| jog0 | so just saw a spike in gate-tempest-devstack-vm-neutron failures | 00:03 |
|---|---|---|
| *** julim has quit IRC | 00:04 | |
| *** otherwiseguy has quit IRC | 00:14 | |
| *** qingluo has joined #openstack-neutron | 00:21 | |
| *** nati_uen_ has joined #openstack-neutron | 00:23 | |
| *** nati_uen_ has quit IRC | 00:25 | |
| *** nati_uen_ has joined #openstack-neutron | 00:26 | |
| *** nati_ueno has quit IRC | 00:27 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Send only one notification is sent on port_udate https://review.openstack.org/58415 | 00:27 |
| *** jasonv has joined #openstack-neutron | 00:32 | |
| *** matsuhashi has joined #openstack-neutron | 00:32 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Send only one agent notification on port update https://review.openstack.org/58415 | 00:36 |
| *** carl_baldwin has joined #openstack-neutron | 00:36 | |
| *** armax has joined #openstack-neutron | 00:37 | |
| idella4 | methinks the brittle internet connection @ the motel cracked. anteaya doesn't like the sound of that <crack> | 00:38 |
| *** pete5 has quit IRC | 00:44 | |
| *** yamahata_ has joined #openstack-neutron | 00:48 | |
| *** otherwiseguy has joined #openstack-neutron | 01:02 | |
| *** carl_baldwin has quit IRC | 01:02 | |
| *** banix has joined #openstack-neutron | 01:04 | |
| qingluo | how to set up network config for two nodes openstack(SINGLE NIC)? i have a controller node->192.168.0.63, a compute node->192.168.0.69. how to set bridge for host. | 01:08 |
| *** jorisroovers has quit IRC | 01:14 | |
| *** nati_ueno has joined #openstack-neutron | 01:14 | |
| *** alagalah has joined #openstack-neutron | 01:16 | |
| *** jorisroovers has joined #openstack-neutron | 01:17 | |
| *** nati_uen_ has quit IRC | 01:17 | |
| openstackgerrit | Zhenguo Niu proposed a change to openstack/neutron: Update Zhenguo Niu's mailmap https://review.openstack.org/53805 | 01:26 |
| openstackgerrit | berlin proposed a change to openstack/neutron: Fix showing nonexistent NetworkGateway throws 500 instead of 404 https://review.openstack.org/58423 | 01:29 |
| *** Jianyong has joined #openstack-neutron | 01:30 | |
| *** jorisroovers has quit IRC | 01:31 | |
| *** SumitNaiksatam has quit IRC | 01:54 | |
| *** yamahata_ has quit IRC | 02:03 | |
| *** ljjjustin has joined #openstack-neutron | 02:03 | |
| *** netredem-jr has joined #openstack-neutron | 02:03 | |
| *** nati_uen_ has joined #openstack-neutron | 02:03 | |
| *** carl_baldwin has joined #openstack-neutron | 02:04 | |
| *** nati_ueno has quit IRC | 02:06 | |
| *** coolsvap has quit IRC | 02:06 | |
| *** netredem-jr has quit IRC | 02:08 | |
| *** carl_baldwin has quit IRC | 02:08 | |
| *** nati_uen_ has quit IRC | 02:09 | |
| *** nati_ueno has joined #openstack-neutron | 02:10 | |
| *** netredem-jr has joined #openstack-neutron | 02:15 | |
| *** nati_ueno has quit IRC | 02:21 | |
| *** armax has left #openstack-neutron | 02:31 | |
| *** dims has quit IRC | 02:38 | |
| *** networkstatic has joined #openstack-neutron | 02:41 | |
| *** networkstatic has quit IRC | 02:41 | |
| *** arosen has quit IRC | 02:45 | |
| *** arosen has joined #openstack-neutron | 02:47 | |
| *** matsuhashi has quit IRC | 03:11 | |
| *** carl_baldwin has joined #openstack-neutron | 03:16 | |
| *** carl_baldwin has quit IRC | 03:20 | |
| *** clev has joined #openstack-neutron | 03:26 | |
| *** nati_ueno has joined #openstack-neutron | 03:28 | |
| *** arosen has quit IRC | 03:28 | |
| *** clev has quit IRC | 03:32 | |
| *** nati_ueno has quit IRC | 03:35 | |
| *** nati_ueno has joined #openstack-neutron | 03:35 | |
| openstackgerrit | Nachi Ueno proposed a change to openstack/neutron: Improve vif attributes related with firewalling https://review.openstack.org/21946 | 03:38 |
| *** banix has quit IRC | 03:41 | |
| *** otherwiseguy has quit IRC | 03:45 | |
| openstackgerrit | Zhang Hua proposed a change to openstack/neutron: Replace uuidutils.generate_uuid() with str(uuid.uuid4()) https://review.openstack.org/58234 | 03:45 |
| *** coolsvap has joined #openstack-neutron | 03:50 | |
| *** jasonv has quit IRC | 03:57 | |
| *** jorisroovers has joined #openstack-neutron | 03:59 | |
| *** jroovers has joined #openstack-neutron | 04:01 | |
| *** jorisroovers has quit IRC | 04:03 | |
| *** suresh12 has quit IRC | 04:05 | |
| *** chandankumar has joined #openstack-neutron | 04:05 | |
| *** yamahata_ has joined #openstack-neutron | 04:14 | |
| *** carl_baldwin has joined #openstack-neutron | 04:16 | |
| *** matsuhashi has joined #openstack-neutron | 04:19 | |
| *** carl_baldwin has quit IRC | 04:20 | |
| openstackgerrit | Jianing YANG proposed a change to openstack/neutron: Exit by raising SystemExit instead of calling sys.exit(). https://review.openstack.org/58433 | 04:23 |
| *** nati_uen_ has joined #openstack-neutron | 04:29 | |
| *** coolsvap has quit IRC | 04:31 | |
| *** nati_ueno has quit IRC | 04:33 | |
| *** yamahata_ has quit IRC | 04:47 | |
| *** banix has joined #openstack-neutron | 04:55 | |
| *** coolsvap_ has joined #openstack-neutron | 05:05 | |
| *** zigo_ has quit IRC | 05:10 | |
| *** zigo has joined #openstack-neutron | 05:11 | |
| *** carl_baldwin has joined #openstack-neutron | 05:16 | |
| *** carl_baldwin has quit IRC | 05:21 | |
| *** coolsvap_ has quit IRC | 05:24 | |
| *** coolsvap has joined #openstack-neutron | 05:29 | |
| *** bashok has joined #openstack-neutron | 05:30 | |
| *** netredem-jr has quit IRC | 05:41 | |
| *** networkstatic has joined #openstack-neutron | 05:41 | |
| *** richardboswell has quit IRC | 05:47 | |
| *** matsuhashi has quit IRC | 05:50 | |
| *** richardboswell has joined #openstack-neutron | 05:50 | |
| *** matsuhashi has joined #openstack-neutron | 05:51 | |
| *** matsuhas_ has joined #openstack-neutron | 05:53 | |
| *** matsuhashi has quit IRC | 05:53 | |
| *** steven-weston_ has joined #openstack-neutron | 05:59 | |
| bashok | salv-orlando, hi | 06:02 |
| *** banix has quit IRC | 06:04 | |
| *** yfried has joined #openstack-neutron | 06:14 | |
