshka | amuller, I really hope you'll find time | 16:54 |
| ajo | that was working before | 16:54 |
| ajo | working on a fix | 16:54 |
| ihrachyshka | ajo, ack | 16:54 |
| amuller | ihrachyshka: yes I can't exactly whine about this stuff then not review your change | 16:54 |
| jlibosva | ajo: in functional or manual tests? | 16:54 |
| ihrachyshka | amuller, isn't it how all reviews go in openstack? | 16:54 |
| ajo | jlibosva, manual tests , the issue is in ml2.plugin when comparing the qos_policy_id for changes | 16:54 |
| jlibosva | ajo: aha | 16:55 |
| ajo | API->ml2.plugin interaction | 16:55 |
| ihrachyshka | ajo, wait, haven't I fixed that? | 16:55 |
| ihrachyshka | ajo, for None | 16:55 |
| ajo | hmm | 16:55 |
| ajo | I'm changing from None to something else | 16:55 |
| ihrachyshka | hm, ok. maybe that one I broke or whatever | 16:55 |
| ihrachyshka | please fix :) | 16:55 |
| ihrachyshka | I need to run | 16:55 |
| ajo | ihrachyshka, ack, doing | 16:55 |
| *** vikram has joined #openstack-neutron | 16:56 | |
| *** alexpilotti_ has joined #openstack-neutron | 16:57 | |
| jlibosva | ajo: https://review.openstack.org/#/c/211262/1/neutron/tests/functional/agent/l2/extensions/test_ovs_agent_qos_extension.py L168 :( | 16:57 |
| *** ildikov has quit IRC | 16:57 | |
| *** sakaYK has quit IRC | 16:58 | |
| *** baoli_ has quit IRC | 16:59 | |
| *** amotoki has quit IRC | 16:59 | |
| *** ihrachyshka has quit IRC | 17:00 | |
| *** baoli has joined #openstack-neutron | 17:00 | |
| *** alexpilotti_ has quit IRC | 17:01 | |
| *** mitchjameson has quit IRC | 17:01 | |
| *** fawadkhaliq has quit IRC | 17:01 | |
| *** arosen has joined #openstack-neutron | 17:01 | |
| ajo | jlibosva, I think your not basing it on the right patch | 17:01 |
| ajo | I have a "wait until rule applied" | 17:02 |
| jlibosva | ajo++ you are awesome man | 17:02 |
| Sukhdev | mestery: Ping | 17:02 |
| ajo | https://review.openstack.org/#/c/211173/ | 17:02 |
| ajo | jlibosva: I'm not :D | 17:02 |
| mestery | Sukhdev: Pongalongdingdong | 17:02 |
| jlibosva | ajo: thought I checked before I started writing that that I'm on the latest | 17:02 |
| Sukhdev | mestery: :-) | 17:02 |
| ajo | jlibosva: btw, there's no regression, I just forgot to configure the ml2 extension | 17:02 |
| ajo | ;D | 17:02 |
| ajo | yikes... | 17:02 |
| jlibosva | good that you found that :) | 17:03 |
| Sukhdev | mestery: Have a QQ about Ironic-neutron integration | 17:03 |
| openstackgerrit | Merged openstack/neutron: Add oslo db retry decorator to the RPC handlers https://review.openstack.org/207532 | 17:04 |
| *** britthou_ has joined #openstack-neutron | 17:04 | |
| mestery | Sukhdev: Please, go ahead | 17:04 |
| openstackgerrit | Merged openstack/neutron: Introduce ItemAllocator class https://review.openstack.org/210128 | 17:04 |
| *** sdake_ has joined #openstack-neutron | 17:04 | |
| openstackgerrit | Merged openstack/neutron: Broadcast service port's arp in DVR https://review.openstack.org/209388 | 17:04 |
| obondarev_ | Ryan Moats: ping | 17:05 |
| obondarev_ | anybody aware of Ryan Moats nic? | 17:05 |
| jckasper | obondarev_: regXboi = Ryan Moats | 17:06 |
| obondarev_ | jckasper: thanks! | 17:06 |
| regXboi | obondarev_: pong | 17:06 |
| obondarev_ | regXboi: hey | 17:06 |
| regXboi | obondarev: the bug changes were unintentional | 17:06 |
| regXboi | I'm picking up a patch set and rebasing it on your patch set | 17:06 |
| regXboi | and that caused a bunch of bug changes | 17:06 |
| *** jckasper has left #openstack-neutron | 17:07 | |
| obondarev_ | regXboi: please check my comment on https://review.openstack.org/#/c/123815/ | 17:07 |
| *** jckasper has joined #openstack-neutron | 17:07 | |
| * regXboi looks | 17:07 | |
| *** britthouser has quit IRC | 17:07 | |
| obondarev_ | regXboi: you mean a bunch of rebases on existing patches? | 17:07 |
| *** sdake has quit IRC | 17:07 | |
| *** mfranc213 has joined #openstack-neutron | 17:08 | |
| obondarev_ | regXboi: also last patchset here https://review.openstack.org/#/c/122024 reverts some important changes from ps 44 | 17:08 |
| *** SourabhP has joined #openstack-neutron | 17:09 | |
| regXboi | obondarev: I'm looking at the test cases to see if that patch is needed for https://bugs.launchpad.net/neutron/+bug/1369721 (the defect that I care about) | 17:09 |
| openstack | Launchpad bug 1369721 in neutron "manually moving dvr-snat router fails" [High,In progress] - Assigned to Ryan Moats (rmoats) | 17:09 |
| regXboi | obondarev_: ack | 17:09 |
| *** britthou_ has quit IRC | 17:10 | |
| *** lynnli has joined #openstack-neutron | 17:10 | |
| regXboi | obondarev_: I've also self -1'd 123815 as it is so WiP at this point | 17:10 |
| *** jamielennox is now known as jamielennox|away | 17:11 | |
| obondarev_ | regXboi: I will revert https://review.openstack.org/#/c/122024 back to th state it was at ps44 | 17:12 |
| *** yfried has quit IRC | 17:12 | |
| *** sdake_ is now known as sdake | 17:14 | |
| regXboi | obondarev_: ack | 17:14 |
| *** achanda has joined #openstack-neutron | 17:14 | |
| obondarev_ | regXboi: or better, can you please check ps44 there (with paren patches, basically can just checkout), if it solves manual rescheduling | 17:15 |
| jlibosva | ajo: yeah, I based it on wrong patch :D | 17:15 |
| obondarev_ | regXboi: then can say if https://review.openstack.org/#/c/123815/ is needed at all | 17:15 |
| *** bapalm has quit IRC | 17:16 | |
| *** salv-orlando has quit IRC | 17:16 | |
| obondarev_ | regXboi: sorry, have to leave now, will get back to it tomorrow | 17:16 |
| *** bapalm has joined #openstack-neutron | 17:16 | |
| *** abhiraut has joined #openstack-neutron | 17:16 | |
| regXboi | obondarev_ no worries | 17:17 |
| *** matrohon has quit IRC | 17:17 | |
| regXboi | carl_baldwin: ping - what's up with https://bugs.launchpad.net/neutron/+bug/1383571 | 17:17 |
| openstack | Launchpad bug 1383571 in neutron "The fip namespace can be destroyed on L3 agent restart" [Medium,Confirmed] - Assigned to Carl Baldwin (carl-baldwin) | 17:17 |
| *** moshele has quit IRC | 17:18 | |
| lynnli | amuller: ping | 17:19 |
| carl_baldwin | regXboi: I’m not sure why I was the assignee. | 17:20 |
| *** bapalm has quit IRC | 17:21 | |
| *** yushiro has joined #openstack-neutron | 17:22 | |
| *** lilynn2 has quit IRC | 17:23 | |
| *** banix has quit IRC | 17:23 | |
| *** mohankumar has joined #openstack-neutron | 17:23 | |
| yushiro | ping mestery | 17:24 |
| *** dave-mccowan has quit IRC | 17:26 | |
| *** saksham has joined #openstack-neutron | 17:27 | |
| *** jlibosva has quit IRC | 17:27 | |
| *** chadix has quit IRC | 17:28 | |
| mestery | yushiro: pong | 17:28 |
| *** abhiraut has quit IRC | 17:29 | |
| *** rushil has quit IRC | 17:30 | |
| yushiro | mestery, thank you :) I have a question https://review.openstack.org/#/c/205902/ . Are more reviewer necessary? Yesterday, depended-on patch has been merged. | 17:30 |
| *** ritesh has quit IRC | 17:30 | |
| openstackgerrit | Merged openstack/neutron: TESTING.rst love https://review.openstack.org/210754 | 17:30 |
| *** sdake_ has joined #openstack-neutron | 17:32 | |
| *** rushil has joined #openstack-neutron | 17:32 | |
| openstackgerrit | Sukhdev Kapur proposed openstack/neutron: Neutron-Ironic integration patch https://review.openstack.org/197774 | 17:33 |
| *** sridhar_ram has joined #openstack-neutron | 17:35 | |
| mestery | yushiro: looking | 17:35 |
| *** sdake has quit IRC | 17:35 | |
| mestery | yushiro: I just did a re-check, that should land it | 17:36 |
| mestery | stay tuned | 17:36 |
| *** abhiraut has joined #openstack-neutron | 17:38 | |
| yushiro | mestery, I see. Thank you for your help. | 17:38 |
| mestery | yushiro: yw | 17:38 |
| *** ritesh has joined #openstack-neutron | 17:38 | |
| *** nyechiel has quit IRC | 17:38 | |
| *** baoli has quit IRC | 17:39 | |
| *** sthillma has joined #openstack-neutron | 17:40 | |
| *** yushiro has quit IRC | 17:40 | |
| *** Sukhdev has quit IRC | 17:41 | |
| *** Sukhdev has joined #openstack-neutron | 17:42 | |
| *** baoli has joined #openstack-neutron | 17:43 | |
| Sukhdev | kevinbenton: Ping | 17:44 |
| *** shivrao has joined #openstack-neutron | 17:44 | |
| *** abehl has joined #openstack-neutron | 17:44 | |
| *** [Outcast] has joined #openstack-neutron | 17:44 | |
| *** britthouser has joined #openstack-neutron | 17:45 | |
| *** Marga_ has quit IRC | 17:46 | |
| *** Marga_ has joined #openstack-neutron | 17:47 | |
| *** rkukura has quit IRC | 17:47 | |
| amuller | lynnli: hello | 17:49 |
| *** madhu_ak has joined #openstack-neutron | 17:50 | |
| *** britthouser has quit IRC | 17:50 | |
| *** britthouser has joined #openstack-neutron | 17:50 | |
| lynnli | amuller: question about 178880 | 17:51 |
| lynnli | I added some more functional test cases along with yours | 17:51 |
| amuller | lynnli: ok | 17:52 |
| lynnli | but I need to refactored some codes as your original codes could be re-used.. question is ... should I just go ahead upload the new changes for review, and add my name as co-author? is that ok with you ? | 17:53 |
| amuller | lynnli: sounds great | 17:54 |
| *** piyanai has quit IRC | 17:55 | |
| *** piyanai has joined #openstack-neutron | 17:56 | |
| *** dave-mccowan has joined #openstack-neutron | 17:56 | |
| *** piyanai has quit IRC | 17:56 | |
| lynnli | amuller: OK .. let me put those together and upload it. Should be done soon. Thanks. Meanwhile, since there are more test cases need to be covered. I would think we should get this patch merged to upstream when it is in the good shape. And we can maintain it after that. Because some other patches may need to add more functional test cases and need this patch. Does that make sense ? | 17:56 |
| amuller | lynnli: yes the intention is to keep the original patch small | 17:56 |
| amuller | lynnli: just a few tests, but get the infrastructure in place | 17:57 |
| amuller | lynnli: then add more and more in follow up patches | 17:57 |
| lynnli | amuller: agree | 17:57 |
| *** SourabhP has quit IRC | 17:57 | |
| *** mohankumar has quit IRC | 17:58 | |
| *** dave-mcc_ has joined #openstack-neutron | 17:58 | |
| *** piyanai has joined #openstack-neutron | 17:58 | |
| *** dave-mccowan has quit IRC | 18:01 | |
| *** madhu_ak has quit IRC | 18:01 | |
| *** chadix has joined #openstack-neutron | 18:02 | |
| *** geekinutah has joined #openstack-neutron | 18:03 | |
| *** vivek-ebay has quit IRC | 18:04 | |
| *** rushil has quit IRC | 18:05 | |
| *** merooney has quit IRC | 18:05 | |
| *** rushil has joined #openstack-neutron | 18:07 | |