| *** carl_baldwin has joined #openstack-neutron | 06:16 | |
| *** matsuhas_ has quit IRC | 06:20 | |
| *** carl_baldwin has quit IRC | 06:20 | |
| *** coolsvap has quit IRC | 06:26 | |
| *** matsuhashi has joined #openstack-neutron | 06:29 | |
| *** coolsvap has joined #openstack-neutron | 06:33 | |
| *** yongli has quit IRC | 06:34 | |
| *** coolsvap has quit IRC | 06:36 | |
| *** yongli has joined #openstack-neutron | 06:36 | |
| openstackgerrit | Jenkins proposed a change to openstack/neutron: Imported Translations from Transifex https://review.openstack.org/58446 | 06:39 |
| *** amritanshu_RnD has joined #openstack-neutron | 06:43 | |
| *** mihgen has joined #openstack-neutron | 06:47 | |
| *** yongli has quit IRC | 06:48 | |
| *** yongli has joined #openstack-neutron | 06:48 | |
| *** yongli has quit IRC | 06:50 | |
| *** yongli has joined #openstack-neutron | 06:50 | |
| *** matsuhashi has quit IRC | 06:53 | |
| *** matsuhashi has joined #openstack-neutron | 06:54 | |
| *** nati_uen_ has quit IRC | 06:56 | |
| *** matsuhashi has quit IRC | 06:58 | |
| *** matsuhashi has joined #openstack-neutron | 06:59 | |
| *** gdubreui has quit IRC | 07:05 | |
| *** introom has joined #openstack-neutron | 07:06 | |
| introom | hi | 07:06 |
| *** mengxd has joined #openstack-neutron | 07:06 | |
| introom | Can I install multiple openvswitch in openstack and arrange the topology as I like? | 07:06 |
| *** yongli has quit IRC | 07:06 | |
| *** mengxd has quit IRC | 07:07 | |
| *** yongli has joined #openstack-neutron | 07:07 | |
| *** introom has quit IRC | 07:09 | |
| *** nati_ueno has joined #openstack-neutron | 07:12 | |
| *** afazekas has joined #openstack-neutron | 07:13 | |
| *** steven-weston_ has quit IRC | 07:16 | |
| *** steven-weston has joined #openstack-neutron | 07:23 | |
| *** matsuhashi has quit IRC | 07:26 | |
| *** matsuhashi has joined #openstack-neutron | 07:27 | |
| yfried | marun: ping | 07:34 |
| *** yongli has quit IRC | 07:35 | |
| *** yongli has joined #openstack-neutron | 07:35 | |
| *** alagalah has quit IRC | 07:37 | |
| *** matsuhas_ has joined #openstack-neutron | 07:39 | |
| *** matsuhashi has quit IRC | 07:42 | |
| *** SumitNaiksatam has joined #openstack-neutron | 07:43 | |
| *** x86brandon has joined #openstack-neutron | 07:49 | |
| *** jlibosva has joined #openstack-neutron | 07:59 | |
| openstackgerrit | A change was merged to openstack/neutron: Sync openstack.common.local from oslo https://review.openstack.org/57909 | 08:00 |
| sgran | obondarev: ping | 08:01 |
| *** amuller has joined #openstack-neutron | 08:01 | |
| *** matsuhas_ has quit IRC | 08:05 | |
| *** matsuhashi has joined #openstack-neutron | 08:05 | |
| sgran | and, morning all | 08:06 |
| steven-weston | morning, sgran | 08:09 |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Send only one agent notification on port update https://review.openstack.org/58415 | 08:11 |
| *** mihgen has quit IRC | 08:11 | |
| *** fouxm has joined #openstack-neutron | 08:15 | |
| *** steven-weston is now known as sw2 | 08:16 | |
| openstackgerrit | Marios Andreou proposed a change to openstack/neutron: Make flow operation parameters explicit https://review.openstack.org/58308 | 08:16 |
| *** steven-weston has joined #openstack-neutron | 08:17 | |
| openstackgerrit | berlin proposed a change to openstack/neutron: Fix showing nonexistent NetworkGateway throws 500 instead of 404 https://review.openstack.org/58423 | 08:19 |
| *** akamyshnikova has quit IRC | 08:20 | |
| Qlawy | Hi, when I try to get metadata from inside instance I recive 500 Internal Server Error | 08:26 |
| Qlawy | stacktrace http://wklej.org/hash/7e9cb2b5674/ | 08:26 |
| Qlawy | Its first time I have this issue | 08:26 |
| *** steven-weston has quit IRC | 08:28 | |
| *** x86brandon has quit IRC | 08:29 | |
| *** nati_ueno has quit IRC | 08:29 | |
| EmilienM | salv-orlando: ping | 08:30 |
| salv-orlando | EmilienM: in a meeting, will ping you back when I'm done | 08:31 |
| sgran | Qlawy: you seem to be using the nova extension os-virtual-interfaces, which neutron doesn't support | 08:34 |
| *** akamyshnikova has joined #openstack-neutron | 08:36 | |
| *** steven-weston has joined #openstack-neutron | 08:36 | |
| steven-weston | sw2: ping | 08:36 |
| Qlawy | sgran: yeah, just realized it ;) | 08:42 |
| Qlawy | sgran: but thx | 08:42 |
| *** steven-weston_ has joined #openstack-neutron | 08:45 | |
| *** yongli has quit IRC | 08:46 | |
| *** yongli has joined #openstack-neutron | 08:47 | |
| *** steven-weston_ has quit IRC | 08:50 | |
| *** steven-weston has quit IRC | 08:50 | |
| *** mihgen has joined #openstack-neutron | 08:55 | |
| *** mihgen has quit IRC | 09:00 | |
| *** mihgen has joined #openstack-neutron | 09:00 | |
| *** steven-weston has joined #openstack-neutron | 09:00 | |
| *** yonglihe_ has joined #openstack-neutron | 09:05 | |
| *** jistr has joined #openstack-neutron | 09:06 | |
| *** safchain has joined #openstack-neutron | 09:06 | |
| *** _coolsvap_ has joined #openstack-neutron | 09:06 | |
| *** yonglihe_ has quit IRC | 09:06 | |
| *** yonglihe_ has joined #openstack-neutron | 09:07 | |
| *** yongli has quit IRC | 09:07 | |
| *** yongli has joined #openstack-neutron | 09:07 | |
| *** nati_ueno has joined #openstack-neutron | 09:09 | |
| *** sw2 has quit IRC | 09:10 | |
| *** jpich has joined #openstack-neutron | 09:11 | |
| *** yongli has quit IRC | 09:13 | |
| *** yongli has joined #openstack-neutron | 09:13 | |
| openstackgerrit | Youcef Laribi proposed a change to openstack/neutron: Implements an LBaaS driver for NetScaler devices. https://review.openstack.org/57524 | 09:16 |
| *** richardboswell has quit IRC | 09:17 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Send only one agent notification on port update https://review.openstack.org/58415 | 09:24 |
| *** gdubreui has joined #openstack-neutron | 09:27 | |
| openstackgerrit | Zhang Hua proposed a change to openstack/neutron: Replace uuidutils.generate_uuid() with str(uuid.uuid4()) https://review.openstack.org/58234 | 09:29 |