| *** rkukura has joined #openstack-neutron | 18:07 | |
| *** bapalm has joined #openstack-neutron | 18:07 | |
| *** bapalm has quit IRC | 18:07 | |
| *** bapalm has joined #openstack-neutron | 18:08 | |
| *** neophy has quit IRC | 18:08 | |
| *** merooney has joined #openstack-neutron | 18:08 | |
| *** banix has joined #openstack-neutron | 18:09 | |
| *** britthou_ has joined #openstack-neutron | 18:09 | |
| *** piyanai has quit IRC | 18:09 | |
| *** salv-orlando has joined #openstack-neutron | 18:10 | |
| sc68cal | mestery: quick question for you - I think https://review.openstack.org/#/c/206691/ is pretty much good to go, but two of the files don't have a copyright set - can I just stick 'OpenStack Foundation' in - or should I wait for ihar to respond. | 18:10 |
| sc68cal | mestery: https://review.openstack.org/#/c/206691/2/neutron_fwaas/db/migration/alembic_migrations/versions/liberty/expand/c40fbb377ad_initial.py,cm | 18:11 |
| amuller | sc68cal: it is incorrect to add a foundation copyright when it is not a foundation employee that wrote the code | 18:11 |
| mestery | sc68cal: I think we shoudl wait and let ihar respin those and add a correct copyright | 18:11 |
| sc68cal | amuller: ok. then can I assume redhat? | 18:11 |
| amuller | sc68cal: it should be a copyright of the author's company, or none at all | 18:11 |
| mestery | yes | 18:11 |
| amuller | sc68cal: I think Ihar prefers not to have a copyright last time I checked | 18:12 |
| *** ildikov has joined #openstack-neutron | 18:12 | |
| *** britthou_ has quit IRC | 18:12 | |
| sc68cal | amuller: ok - I'll -1 it and let him update instead of trying to guess what he'd do | 18:12 |
| *** yfried has joined #openstack-neutron | 18:12 | |
| *** britthouser has quit IRC | 18:12 | |
| *** jwy has joined #openstack-neutron | 18:12 | |
| *** abehl has quit IRC | 18:13 | |
| *** jlibosva has joined #openstack-neutron | 18:14 | |
| *** britthouser has joined #openstack-neutron | 18:16 | |
| *** [Outcast] has quit IRC | 18:17 | |
| *** sudipto has quit IRC | 18:17 | |
| regXboi | carl_baldwin: I'm not sure anybody is working on it | 18:17 |
| *** Aish has joined #openstack-neutron | 18:18 | |
| *** gampel has joined #openstack-neutron | 18:20 | |
| Sukhdev | kevinbenton: Hey Kevin, are you around? | 18:20 |
| *** yamahata has quit IRC | 18:20 | |
| *** britthouser has quit IRC | 18:21 | |
| *** salv-orlando has quit IRC | 18:22 | |
| *** bharath has joined #openstack-neutron | 18:23 | |
| *** abregman has quit IRC | 18:23 | |
| *** vivek-ebay has joined #openstack-neutron | 18:23 | |
| *** josecastroleon has joined #openstack-neutron | 18:24 | |
| *** baoli has quit IRC | 18:25 | |
| *** ihrachyshka has joined #openstack-neutron | 18:26 | |
| ihrachyshka | ajo, mestery: I guess we can +W: https://review.openstack.org/#/c/210998/ | 18:26 |
| *** baoli_ has joined #openstack-neutron | 18:26 | |
| mestery | ihrachyshka: Indeed we can | 18:26 |
| *** bapalm__ has joined #openstack-neutron | 18:27 | |
| sc68cal | ihrachyshka: hey, was asking about https://review.openstack.org/#/c/206691/ - I think it just needs a small respin then I think it's good to merge | 18:29 |
| *** merooney has quit IRC | 18:29 | |
| sc68cal | aveiga: you need to be in #openstack-infra | 18:30 |
| ihrachyshka | sc68cal, pfff, I am so dumb to get +2 reset for that nonsense :) | 18:30 |
| aveiga | sc68cal: why's that? | 18:30 |
| ihrachyshka | I wonder how people missed it though :) | 18:30 |
| sc68cal | ihrachyshka: well, in a vulgar display of power I'll +2+A it since dougwig +2'd it and I'm just being nit picky | 18:30 |
| *** bapalm has quit IRC | 18:30 | |
| ihrachyshka | sc68cal, nah, let's do it right. | 18:30 |
| sc68cal | aveiga: ipv6 packets getting dropped between SEA and DEN on the way to RAX via Comcast net | 18:30 |
| *** merooney has joined #openstack-neutron | 18:30 | |
| ihrachyshka | sc68cal, I think I can respin for that and you +2/+W, I don't think we need to wait again for second reviewer for such a nit | 18:31 |
| *** salv-orlando has joined #openstack-neutron | 18:31 | |
| sc68cal | ihrachyshka: works for me - apologies for the nit | 18:31 |
| sc68cal | worst case we blame dougwig :) | 18:32 |
| *** SourabhP has joined #openstack-neutron | 18:32 | |
| sc68cal | for making me core in something | 18:32 |
| ihrachyshka | :P | 18:32 |
| openstackgerrit | Ihar Hrachyshka proposed openstack/neutron-fwaas: switch to multi branch migration tree for liberty https://review.openstack.org/206691 | 18:32 |
| ihrachyshka | sc68cal, ^ | 18:32 |
| mestery | ihrachyshka sc68cal: +2 from me, thanks for the quick re-spin | 18:34 |
| *** jlibosva has left #openstack-neutron | 18:34 | |
| *** jlibosva has joined #openstack-neutron | 18:34 | |
| ihrachyshka | mestery, oh, also that one popped up before: https://review.openstack.org/#/c/210625/ that fixes a bug when ml2 plugin fails to determine a change to notify about when qos_policy_id is unset for a port | 18:34 |
| mestery | ihrachyshka: Really? Looking. | 18:34 |
| * sc68cal pushes the big red shiny button | 18:35 | |
| ihrachyshka | mestery, yeah, that's the price of not having fullstack :( | 18:35 |
| mestery | heh :) | 18:35 |
| *** sdake_ is now known as sdake | 18:35 | |
| ihrachyshka | mestery, btw we have first fullstack tests in review but they will wait for master merge and client merge since fullstack relies on real client | 18:35 |
| mestery | Ack | 18:36 |
| *** rm_work|away is now known as rm_work | 18:36 | |
| jlibosva | ihrachyshka: ajo this a good story to read before going to bed https://review.openstack.org/#/c/211262/ | 18:37 |
| ihrachyshka | jlibosva, long?.. I am about to end it. do you have more??? | 18:37 |
| *** Sukhdev has quit IRC | 18:37 | |
| aveiga | sc68cal: thanks for the heads up; we'll teake care of it | 18:37 |
| jlibosva | ihrachyshka: it's a short story | 18:38 |
| ihrachyshka | that would work for my son | 18:38 |
| jlibosva | any hrachyshka is good I guess | 18:38 |
| *** jlibosva has left #openstack-neutron | 18:39 | |
| *** jlibosva has joined #openstack-neutron | 18:39 | |
| sc68cal | aveiga: no problem - i'll send you some beer as thanks | 18:41 |
| aveiga | ha, no need. Fixing IPv6 is its own reward ;) | 18:41 |
| ihrachyshka | jreeves, hi. | 18:42 |
| regXboi | that sounds like a koan | 18:42 |
| jreeves | Hello | 18:42 |
| jreeves | ihrachyshka: something you need from me? | 18:43 |
| ihrachyshka | jreeves, I see functional test_qos_bw_limit fails. do you test the patch locally? I think it may be more effective for you than hitting the gate and waiting for response (the queue is quite long, so no immediate replies) | 18:43 |
| banix | apuimedo: Hi Toni, Just noticed this: https://github.com/docker/libnetwork/issues/414 should tell tfukushima: So the short answer is we cannot do that as of now | 18:43 |
| jreeves | yes, I ran it locally and it passed. not sure why it failed there | 18:43 |
| *** Tedster has quit IRC | 18:44 | |
| *** claudiub has quit IRC | 18:44 | |
| *** Tedster has joined #openstack-neutron | 18:44 | |
| *** yamahata has joined #openstack-neutron | 18:45 | |
| *** crose has joined #openstack-neutron | 18:45 | |
| *** slaweq has quit IRC | 18:46 | |
| jlibosva | ihrachyshka: jreeves any chance there is a race between greenthread switching to the one handling port? I experienced that locally | 18:47 |
| jlibosva | race in the test itself I mean, not in the code | 18:48 |
| *** andyhky has quit IRC | 18:49 | |
| *** andyhky has joined #openstack-neutron | 18:49 | |
| *** crose has quit IRC | 18:51 | |
| jreeves | jlibosva: hmmm...perhaps | 18:51 |
| *** slaweq has joined #openstack-neutron | 18:52 | |
| ihrachyshka | jreeves, hm, so I look in the code... you hit some Exception, and try to cleanup from it. | 18:52 |
| ihrachyshka | hm... TypeError: string indices must be integers | 18:53 |
| jreeves | yes, correct | 18:53 |
| *** enikanorov2 has quit IRC | 18:53 | |
| *** gsagie__ has joined #openstack-neutron | 18:53 | |
| *** piyanai has joined #openstack-neutron | 18:53 | |
| *** josecastroleon has quit IRC | 18:54 | |
| *** crose has joined #openstack-neutron | 18:54 | |
| *** dave-mcc_ has quit IRC | 18:55 | |
| *** marun has joined #openstack-neutron | 18:55 | |
| *** abhiraut has quit IRC | 18:55 | |
| *** abhiraut has joined #openstack-neutron | 18:56 | |
| *** rushil has quit IRC | 18:58 | |
| ajo | jlibosva: a few nits here: https://review.openstack.org/#/c/211262/ | 18:58 |
| ajo | but I +2'd because I think those are not important, the test should work as you wrote it :) | 18:58 |
| ajo | if you do it now, we can re+2 / +A or ... otherwise we can follow up later with cleanups | 18:59 |
| ajo | but awesome work dude, I thought I was going to need more work in the test framework for ovs agent :) | 18:59 |
| *** rushil has joined #openstack-neutron | 19:00 | |
| *** baoli_ has quit IRC | 19:01 | |
| *** baoli has joined #openstack-neutron | 19:02 | |
| *** britthouser has joined #openstack-neutron | 19:02 | |
| *** crose has quit IRC | 19:02 | |
| *** enikanorov2 has joined #openstack-neutron | 19:03 | |
| *** crc32 has joined #openstack-neutron | 19:05 | |
| *** agireud_ has joined #openstack-neutron | 19:05 | |
| *** madhu_ak has joined #openstack-neutron | 19:05 | |
| jlibosva | ajo: I can do a fast re-spin if you want | 19:05 |
| *** agireud has quit IRC | 19:06 | |
| *** agireud_ is now known as agireud | 19:06 | |
| openstackgerrit | Lynn Li proposed openstack/neutron: Add high-level functional/integration DVR tests https://review.openstack.org/178880 | 19:07 |
| *** dave-mccowan has joined #openstack-neutron | 19:07 | |
| *** pai15 has joined #openstack-neutron | 19:07 | |
| *** mctaylor has joined #openstack-neutron | 19:08 | |
| *** gangil has quit IRC | 19:08 | |
| *** bapalm_ has quit IRC | 19:08 | |
| *** abhiraut has quit IRC | 19:08 | |
| *** vikram has quit IRC | 19:10 | |
| jlibosva | ajo: done :) | 19:13 |
| *** gsagie__ has quit IRC | 19:13 | |
| *** Sukhdev has joined #openstack-neutron | 19:14 | |
| *** rkukura has quit IRC | 19:14 | |
| apuimedo | banix: thanks. Good find! | 19:17 |
| kevinbenton | Sukhdev: ping | 19:17 |
| *** fawadkhaliq has joined #openstack-neutron | 19:18 | |
| openstackgerrit | Cyril Roelandt proposed openstack/neutron: Python 3: encode unicode response bodies https://review.openstack.org/204791 | 19:18 |
| Sukhdev | kevinbenton: Hey Kevin - can you have a look at this - https://review.openstack.org/#/c/197774/ | 19:19 |
| ihrachyshka | otherwiseguy, what does .execute() return in ovs_lib? | 19:20 |
| kevinbenton | Sukhdev: done | 19:20 |