| steven-weston | salv-orlando: are you available for a quick momento? :-) | 09:30 |
| *** yamahata_ has joined #openstack-neutron | 09:30 | |
| salv-orlando | steven-weston: I have just 5 minutes I am afraid | 09:32 |
| salv-orlando | shoot | 09:32 |
| steven-weston | salv-orlando: that's okay, I can wait .. I know you are busy ... just would like to know when we can work on the auto ip assignment bp, or of there is anything else I can help with instead | 09:33 |
| *** ljjjustin has quit IRC | 09:33 | |
| salv-orlando | steven-weston: yeah that slipped off my mind, you're right. I have a 3-hour appointment now. Will you be online around 2 PM GMT? | 09:35 |
| *** rpodolyaka has left #openstack-neutron | 09:35 | |
| steven-weston | salv-orlando: yes, let's talk more then? | 09:35 |
| salv-orlando | sounds good to me. | 09:37 |
| *** yfried has quit IRC | 09:37 | |
| salv-orlando | steven-weston: ping me because I'll probably forget about it. | 09:37 |
| *** _coolsvap_ has quit IRC | 09:37 | |
| steven-weston | salv-orlando: okay! have a great appointment :-) | 09:37 |
| *** nati_ueno has quit IRC | 09:40 | |
| openstackgerrit | Joe Mills proposed a change to openstack/neutron: Blackhole traffic not destined to existing port https://review.openstack.org/58474 | 09:45 |
| *** networkstatic has quit IRC | 09:49 | |
| sgran | can I have a pair of eyes on https://review.openstack.org/#/c/56815/ and https://review.openstack.org/#/c/40381/ ? | 09:57 |
| sgran | or are people holding off approving/merging at the moment? I saw something about gate issues | 09:57 |
| *** matsuhas_ has joined #openstack-neutron | 09:58 | |
| *** matsuhas_ has quit IRC | 09:58 | |
| *** matsuhashi has quit IRC | 10:02 | |
| steven-weston | sgran: how are you choosing the 192.168 address as the load balancer ip? should it be chosen from the range on the network you are testing the lb against? | 10:06 |
| sgran | where? In the test? | 10:06 |
| steven-weston | yes | 10:07 |
| sgran | self.member() defaults to 192.168.1.100 | 10:07 |
| sgran | so I'm just making it explicit to make it obvious | 10:07 |
| steven-weston | ok, gotcha ... i don't currently have a way to test it, I am rebuilding my dev gate, but it looks okay to me. | 10:08 |
| sgran | great | 10:08 |
| *** Jianyong has quit IRC | 10:12 | |
| enikanorov | steven-weston: what kind of gate are you reffering in the review? | 10:19 |
| steven-weston | enikanorov: the jenkins gate ... i thought it was down | 10:20 |
| enikanorov | jenkins +1ed the patch | 10:20 |
| steven-weston | oh, i didn't see that .. just a second | 10:20 |
| *** qingluo has quit IRC | 10:22 | |
| EmilienM | salv-orlando: it was just to let you know i started a work on grenade for Neutron | 10:24 |
| steven-weston | enikanorov: okay, better? | 10:26 |
| enikanorov | steven-weston: sure, if you have confidence in it :-) | 10:27 |
| steven-weston | enikanorov: yes, unless I shouldn't be ... if it works in devstack gate then it is good enough for jenkins? | 10:29 |
| *** yamahata_ has quit IRC | 10:36 | |
| *** jroovers has quit IRC | 10:54 | |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: Implement testing of migrations https://review.openstack.org/48594 | 11:02 |
| *** rossella_s has joined #openstack-neutron | 11:03 | |
| *** yfried has joined #openstack-neutron | 11:06 | |
| *** gdubreui has quit IRC | 11:10 | |
| *** roeyc has joined #openstack-neutron | 11:23 | |
| *** jorisroovers has joined #openstack-neutron | 11:26 | |
| *** jorisroovers has quit IRC | 11:27 | |
| *** jorisroovers has joined #openstack-neutron | 11:27 | |
| *** nissim_ has joined #openstack-neutron | 11:38 | |
| nissim_ | hi | 11:38 |
| nissim_ | can anyone assist with a small neutron/metadata proxy issue? | 11:38 |
| *** pcm_ has joined #openstack-neutron | 11:41 | |
| decede | nissim_: whats the issue? | 11:44 |
| *** jroovers has joined #openstack-neutron | 11:44 | |
| *** francois_eleouet has joined #openstack-neutron | 11:44 | |
| nissim_ | I am testing openstack grizzly and have 2 private networks connected to two virtual routers via vlans, thing is I can't connect to http://169.254.169.254 to get instance metadata | 11:45 |
| *** jorisroovers has quit IRC | 11:46 | |
| *** feleouet has quit IRC | 11:46 | |
| nissim_ | my setup include: controller-node, network-node, mysql-node (includes keystone) & nova-node | 11:46 |
| nissim_ | I am trying to access the metadata on ubuntu 12.0.4-3 instance not cirros which I know is not working | 11:47 |
| nissim_ | decede? | 11:48 |
| decede | nissim_: for support join #openstack you'll get a better response | 11:54 |
| nissim_ | tried that, looks like there is no traffic there , but thanks. | 11:55 |
| *** nissim_ has left #openstack-neutron | 11:55 | |
| *** pcm_ has quit IRC | 11:55 | |
| *** pcm_ has joined #openstack-neutron | 11:56 | |
| roeyc | Hi there, can anyone help to clarify some points regarding external third party testing of vendor specific plugin ? | 12:07 |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: Sync models with migrations https://review.openstack.org/55411 | 12:09 |
| *** yongli has quit IRC | 12:17 | |
| *** mihgen has quit IRC | 12:26 | |
| EmilienM | salv-orlando: ping | 12:28 |
| *** mihgen has joined #openstack-neutron | 12:33 | |
| *** amrit_ has joined #openstack-neutron | 12:41 | |
| *** amritanshu_RnD has quit IRC | 12:43 | |
| *** amrit_ has quit IRC | 12:43 | |
| *** amritanshu_RnD has joined #openstack-neutron | 12:43 | |
| *** yamahata_ has joined #openstack-neutron | 12:47 | |
| *** jasonv has joined #openstack-neutron | 12:51 | |
| *** jasonv has quit IRC | 12:52 | |
| *** dims has joined #openstack-neutron | 12:55 | |
| *** markmcclain has joined #openstack-neutron | 12:56 | |
| *** arosen has joined #openstack-neutron | 13:00 | |
| *** ygbo has joined #openstack-neutron | 13:12 | |
| *** insanidade has joined #openstack-neutron | 13:13 | |
| insanidade | Hi all. How do I find out my neutron client's version? I'm trying to figure out the reason why 'neutron agent-list' is trying to point to a wrong endpoint (it appends '/v2' to the endpoint registered in keystone) | 13:14 |