| Sukhdev | kevinbenton: thanks | 19:21 |
| anteaya | so meeting is tomorrow? https://wiki.openstack.org/wiki/Network/Meetings | 19:21 |
| kevinbenton | anteaya: yeah | 19:22 |
| anteaya | kevinbenton: thanks | 19:22 |
| *** rkukura has joined #openstack-neutron | 19:24 | |
| openstackgerrit | Sean M. Collins proposed openstack/neutron-fwaas: Add oslotest dependency to test-requirements https://review.openstack.org/211302 | 19:25 |
| *** Aish has quit IRC | 19:25 | |
| *** britthouser has quit IRC | 19:26 | |
| *** marun has left #openstack-neutron | 19:26 | |
| kevinbenton | sc68cal: thanks for the pointer. i'll check that out | 19:26 |
| *** britthouser has joined #openstack-neutron | 19:26 | |
| openstackgerrit | Merged openstack/neutron: DVR: fix router scheduling https://review.openstack.org/199032 | 19:27 |
| openstackgerrit | Ryan Moats proposed openstack/neutron: manual add/remove router for dvr_snat agent https://review.openstack.org/122024 | 19:29 |
| openstackgerrit | Ryan Moats proposed openstack/neutron: DVR: fix router rescheduling https://review.openstack.org/199514 | 19:29 |
| *** sdake_ has joined #openstack-neutron | 19:29 | |
| *** moshele has joined #openstack-neutron | 19:30 | |
| *** yapeng has quit IRC | 19:31 | |
| *** sdake has quit IRC | 19:33 | |
| *** sridhar_ram has quit IRC | 19:33 | |
| sc68cal | kevinbenton: no problem | 19:34 |
| *** sdake has joined #openstack-neutron | 19:35 | |
| *** TrevorV has quit IRC | 19:35 | |
| *** britthouser has quit IRC | 19:36 | |
| *** sdake_ has quit IRC | 19:38 | |
| *** jlibosva has quit IRC | 19:39 | |
| *** britthouser has joined #openstack-neutron | 19:40 | |
| *** lori is now known as lori|away | 19:40 | |
| *** nlahouti has quit IRC | 19:41 | |
| *** abregman has joined #openstack-neutron | 19:42 | |
| ihrachyshka | jreeves, I call the command that test calls to create the queue (/usr/bin/ovs-vsctl --timeout=10 --oneline --format=json -- create Queue other_config:burst=70000 other_config:max-rate=700000 other_config:min-rate=700000 external_ids:id=test-port306100) and I get back a single string: "e22275b6-2118-4616-a29c-27efa27e773e" that's why your test fails: first [0] gets the first char in the string, then you apply ['_uui | 19:43 |
| ihrachyshka | d'] and it obviously fails | 19:43 |
| *** ajo_ has joined #openstack-neutron | 19:44 | |
| *** Aish has joined #openstack-neutron | 19:44 | |
| *** bapalm__ has quit IRC | 19:44 | |
| ajo_ | ping ihrachyshka : added some extra content in comments to: https://review.openstack.org/#/c/211259/ | 19:44 |
| ihrachyshka | ajo, ack, I'll update it tomorrow. I wonder whether amuller will be happy about it. | 19:44 |
| *** bapalm has joined #openstack-neutron | 19:45 | |
| ihrachyshka | amuller, can you check the ajo's comment? | 19:45 |
| *** bapalm has quit IRC | 19:46 | |
| openstackgerrit | Miguel Lavalle proposed openstack/neutron: Add dns_label processing for Ports https://review.openstack.org/200952 | 19:46 |
| *** bapalm has joined #openstack-neutron | 19:46 | |
| openstackgerrit | Ryan Moats proposed openstack/neutron: manual add/remove router for dvr_snat agent https://review.openstack.org/122024 | 19:48 |
| amuller | ihrachyshka: in a meeting atm | 19:48 |
| *** ijw has quit IRC | 19:56 | |
| openstackgerrit | sean mooney proposed openstack/neutron: Add config option to specify ovs datapath. https://review.openstack.org/197209 | 19:57 |
| *** annegentle has quit IRC | 19:57 | |
| *** salv-orlando has quit IRC | 19:57 | |
| *** Piet has joined #openstack-neutron | 19:58 | |
| ajo_ | ihrachyshka, https://review.openstack.org/211311 | 19:58 |
| ajo_ | a tiny fix for m2l->agents propagation of qos_profile_id port removal | 19:59 |
| ajo_ | ml2 | 19:59 |
| ihrachyshka | ajo, https://review.openstack.org/#/c/210625/ ? | 20:00 |
| ajo_ | ihrachyshka, looking where unit tests belong | 20:00 |
| ajo_ | ouch | 20:00 |
| ihrachyshka | you reviewed it | 20:00 |
| ihrachyshka | I wonder whether it's the same? | 20:00 |
| ihrachyshka | looks a bit different, but I guess the issue is the same | 20:00 |
| ajo_ | ihrachyshka, yes, but there's a missing part in the plugin.py | 20:00 |
| ajo_ | otherwise it will explode by not finding the key in port_update, hhmmm | 20:01 |
| ajo_ | in your implementation it wont | 20:01 |
| ajo_ | ok I can drop my patch | 20:01 |
| ajo_ | :D | 20:01 |
| ajo_ | I will test with your's on top | 20:01 |
| ihrachyshka | ajo, basically, I will only note that it's worth reading comments | 20:01 |
| ihrachyshka | ajo, I refered to the patch several times in irc and gerrit | 20:02 |
| ajo_ | ihrachyshka, I totally forgot when testing, | 20:02 |
| *** salv-orlando has joined #openstack-neutron | 20:02 | |
| ihrachyshka | ajo, anyway, you paid the price :) | 20:03 |
| ihrachyshka | it's good your solution is similar. meaning, it's the right direction :) | 20:03 |
| ajo_ | ':) | 20:03 |
| *** rushil has quit IRC | 20:03 | |
| *** rushil has joined #openstack-neutron | 20:05 | |
| openstackgerrit | Carl Baldwin proposed openstack/neutron: Add address scope to ports in RPC response to L3 agent https://review.openstack.org/189948 | 20:06 |
| ajo_ | ihrachyshka, yours work, tested | 20:06 |
| ihrachyshka | jreeves, I've posted comments, those should fix the test. | 20:06 |
| ihrachyshka | ajo, cool. | 20:06 |
| *** ijw has joined #openstack-neutron | 20:06 | |
| jreeves | ihrachyshka: thanks! I'll check it out | 20:07 |
| ihrachyshka | jreeves, so basically, db_find returns a list of dicts with _uuid keys. but db_create returns just uuid string | 20:08 |
| jreeves | yep, that's the direction I was looking to | 20:08 |
| *** bapalm has quit IRC | 20:09 | |
| *** bapalm has joined #openstack-neutron | 20:09 | |
| *** ijw has quit IRC | 20:11 | |
| *** bapalm has quit IRC | 20:12 | |
| *** bapalm has joined #openstack-neutron | 20:12 | |
| *** johnbelamaric has quit IRC | 20:12 | |
| openstackgerrit | Doug Wiegley proposed openstack/neutron: Revert "Enable VPN plugin for API test" https://review.openstack.org/211317 | 20:15 |
| openstackgerrit | Carl Baldwin proposed openstack/neutron: Add address scope to floating IPs in RPC response to L3 agent https://review.openstack.org/189500 | 20:16 |
| otherwiseguy | ihrachyshka: it returns what ever is set as the result attribute of the Command. Care should be taken across implementations to ensure that both return the same things. | 20:16 |
| * otherwiseguy catches up on scrollback | 20:17 | |
| *** vivek-ebay has quit IRC | 20:17 | |
| otherwiseguy | I usually just try to make sure the ovsdb implementation matches what the ovs-vsctl does, since there is more control with the ovsdb. | 20:17 |
| ihrachyshka | ajo, what's the version of ovs that started supporting QoS queue? | 20:18 |
| openstackgerrit | Doug Wiegley proposed openstack/neutron: Revert "Enable VPN plugin for API test" https://review.openstack.org/211317 | 20:19 |
| ihrachyshka | otherwiseguy, hm. but overall, seems like db_find returns a list of dicts and db_create returns a string of uuid | 20:19 |
| otherwiseguy | ihrachyshka: yes. | 20:19 |
| ihrachyshka | ok. I think then that was the culprit of the test failure :) | 20:19 |
| otherwiseguy | db_create on the ovsdb side might return the Row object (did last I checked), so probably needs to be updated to return the Row._uuid or something. | 20:20 |
| *** abhiraut has joined #openstack-neutron | 20:20 | |
| ajo_ | ihrachyshka, yey, and the low level OVS limit is effectively working :) | 20:21 |
| ihrachyshka | ajo, nah, who cares? | 20:21 |
| ihrachyshka | :) | 20:21 |
| *** gangil has joined #openstack-neutron | 20:22 | |
| ihrachyshka | otherwiseguy, meh. ok. does it mean we need to patch idl impl itself? | 20:22 |
| ihrachyshka | ajo, so what's the OVS version that has qos support? I want to have it in devref. | 20:22 |
| amuller | ihrachyshka: ajo_: What QoS version supported queues/profiles btw? since gate is running ovs 2.0, it's quite old | 20:22 |
| ihrachyshka | amuller, well, per blogs, it was in 1.9 already | 20:22 |
| amuller | good | 20:23 |
| amuller | so we'll be able to test it with functional/fullstack tests | 20:23 |
| ihrachyshka | amuller, will it suffice to say in devref it was at least in 2.x? | 20:23 |
| ihrachyshka | amuller, we not only will, we have some patches in fly that you should review :) | 20:23 |
| ajo_ | hmm, it's misbehaving now | 20:23 |
| ajo_ | but it was working before | 20:23 |
| * ajo_ checks | 20:23 | |
| *** ronis has quit IRC | 20:24 | |
| ajo_ | $ neutron qos-bandwidth-limit-rule-update 3b1d1444-8807-4c4a-acaf-13bf96cdb6bf bw100k --max_kbps 1000 --max_burst_kbps=100 | 20:24 |
| ajo_ | Updated bandwidth_limit_rule: 3b1d1444-8807-4c4a-acaf-13bf96cdb6bf | 20:24 |
| ajo_ | [vagrant@devstack neutron]$ ssh cirros@10.0.0.3 'dd if=/dev/zero bs=1M count=1000' | 20:24 |
| ajo_ | :-) | 20:24 |
| ajo_ | and then... /me looks at nload over the port | 20:24 |
| *** moshele has quit IRC | 20:24 | |
| *** pc_m has quit IRC | 20:25 | |
| openstackgerrit | Carl Baldwin proposed openstack/neutron: Add address scope to floating IPs in RPC response to L3 agent https://review.openstack.org/189500 | 20:25 |
| *** alex_klimov has joined #openstack-neutron | 20:26 | |
| *** ijw has joined #openstack-neutron | 20:27 | |
| ajo_ | lol | 20:27 |
| ajo_ | we have a magnitude error with ovs low level handling | 20:28 |
| ajo_ | ihrachyshka, amuller ^ | 20:28 |
| ajo_ | setting bps where we should set kbps | 20:28 |
| ihrachyshka | amuller, ajo I updated devref with ovs agent details: https://review.openstack.org/211259 | 20:28 |
| amuller | lol | 20:28 |
| *** johnbelamaric has joined #openstack-neutron | 20:28 | |
| *** fawadkhaliq has quit IRC | 20:29 | |
| ihrachyshka | ajo_, hm. wasn't ovs controlled in bits units? | 20:29 |
| *** johnbelamaric has quit IRC | 20:29 | |
| ihrachyshka | that sounded weird to me before, but I assumed people know better :) | 20:29 |
| *** johnbelamaric has joined #openstack-neutron | 20:29 | |
| ihrachyshka | and also, who cares about such a tiny glitch? | 20:29 |
| ihrachyshka | the main thing is that API works | 20:30 |
| ihrachyshka | :) | 20:30 |
| *** ijw_ has joined #openstack-neutron | 20:30 | |
| ajo_ | lol | 20:30 |
| ajo_ | ihrachyshka, it does not match my initial experiments where I was using bps too: https://github.com/mangelajo/ovs-experiments/blob/master/qos/qos_traffic_shapping.sh#L139 | 20:30 |
| ajo_ | btw it seems that it works if I set /1000 via API | 20:31 |
| *** ijw__ has joined #openstack-neutron | 20:31 | |