| insanidade | any help ? | 13:14 |
| *** dhellmann-afk is now known as dhellmann | 13:15 | |
| salv-orlando | neutron --version | 13:22 |
| salv-orlando | EmilienM: I'm all yours... | 13:22 |
| salv-orlando | …until wife calls for lunch | 13:22 |
| EmilienM | salv-orlando: lol | 13:22 |
| EmilienM | salv-orlando: i started something yesterday night on grenade | 13:22 |
| salv-orlando | EmilienM: very good | 13:23 |
| salv-orlando | do you have a rough idea of the things we need to do. | 13:23 |
| EmilienM | salv-orlando: a WIP, i'll finish it tonight i think. Do you think i have to delete draft stuff and publish it ? | 13:23 |
| EmilienM | salv-orlando: is it a question ? | 13:23 |
| salv-orlando | even if WIP, yes it would be good if you can publishit | 13:23 |
| salv-orlando | yes, but you already answered it I guess | 13:23 |
| EmilienM | salv-orlando: yeah i understand grenade, i'll make it happen for this week | 13:24 |
| salv-orlando | "publishit" is an interesting whitespace failure | 13:24 |
| salv-orlando | EmilienM: awesome | 13:24 |
| EmilienM | published | 13:24 |
| EmilienM | salv-orlando: thanks | 13:24 |
| salv-orlando | cool, I'll check it out later | 13:24 |
| enikanorov_ | insanidade: | 13:27 |
| insanidade | salv-orlando: yes, neutron --version reports the client's version or server version? | 13:27 |
| insanidade | hi, enikanorov_ | 13:27 |
| enikanorov_ | insanidade: afaik, current client only supports 2.0 | 13:27 |
| sgran | roeyc: what do you want to know? | 13:28 |
| salv-orlando | client version | 13:28 |
| sgran | I don't know that I have the answers, but I'm learning about it too, so maybe we can try together :) | 13:28 |
| insanidade | enikanorov_: right. but I don't understand why that 'neutron agent-list' appends a '/v2' to the endpoint registered in keystone | 13:28 |
| salv-orlando | does it append the same /v2 to net-list and all the other cmmands, or just agent-list? | 13:29 |
| insanidade | enikanorov_: actually, it adds '/v2/v2.0' to the endpoint url I see in keystone endpoint-list | 13:29 |
| enikanorov_ | what is in the endpoint list? | 13:29 |
| insanidade | salv-orlando: it appends to net-list as well. | 13:30 |
| enikanorov_ | i guess neutron should not be registered in the keystone with api version in url | 13:30 |
| enikanorov_ | unlike some other OS projects | 13:31 |
| insanidade | enikanorov_: I have http://192.168.100.45:9696/ for all three endpoint values | 13:31 |
| *** arosen has quit IRC | 13:31 | |
| salv-orlando | can you post the whole output from keystone endpoint-list | 13:31 |
| enikanorov_ | and as a result client sends http://192.168.100.45:9696/v2/v2.0/networks.json? | 13:31 |
| salv-orlando | and then the output from neutron -v net-list? | 13:31 |
| sgran | or from neutron --debug net-list | 13:31 |
| *** dhellmann is now known as dhellmann-afk | 13:31 | |
| salv-orlando | use paste.openstack.org if you want | 13:31 |
| sgran | er, yes, -v | 13:32 |
| *** gongysh has joined #openstack-neutron | 13:32 | |
| insanidade | salv-orlando, enikanorov_ : sure. just a sec. | 13:33 |
| *** bashok has quit IRC | 13:34 | |
| insanidade | salv-orlando, enikanorov_ : please, take a look: http://paste.openstack.org/show/53984/ | 13:35 |
| roeyc | sgran: Would like to know if all tests under tempest should be run and against which patchsets ( Neutrons only )? | 13:36 |
| *** amotoki has joined #openstack-neutron | 13:37 | |
| salv-orlando | insanidade: have you always been hitting a 300 error? | 13:38 |
| salv-orlando | out for lunch, back in 20 minutes | 13:38 |
| insanidade | salv-orlando: yes. every neutron command I issue returns that 300 error. | 13:39 |
| enikanorov_ | can you post neutron.conf? | 13:40 |
| insanidade | enikanorov_: one sec. copying it. | 13:42 |
| *** alagalah has joined #openstack-neutron | 13:44 | |
| insanidade | enikanorov_: neutron.conf: http://paste.openstack.org/show/53985/ | 13:45 |
| *** arosen has joined #openstack-neutron | 13:47 | |
| *** markvoelker1 has joined #openstack-neutron | 13:50 | |
| *** markmcclain has quit IRC | 13:50 | |
| *** gongysh has quit IRC | 13:57 | |
| insanidade | enikanorov_: did you have the time to check it ? | 13:59 |
| enikanorov_ | looking | 14:00 |
| *** yfried has quit IRC | 14:01 | |
| sgran | roeyc: that's my understanding, yeah | 14:09 |
| steven-weston | salv-orlando: may I steal you back from your wife now? :-) https://etherpad.openstack.org/p/NeutronFloatingIPAutoAssociation | 14:11 |
| *** dhellmann-afk is now known as dhellmann | 14:11 | |
| *** julim has joined #openstack-neutron | 14:12 | |
| *** salv-orlando has quit IRC | 14:15 | |
| roeyc | sgran: Allright. Thanks | 14:17 |
| *** arosen has quit IRC | 14:17 | |
| *** salv-orlando has joined #openstack-neutron | 14:21 | |
| salv-orlando | steven-weston: ping | 14:21 |
| steven-weston | salv-orlando: I'm here ;-) | 14:22 |
| *** clev has joined #openstack-neutron | 14:22 | |
| steven-weston | salv-orlando: ping | 14:23 |
| steven-weston | I think ... | 14:24 |
| salv-orlando | steven-weston: hi. Can you refresh me on the workflow when one uses nova-network? | 14:24 |
| salv-orlando | I mean auto-assign-floatingIP with nova-network | 14:24 |
| steven-weston | I think when external network connectivity is needed when the vm boots | 14:26 |
| ygbo | Hi all, is it normal that a non-admin user can not see the internet IP of his router (neutron router-port-list <router_id>)? Because it is a bit of an issue for VPNaaS where one needs to know his external IP to configure it. | 14:26 |
| insanidade | enikanorov_: I keep receiving a 300 error message stating that there are 'multiple options' (to be returned, probably). Do I have to explicitly tell the command the version I want to use ? | 14:28 |
| salv-orlando | ygbo: that's how it works at the moment; it can be changed, but is not trivial. If you have ideas perhaps share them on the mailing list | 14:29 |