| ihrachyshka | ajo, ok, let's then just kill the multiplier. it's small. I wonder whether it may be a difference in ovs versions? | 20:31 |
| ihrachyshka | I don't know how compat new versions usually are | 20:31 |
| *** ijw has quit IRC | 20:32 | |
| ajo_ | ihrachyshka I will check what's wrong, gsagie didn't set things 100% exactly to my example, I'm testing yet | 20:33 |
| ihrachyshka | ajo, ack. and I'll leave to the bed :) | 20:33 |
| ihrachyshka | cu tomorrow | 20:33 |
| *** rushil has quit IRC | 20:34 | |
| ajo_ | ihrachyshka, thanks, gn, :) | 20:34 |
| *** lazy_prince has quit IRC | 20:34 | |
| *** ijw_ has quit IRC | 20:34 | |
| *** ihrachyshka has quit IRC | 20:38 | |
| *** lazy_prince has joined #openstack-neutron | 20:39 | |
| *** john-davidge has quit IRC | 20:40 | |
| *** mctaylor has quit IRC | 20:41 | |
| *** annegentle has joined #openstack-neutron | 20:41 | |
| *** claudiub has joined #openstack-neutron | 20:41 | |
| *** regXboi is now known as regXboi_awa | 20:41 | |
| *** ajo_ has quit IRC | 20:42 | |
| *** chadix has quit IRC | 20:43 | |
| *** ecrosso has quit IRC | 20:44 | |
| *** bapalm has quit IRC | 20:44 | |
| *** nlahouti has joined #openstack-neutron | 20:44 | |
| *** bapalm has joined #openstack-neutron | 20:45 | |
| *** bapalm_ has joined #openstack-neutron | 20:46 | |
| *** nyechiel has joined #openstack-neutron | 20:47 | |
| *** padkrish has joined #openstack-neutron | 20:47 | |
| *** piyanai has quit IRC | 20:47 | |
| *** moshele has joined #openstack-neutron | 20:47 | |
| *** mwagner_lap has quit IRC | 20:47 | |
| *** chlong has joined #openstack-neutron | 20:49 | |
| *** bapalm has quit IRC | 20:49 | |
| *** bapalm_ has quit IRC | 20:50 | |
| *** piyanai has joined #openstack-neutron | 20:50 | |
| *** changbl has quit IRC | 20:51 | |
| *** rkukura has quit IRC | 20:58 | |
| *** thorst has quit IRC | 21:02 | |
| *** merooney has quit IRC | 21:02 | |
| *** tongli has quit IRC | 21:03 | |
| *** sridhar_ram has joined #openstack-neutron | 21:04 | |
| *** tiswanso has quit IRC | 21:06 | |
| *** achanda has quit IRC | 21:07 | |
| *** sthillma has quit IRC | 21:09 | |
| *** jckasper has quit IRC | 21:09 | |
| *** vivek-ebay has joined #openstack-neutron | 21:09 | |
| *** sthillma has joined #openstack-neutron | 21:10 | |
| *** yamamoto has joined #openstack-neutron | 21:11 | |
| *** ijw has joined #openstack-neutron | 21:11 | |
| *** thumpba has quit IRC | 21:12 | |
| *** baoli has quit IRC | 21:13 | |
| *** ijw__ has quit IRC | 21:13 | |
| *** Sukhdev has quit IRC | 21:13 | |
| *** pradk has quit IRC | 21:14 | |
| openstackgerrit | Merged openstack/neutron: Neutron-Ironic integration patch https://review.openstack.org/197774 | 21:14 |
| *** sthillma_ has joined #openstack-neutron | 21:15 | |
| *** nyechiel has quit IRC | 21:15 | |
| *** erlarese has quit IRC | 21:17 | |
| *** piyanai has quit IRC | 21:17 | |
| *** abhiraut has quit IRC | 21:17 | |
| *** sthillma has quit IRC | 21:18 | |
| *** sthillma_ is now known as sthillma | 21:18 | |
| *** abhiraut has joined #openstack-neutron | 21:18 | |
| openstackgerrit | Merged openstack/neutron-fwaas: switch to multi branch migration tree for liberty https://review.openstack.org/206691 | 21:19 |
| *** piyanai has joined #openstack-neutron | 21:19 | |
| *** yamamoto has quit IRC | 21:19 | |
| *** julim has quit IRC | 21:22 | |
| *** rushil has joined #openstack-neutron | 21:23 | |
| *** abregman has quit IRC | 21:25 | |
| *** Aish has left #openstack-neutron | 21:26 | |
| *** rushil_ has joined #openstack-neutron | 21:28 | |
| *** pai15 has quit IRC | 21:29 | |
| *** rushil has quit IRC | 21:31 | |
| *** thorst has joined #openstack-neutron | 21:34 | |
| *** abregman has joined #openstack-neutron | 21:37 | |
| *** achanda has joined #openstack-neutron | 21:40 | |
| *** sthillma_ has joined #openstack-neutron | 21:41 | |
| *** zhipeng has quit IRC | 21:42 | |
| *** padkrish has quit IRC | 21:43 | |
| *** padkrish has joined #openstack-neutron | 21:44 | |
| *** sthillma has quit IRC | 21:44 | |
| *** sthillma_ is now known as sthillma | 21:44 | |
| *** amuller has quit IRC | 21:47 | |
| *** johnbelamaric has quit IRC | 21:47 | |
| *** tfukushima has joined #openstack-neutron | 21:47 | |
| *** padkrish has quit IRC | 21:48 | |
| *** busterswt has quit IRC | 21:54 | |
| *** regXboi_awa is now known as regXboi | 21:55 | |
| *** rushil_ has quit IRC | 21:55 | |
| *** bapalm has joined #openstack-neutron | 21:56 | |
| *** annegentle has quit IRC | 21:56 | |
| *** abhiraut has quit IRC | 21:56 | |
| *** regXboi has quit IRC | 21:57 | |
| *** mwagner_lap has joined #openstack-neutron | 21:57 | |
| *** yamamoto has joined #openstack-neutron | 21:58 | |
| *** abhiraut has joined #openstack-neutron | 21:59 | |
| *** sdake has quit IRC | 22:01 | |
| *** kbringard has quit IRC | 22:02 | |
| *** bapalm has quit IRC | 22:02 | |
| *** bapalm has joined #openstack-neutron | 22:03 | |
| *** yamamoto has quit IRC | 22:05 | |
| *** bapalm has quit IRC | 22:07 | |
| *** dane_leblanc has quit IRC | 22:08 | |
| *** padkrish has joined #openstack-neutron | 22:12 | |
| *** yamamoto has joined #openstack-neutron | 22:12 | |
| *** crc32 has quit IRC | 22:16 | |
| *** padkrish has quit IRC | 22:17 | |
| *** bapalm has joined #openstack-neutron | 22:18 | |
| *** lindycoder has quit IRC | 22:18 | |
| *** sthillma_ has joined #openstack-neutron | 22:22 | |
| *** agireud has quit IRC | 22:24 | |
| *** edmondsw has quit IRC | 22:25 | |
| *** sthillma has quit IRC | 22:25 | |
| *** sthillma_ is now known as sthillma | 22:25 | |
| *** abregman has quit IRC | 22:25 | |
| *** moshele has quit IRC | 22:26 | |
| *** bapalm has quit IRC | 22:29 | |
| *** salv-orl_ has joined #openstack-neutron | 22:31 | |
| *** salv-orlando has quit IRC | 22:33 | |
| openstackgerrit | Cedric Brandily proposed openstack/neutron: Python 3: encode unicode response bodies https://review.openstack.org/204791 | 22:33 |
| *** jecarey has quit IRC | 22:33 | |
| *** annegentle has joined #openstack-neutron | 22:36 | |
| *** Sukhdev has joined #openstack-neutron | 22:37 | |
| *** abregman has joined #openstack-neutron | 22:37 | |
| *** padkrish has joined #openstack-neutron | 22:37 | |
| *** padkrish has quit IRC | 22:38 | |
| *** jamielennox|away is now known as jamielennox | 22:38 | |
| *** padkrish has joined #openstack-neutron | 22:38 | |
| *** r-daneel has quit IRC | 22:39 | |
| *** PabloPhoenix has quit IRC | 22:39 | |
| *** zzzeek has quit IRC | 22:42 | |
| *** julim has joined #openstack-neutron | 22:43 | |
| *** vivek-ebay has quit IRC | 22:44 | |
| *** claudiub has quit IRC | 22:47 | |
| *** banix has quit IRC | 22:50 | |
| *** alex_klimov has quit IRC | 22:54 | |
| *** krtaylor has quit IRC | 22:59 | |
| *** merooney has joined #openstack-neutron | 23:02 | |
| *** ivar-laz_ has joined #openstack-neutron | 23:02 | |
| *** mlavalle has quit IRC | 23:04 | |
| *** sthillma_ has joined #openstack-neutron | 23:05 | |
| *** ivar-lazzaro has quit IRC | 23:05 | |
| *** sthillma has quit IRC | 23:07 | |
| *** sthillma_ is now known as sthillma | 23:07 | |
| *** thedodd has quit IRC | 23:10 | |
| *** ijw_ has joined #openstack-neutron | 23:13 | |
| *** yfried is now known as yfried|afk | 23:14 | |
| *** ijw has quit IRC | 23:16 | |
| *** bharath has quit IRC | 23:20 | |
| *** bharath has joined #openstack-neutron | 23:21 | |
| *** julim has quit IRC | 23:21 | |
| *** shwetaap has quit IRC | 23:23 | |
| *** merooney has quit IRC | 23:25 | |
| *** vivek-ebay has joined #openstack-neutron | 23:26 | |
| *** annegentle has quit IRC | 23:26 | |
| *** yamamoto has quit IRC | 23:27 | |
| *** yushiro has joined #openstack-neutron | 23:29 | |
| *** hichihara has joined #openstack-neutron | 23:29 | |
| *** piyanai has quit IRC | 23:30 | |
| *** sdake has joined #openstack-neutron | 23:31 | |
| *** ijw_ has quit IRC | 23:31 | |
| *** ijw has joined #openstack-neutron | 23:31 | |
| *** tfukushima has quit IRC | 23:32 | |
| *** ijw has quit IRC | 23:33 | |
| *** ijw has joined #openstack-neutron | 23:33 | |
| *** yfried|afk is now known as yfried | 23:34 | |
| *** bharath has quit IRC | 23:36 | |
| *** bharath has joined #openstack-neutron | 23:36 | |
| *** bharath has quit IRC | 23:37 | |
| *** changbl has joined #openstack-neutron | 23:38 | |
| *** ijw has quit IRC | 23:39 | |
| *** merooney has joined #openstack-neutron | 23:39 | |
| *** sthillma_ has joined #openstack-neutron | 23:40 | |
| *** ijw has joined #openstack-neutron | 23:40 | |
| *** sthillma has quit IRC | 23:42 | |
| *** sthillma_ is now known as sthillma | 23:42 | |
| *** jaypipes has quit IRC | 23:42 | |
| abhiraut | ping: HenryG | 23:42 |
| HenryG | abhiraut: pong | 23:43 |
| yushiro | Good evening mestery, thank you for rechecking my patch https://review.openstack.org/#/c/205902/ | 23:43 |
| abhiraut | what is the purpose of alembic_migrations/external.py module? | 23:43 |
| *** ijw_ has joined #openstack-neutron | 23:44 | |
| *** daneyon has quit IRC | 23:44 | |
| *** daneyon has joined #openstack-neutron | 23:44 | |
| abhiraut | more specifically .. do I update this file if I were to move models of a monolithic plugin from neutron to vendor repo? | 23:44 |
| HenryG | When an external repo (like *aas, or networking-foo) takes control of the alembic migrations of the DB tables it owns, then those tables must be listed in external.py | 23:45 |
| HenryG | abhiraut: yes, when you move the models out of the neutron tree is when you add the table names to external.py | 23:45 |
| *** PabloPhoenix has joined #openstack-neutron | 23:45 | |
| abhiraut | HenryG: ok .. so it doesnt matter if it is a ML2 mech driver or a monoltihic plugin.. i was just thrown off by the DRIVER_TABLE .. sounded more ML2 to me :) | 23:46 |
| *** ijw has quit IRC | 23:46 | |
| HenryG | abhiraut: If you want you can rename from DRIVER_TABLES to EXTERNAL_PROJECT_TABLES :) | 23:47 |
| yushiro | haleyb, ping | 23:48 |
| *** gildub has joined #openstack-neutron | 23:48 | |
| *** ajmiller has quit IRC | 23:50 | |
| *** gampel has quit IRC | 23:50 | |
| abhiraut | HenryG: ha! thanks for the suggestion. | 23:51 |
| *** dannywilson has quit IRC | 23:51 | |
| *** tfukushima has joined #openstack-neutron | 23:52 | |
| *** zzzeek has joined #openstack-neutron | 23:52 | |
| HenryG | abhiraut: A better name would actually be DECOMPOSED_PROJECT_TABLES | 23:52 |
| *** sdake has quit IRC | 23:53 | |
| abhiraut | HenryG: long.. but more correct | 23:53 |
| *** sdake has joined #openstack-neutron | 23:56 | |
| *** tfukushima has quit IRC | 23:57 | |
| *** carl_baldwin has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