| openstackgerrit | Aleksandr Chirko proposed a change to openstack/neutron: Bugfix and refactoring for ovs_lib.OVSBridge flow managment methods https://review.openstack.org/58533 | 14:29 |
| sgran | hurrah, the gate is unwedged, they say | 14:29 |
| salv-orlando | steven-weston: when nova boots a vm to be associated with a floating IP, does it make separate calls to nova network, or a single call to allocate both the vifs and the floating ip | 14:30 |
| sgran | can I ask for an approval on https://review.openstack.org/#/c/56815/ and https://review.openstack.org/#/c/40381/ ? | 14:30 |
| sgran | pretty please? :) | 14:30 |
| salv-orlando | sgran: were they already approved by core devs and then the patch did not pass the gate? | 14:31 |
| ygbo | salv-orlando: Thanks, is this a keystone limitation? or from the neutron side? | 14:31 |
| salv-orlando | ygbo: neutron authZ model | 14:31 |
| sgran | salv-orlando: they're waiting for approval | 14:31 |
| steven-weston | salv-orlando: i don't know, let me check ... | 14:31 |
| sgran | I was assuming approvals were being withheld because the gate was wedged | 14:31 |
| salv-orlando | sgran: so they were already +2'ed from 2 cores, I guess | 14:31 |
| ygbo | salv-orlando: ok, thanks. | 14:31 |
| sgran | 40381 has a +2, 56815 still needs one | 14:32 |
| salv-orlando | sgran: ok. I will review then | 14:32 |
| sgran | \o/ | 14:32 |
| sgran | thanks :) | 14:32 |
| salv-orlando | might take a while 40381 is huge | 14:32 |
| sgran | yes, yes it is. It's not my change, but I'm kind of trying to champion it because I think it's a good change, and I have some work that will need refactoring once it gets merged | 14:33 |
| sgran | 56815 should be fairly trivial, though | 14:33 |
| openstackgerrit | A change was merged to openstack/neutron: Imported Translations from Transifex https://review.openstack.org/58446 | 14:38 |
| *** clev has quit IRC | 14:38 | |
| *** aveiga has joined #openstack-neutron | 14:39 | |
| *** dyerm has joined #openstack-neutron | 14:39 | |
| *** insanida1e has joined #openstack-neutron | 14:43 | |
| *** banix has joined #openstack-neutron | 14:43 | |
| steven-weston | salv-orlando: looks like it is currently done as separate calls | 14:45 |
| *** insanidade has quit IRC | 14:46 | |
| salv-orlando | steven-weston: I was asking because when interfacing with neutron we should keep the same interface | 14:46 |
| insanida1e | salv-orlando, enikanorov_ : I think I found the problem. | 14:46 |
| salv-orlando | nova-api would call into nova.network api which instead of calling nova-network service will call neutron. does that make sense? | 14:47 |
| insanida1e | salv-orlando, enikanorov_ : pointing to nova-network instead of neutron (copy & paste error) | 14:47 |
| salv-orlando | insanida1e: I finally got your nick right. What was it? | 14:47 |
| salv-orlando | ok, I see | 14:47 |
| steven-weston | salv-orlando: ok. perhaps I am being a bit too optimistic in the blueprint, it should simply be considered from a nova parity perspective? | 14:47 |
| insanida1e | salv-orlando, enikanorov_ : thanks :) learning a lot trying to isntall havana from the ground up. | 14:47 |
| *** insanida1e is now known as insanidade | 14:48 | |
| steven-weston | salv-orlando: yes, that makes sense | 14:48 |
| salv-orlando | steven-weston: not really, but is a requirement that nova should be able to auto-assign the floating IP with neutron as well | 14:48 |
| salv-orlando | And I was try to understand if we can have a solution that minimises the amount of change on the nova side. | 14:49 |
| *** thedodd has joined #openstack-neutron | 14:50 | |
| steven-weston | salv-orlando: it all depends on how you want it done. I think it would be possible that there would be no change required from the nova side at all, besides changing the ip manager from the conf file, that is ... | 14:51 |
| *** clev has joined #openstack-neutron | 14:53 | |
| salv-orlando | steven-weston: on the neutron side how would the API change to allow for auto associating floating IPs? | 14:56 |
| *** jroovers has quit IRC | 14:56 | |
| *** clev has quit IRC | 14:57 | |
| *** peristeri has joined #openstack-neutron | 14:58 | |
| *** clev has joined #openstack-neutron | 15:00 | |
| steven-weston | salv-orlando: it may not have to change at all. i'm looking at allocate_for_instance in nova/network/floating_ips.py and since the calls are currently separate I can look into getting it working by simply changing the floating ip manager in the nova.conf file | 15:01 |
| steven-weston | salv-orlando: it may work already, i haven't tried it yet ;-) | 15:01 |
| *** armax has joined #openstack-neutron | 15:01 | |
| salv-orlando | steven-weston: let's start from this and then we'll look at how to optimise with api changes | 15:02 |
| *** jecarey has joined #openstack-neutron | 15:04 | |
| steven-weston | salv-orlando: sounds like a plan! Any changes that should be made to the bp or the etherpad right now, or should we wait until I can answer the questions you asked today? | 15:04 |
| *** fouxm_ has joined #openstack-neutron | 15:08 | |
| *** fouxm_ has quit IRC | 15:09 | |
| *** fouxm has quit IRC | 15:09 | |
| *** fouxm has joined #openstack-neutron | 15:09 | |
| *** fouxm_ has joined #openstack-neutron | 15:11 | |
| *** fouxm_ has quit IRC | 15:12 | |
| *** fouxm_ has joined #openstack-neutron | 15:13 | |
| *** roeyc has quit IRC | 15:14 | |
| *** fouxm has quit IRC | 15:15 | |
| salv-orlando | I think we can wait | 15:15 |
| sgran | salv-orlando: both good nits | 15:16 |
| sgran | I'll reupload | 15:16 |
| *** amuller has quit IRC | 15:17 | |
| steven-weston | salv-orlando: ok, great! do you need anything else right now? I can get back to you before the end of your day tomorrrow with my answers, at the very latest. | 15:18 |
| salv-orlando | steven-weston: thanks for everything your plan sounds good. | 15:20 |
| steven-weston | salv-orlando: absolutely! what time should I meet you back here tomorrow? | 15:21 |
| salv-orlando | I am online usually from 8AM-1PM GMT and 5PM-11PM GMT; This way I can work with Europe, Asia, and US | 15:22 |
| steven-weston | salv-orlando: very good. thanks for your direction today!! | 15:24 |
| *** carl_baldwin has joined #openstack-neutron | 15:24 | |
| *** amritanshu_RnD has quit IRC | 15:32 | |
| *** alexpilotti has joined #openstack-neutron | 15:38 | |
| *** dhellmann is now known as dhellmann-afk | 15:41 | |
| *** peristeri has quit IRC | 15:43 | |
| *** briancline has joined #openstack-neutron | 15:50 | |
| *** carl_baldwin has quit IRC | 15:57 | |
| *** yfried has joined #openstack-neutron | 15:57 | |
| *** dhellmann-afk is now known as dhellmann | 15:58 | |
| *** carl_baldwin has joined #openstack-neutron | 16:01 | |
| *** afazekas has quit IRC | 16:02 | |
| *** armax has left #openstack-neutron | 16:02 | |
| *** armax has joined #openstack-neutron | 16:02 | |
| *** SumitNaiksatam has quit IRC | 16:03 | |
| openstackgerrit | Dane LeBlanc proposed a change to openstack/neutron: Improve unit test coverage for Cisco plugin model code https://review.openstack.org/58125 | 16:08 |
| ygbo | salv-orlando: I actually logged https://bugs.launchpad.net/neutron/+bug/1255142 | 16:09 |
| *** yamahata_ has quit IRC | 16:10 | |
| *** dhellmann is now known as dhellmann-afk | 16:17 | |
| *** idella4 has quit IRC | 16:17 | |
| *** mihgen has quit IRC | 16:22 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Rebind allowed address pairs only if they changed https://review.openstack.org/58566 | 16:26 |
| openstackgerrit | Stephen Gran proposed a change to openstack/neutron: Enforce unique constraint on neutron pool members https://review.openstack.org/56815 | 16:26 |
| *** networkstatic has joined #openstack-neutron | 16:28 | |
| marun | salv-orlando: project_name vs tenant_name? | 16:29 |
| salv-orlando | do you want me to make a random choice? | 16:30 |
| marun | salv-orlando: I'm wondering wtf, actually, hoping you can fill me in | 16:31 |
| marun | salv-orlando: I mean, Keystone v3 has switched from tenant to project | 16:32 |
| marun | salv-orlando: I think nova must have always been project | 16:32 |
| salv-orlando | Yes, we were warned about this. Somewhere before the Folsom release. | 16:32 |
| marun | salv-orlando: but where does that leave neutron? do we have a preference? | 16:32 |
| salv-orlando | At the time we chose to go with whatever keystone decided. | 16:32 |
| marun | salv-orlando: so if that's the case, is there any reason to continue using 'tenant' instead of 'project'? | 16:33 |
| marun | salv-orlando: i.e. can a search-replace patch solve this issue once and for all? | 16:33 |
| salv-orlando | I think that might be good | 16:33 |
| marun | salv-orlando: ok. I'll file a bug and see if I can get buy-in on the mailing list. | 16:34 |
| salv-orlando | marun: enikanorov has a patch that puts both tenant_name and project_name in context | 16:34 |
| marun | salv-orlando: I was working on a similar patch:https://review.openstack.org/#/c/58348/ | 16:34 |
| salv-orlando | ah I see | 16:34 |
| *** SumitNaiksatam has joined #openstack-neutron | 16:35 | |
| marun | salv-orlando: oleg's question of whether tenant_name was needed prompted me to pursue this line of questioning | 16:35 |
| marun | salv-orlando: do you have a link to his patch? | 16:35 |
| *** ywu has joined #openstack-neutron | 16:42 | |
| *** beagles has quit IRC | 16:45 | |
| openstackgerrit | Ralf Haferkamp proposed a change to openstack/neutron: Reassign IP to vlan interface when deleting a VLAN bridge https://review.openstack.org/58575 | 16:46 |
| *** beagles has joined #openstack-neutron | 16:47 | |
| openstackgerrit | stephen-ma proposed a change to openstack/neutron: L3 Agent restart causes network outage https://review.openstack.org/30988 | 16:52 |
| *** rkukura has left #openstack-neutron | 16:53 | |
| openstackgerrit | Maru Newby proposed a change to openstack/neutron: Add missing key to context dict to fix rpc logging https://review.openstack.org/58348 | 16:56 |
| morganfainberg | marun, salv-orlando, (let me put on my Keystone hat), moving to "project" from "tenant" seems to make the most sense based upon the thread on the ML. As a keystone person, I'd like to see it all unified under project. | 16:57 |
| marun | morganfainberg: https://blueprints.launchpad.net/neutron/+spec/rename-tenant-to-project | 16:58 |
| morganfainberg | marun, yay! | 16:58 |
| marun | morganfainberg: I need to send an email to the ml to get everyone on board. | 16:58 |
| morganfainberg | marun, totally understand. i'll keep my eyes open and jump in if needed to champion "project" ;) | 16:59 |
| marun | morganfainberg: I don't think there will be resistance, I'm just not sure of the priority it will be given early in the cycle where testing is our biggest issue. | 16:59 |
| morganfainberg | more important than personal bias thouhg, it'll be nice once all projects reference the "project"/"tenant" grouping as the same thing (UX wise) | 17:00 |
| *** pete5 has joined #openstack-neutron | 17:04 | |
| *** rkukura has joined #openstack-neutron | 17:06 | |
| marun | morganfainberg: agreed, consistency is a very nice property | 17:07 |
| *** markmcclain has joined #openstack-neutron | 17:09 | |
| enikanorov | salv-orlando: i think that it's marun who have a patch, not me :) | 17:12 |
| *** carl_baldwin has quit IRC | 17:18 | |
| salv-orlando | ok, I'm confused then | 17:19 |
| *** clev has quit IRC | 17:19 | |
| marun | salv-orlando: feel free to review ;) https://review.openstack.org/#/c/58348/ | 17:20 |
| openstackgerrit | Armando Migliaccio proposed a change to openstack/neutron: Handle IPAddressGenerationFailure on create_dhcp_port https://review.openstack.org/57812 | 17:21 |
| salv-orlando | marun: I +2 but there is a comment inline, which is a draft I had from yesterday. I think what we said today clears that comment | 17:23 |
| insanidade | is there any guide on giving a machine a floating ip after it is provisioned through heat ? | 17:23 |
| insanidade | machine= vm | 17:23 |
| marun | insanidade: not sure many here are too familiar with heat | 17:25 |
| marun | insanidade: may want to ask on #openstack | 17:25 |
| sgran | insanidade: heat simply provisions an instance | 17:26 |
| sgran | afterwards, you can do the same thing you would normally do without heat | 17:26 |
| insanidade | sgran, marun: yes. I believe I might be missing something for setting that floating ip to that vm - I understand the "heat" factor should not be important for that matter. | 17:26 |
| *** ygbo has quit IRC | 17:27 | |
| *** reaper has joined #openstack-neutron | 17:27 | |
| *** clev has joined #openstack-neutron | 17:28 | |
| *** safchain has quit IRC | 17:31 | |
| *** clev has quit IRC | 17:32 | |
| *** networkstatic has quit IRC | 17:34 | |
| *** clev has joined #openstack-neutron | 17:35 | |
| *** markmcclain has quit IRC | 17:39 | |
| *** jistr has quit IRC | 17:39 | |
| *** rossella_s has quit IRC | 17:49 | |
| *** clev has quit IRC | 17:56 | |
| *** jlibosva has quit IRC | 17:58 | |
| *** jlibosva has joined #openstack-neutron | 17:58 | |
| *** nati_ueno has joined #openstack-neutron | 17:58 | |
| *** nati_ueno has quit IRC | 17:59 | |
| *** nati_ueno has joined #openstack-neutron | 18:00 | |
| *** otherwiseguy has joined #openstack-neutron | 18:00 | |
| *** suresh12 has joined #openstack-neutron | 18:01 | |
| *** alagalah has quit IRC | 18:03 | |
| *** alagalah has joined #openstack-neutron | 18:03 | |
| *** alagalah_ has joined #openstack-neutron | 18:04 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Rebind security groups only when they're updated https://review.openstack.org/58597 | 18:11 |
| marun | https://bugs.launchpad.net/neutron/+bug/1251448 | 18:16 |
| openstackgerrit | Nachi Ueno proposed a change to openstack/neutron: Improve vif attributes related with firewalling https://review.openstack.org/21946 | 18:23 |
| *** alagalah_ has quit IRC | 18:23 | |
| *** alagalah_ has joined #openstack-neutron | 18:24 | |
| *** clev has joined #openstack-neutron | 18:26 | |
| *** networkstatic has joined #openstack-neutron | 18:35 | |
| *** jprovazn has joined #openstack-neutron | 18:38 | |
| *** jlibosva has quit IRC | 18:41 | |
| *** fouxm_ has quit IRC | 18:45 | |
| *** clev has quit IRC | 18:52 | |
| *** jistr has joined #openstack-neutron | 18:57 | |
| *** jlibosva has joined #openstack-neutron | 18:58 | |
| *** suresh12 has quit IRC | 19:00 | |
| *** mlavalle has joined #openstack-neutron | 19:15 | |
| *** aymenfrikha has joined #openstack-neutron | 19:16 | |
| *** nati_uen_ has joined #openstack-neutron | 19:26 | |
| *** nati_ueno has quit IRC | 19:28 | |
| *** nati_uen_ has quit IRC | 19:30 | |
| *** nati_ueno has joined #openstack-neutron | 19:30 | |
| *** suresh12 has joined #openstack-neutron | 19:30 | |
| openstackgerrit | Nachi Ueno proposed a change to openstack/neutron: VPNaaS integration with service type framework https://review.openstack.org/41827 | 19:31 |
| *** opilotte has joined #openstack-neutron | 19:31 | |
| opilotte | can someone help me with this patch-set? can't figure out what is wrong: https://review.openstack.org/#/c/54170/ | 19:31 |
| opilotte | used to pass every jenkins check | 19:32 |
| opilotte | now it's a mess | 19:32 |
| *** suresh12 has quit IRC | 19:36 | |
| *** suresh12 has joined #openstack-neutron | 19:37 | |
| openstackgerrit | Armando Migliaccio proposed a change to openstack/neutron: Handle IPAddressGenerationFailure on create_dhcp_port https://review.openstack.org/57812 | 19:37 |
| *** yfried has quit IRC | 19:42 | |
| *** jecarey has quit IRC | 19:44 | |
| openstackgerrit | Armando Migliaccio proposed a change to openstack/neutron: Handle IPAddressGenerationFailure on create_dhcp_port https://review.openstack.org/57812 | 19:48 |
| sgran | I think I've now addressed all the comments on https://review.openstack.org/#/c/56815/ , if someone has a moment to review it | 19:51 |
| *** alagalah has quit IRC | 19:55 | |
| *** alagalah_ has quit IRC | 19:55 | |
| *** alagalah_ has joined #openstack-neutron | 19:55 | |
| *** alagalah has joined #openstack-neutron | 19:56 | |
| *** gdubreui has joined #openstack-neutron | 20:01 | |
| *** jlibosva has quit IRC | 20:11 | |
| nati_ueno | mestery: Hi! | 20:14 |
| mestery | nati_ueno: on phone give me 15 | 20:15 |
| nati_ueno | mestery: sure! Thanks | 20:15 |
| *** julim has quit IRC | 20:17 | |
| *** suresh12 has quit IRC | 20:18 | |
| *** nati_uen_ has joined #openstack-neutron | 20:18 | |
| *** nati_ueno has quit IRC | 20:21 | |
| openstackgerrit | Nachi Ueno proposed a change to openstack/neutron: Improve vif attributes related with firewalling https://review.openstack.org/21946 | 20:26 |
| openstackgerrit | Maru Newby proposed a change to openstack/neutron: Fix format errors seen in rpc logging https://review.openstack.org/58348 | 20:30 |
| *** jecarey has joined #openstack-neutron | 20:33 | |
| *** dhellmann-afk is now known as dhellmann | 20:44 | |
| *** jistr has quit IRC | 20:49 | |
| mestery | nati_uen_: I am back, whats up? | 20:54 |
| mestery | marun: Apologies for approving your WIP patch earlier. :( | 20:58 |
| marun | mestery: no worries. i didn't realize that was possible or i would have -1d | 20:58 |
| mestery | marun: Me either! I guess I'm still learning something new everyday. ;) | 20:59 |
| *** alagalah has quit IRC | 20:59 | |
| *** alagalah_ has quit IRC | 20:59 | |
| openstackgerrit | Maru Newby proposed a change to openstack/neutron: Fix format errors seen in rpc logging https://review.openstack.org/58348 | 21:01 |
| *** alagalah has joined #openstack-neutron | 21:01 | |
| *** x86brandon has joined #openstack-neutron | 21:02 | |
| *** prometheanfire has quit IRC | 21:05 | |
| *** prometheanfire has joined #openstack-neutron | 21:05 | |
| *** banix has quit IRC | 21:09 | |
| *** clev has joined #openstack-neutron | 21:13 | |
| openstackgerrit | Akihiro Motoki proposed a change to openstack/neutron: Return request-id in API response https://review.openstack.org/58270 | 21:13 |
| *** suresh12 has joined #openstack-neutron | 21:14 | |
| *** mlavalle has quit IRC | 21:24 | |
| *** safchain has joined #openstack-neutron | 21:25 | |
| *** x86brandon has quit IRC | 21:25 | |
| openstackgerrit | Sylvain Afchain proposed a change to openstack/neutron: Add LeastRouters Scheduler to Neutron L3 Agent https://review.openstack.org/47490 | 21:26 |
| *** x86brandon has joined #openstack-neutron | 21:26 | |
| *** steven-weston has quit IRC | 21:26 | |
| *** julim has joined #openstack-neutron | 21:26 | |
| *** banix has joined #openstack-neutron | 21:31 | |
| *** devlaps has joined #openstack-neutron | 21:35 | |
| *** yamahata_ has joined #openstack-neutron | 21:36 | |
| *** banix has quit IRC | 21:37 | |
| *** dims has quit IRC | 21:40 | |
| *** mlavalle has joined #openstack-neutron | 21:40 | |
| *** networkstatic has quit IRC | 21:41 | |
| *** jprovazn has quit IRC | 21:42 | |
| *** suresh12 has quit IRC | 21:43 | |
| *** suresh12 has joined #openstack-neutron | 21:44 | |
| *** suresh12 has quit IRC | 21:46 | |
| *** alexpilotti has quit IRC | 21:52 | |
| *** dims has joined #openstack-neutron | 21:54 | |
| *** yamahata_ has quit IRC | 22:00 | |
| marun | amotoki: ping | 22:00 |
| amotoki | marun: pong | 22:00 |
| *** markvoelker1 has quit IRC | 22:01 | |
| marun | amotoki: ah, nmind. i just saw your review | 22:01 |
| *** gdubreui has quit IRC | 22:05 | |
| nati_uen_ | mestery: Hi | 22:05 |
| mestery | nati_eun_: Yo, whats up? | 22:06 |
| nati_uen_ | mestery: why ML2 stores port_filter information on DB? | 22:06 |
| mestery | nati_uen_: Checking ... | 22:06 |
| nati_uen_ | cap_port_filter isn't used anywhere, but it is stored in the DB model | 22:07 |
| mestery | Well, I think the idea was to store all the port information needed for port binding. | 22:07 |
| mestery | Which included cap_port_filter. But it's not used anymore? Checking that now too. | 22:08 |
| nati_uen_ | yes. That's attribute isn't used anywhere for functionalities | 22:08 |
| *** aymenfrikha has quit IRC | 22:08 | |
| mestery | OK, so perhaps it should be removed then. :) | 22:09 |
| mestery | Maybe it was previously used when rkukura did the port binding work? | 22:09 |
| nati_uen_ | mestery: yeah, that was intented to be used, but the spec itself changed | 22:09 |
| nati_uen_ | so my question is is it really needed store all information in DB even if it is static? | 22:10 |
| mestery | nati_uen_: OK | 22:10 |
| rkukura | mestery, nati_uen_: what's the question? | 22:10 |
| mestery | nati_uen_: I think the idea was to store information necessary for port binding. | 22:10 |
| nati_uen_ | rkukura I'm working on this one https://review.openstack.org/#/c/21946/ | 22:11 |
| mestery | rkukura: nati_uen_ is asking specifically about why we store cap_port_filter in the DB | 22:11 |
| nati_uen_ | I'm going to remove cap_port_filter and add vif_security | 22:11 |
| nati_uen_ | it looks ML2 stores cap_port_filter | 22:11 |
| rkukura | The DB is actually storing the result of port binding - that's one of the things that the MechanismDriver supplies when it calls PortContext.set_binding() | 22:11 |
| nati_uen_ | so my question is why we are storing immutable information in db such as cap_port_filter? | 22:12 |
| rkukura | looking at the patch to driver_api.py, I think we'd need to be able to set whatever is replacing cap_port_filter, and store that instead | 22:12 |
| nati_uen_ | so I should store vif_security in the DB? | 22:13 |
| rkukura | nati_uen_: What makes it immutable? | 22:13 |
| nati_uen_ | vif_security value is defined per plugin | 22:13 |
| nati_uen_ | so it will not be different per port | 22:13 |
| rkukura | Couldn't it at least very depending on what MechanismDriver does the binding, and maybe based on information it has about the node? | 22:13 |
| *** yamahata_ has joined #openstack-neutron | 22:14 | |
| rkukura | nati_uen_: I don't think that makes sense with ML2 | 22:14 |
| nati_uen_ | we can let MechanismDriver define vif_security value | 22:14 |
| nati_uen_ | but even if so, it is static per MechanismDriver | 22:14 |
| rkukura | Maybe not different per port, but could be different for different drivers in the same deployment (think of a mix of KVM and HyperV nodes) | 22:14 |
| nati_uen_ | rkukura: That's makes sence | 22:15 |
| rkukura | So we could have ML2 call into the bound MechanismDriver, but are you sure it will always be static? | 22:15 |
| nati_uen_ | so IMO, driver -> vif security value mappings is static | 22:15 |
| nati_uen_ | at least for current specs | 22:15 |
| nati_uen_ | so if we know vif_type | 22:16 |
| nati_uen_ | vif_security_value will be defined | 22:16 |
| rkukura | Has anyone worked out how things like PCI-passthru fit into this? | 22:16 |
| nati_uen_ | rkukura: I think no one is thinking about it | 22:17 |
| rkukura | That's another case where a MechanismDriver would need to include information related to the binding that might vary from port to port | 22:17 |
| nati_uen_ | ah so in that mode, we wanna let configure different value for vif_security? | 22:18 |
| nati_uen_ | How we know the port is in the PC-passthru mode? | 22:19 |
| rkukura | Not sure about vif_security, but I think it hints that a more general mechanism may be needed for feed information from the ML2 MechanismDriver that binds to the GenricVIFDriver in nova | 22:19 |
| *** banix has joined #openstack-neutron | 22:20 | |
| mestery | rkukura: I think that makes sense to me. I agree that even with PCI passthrough vif_security may be static though. | 22:20 |
| rkukura | It "may" be static, but do we know it will always be static? | 22:21 |
| nati_uen_ | rkukura: Everything should be static until we get usecase in which it is dynamic | 22:21 |
| nati_uen_ | In current spec, it is static | 22:22 |
| *** gdubreui has joined #openstack-neutron | 22:22 | |
| nati_uen_ | I agree in future, we can have a bp for let admin or user configure the vif_security value per port | 22:22 |
| mestery | I think per-MechanismDriver it will be static for now, unless someone wants to change it. | 22:22 |
| rkukura | Is there a spec I should look at? | 22:22 |
| nati_uen_ | https://bugs.launchpad.net/nova/+bug/1112912 | 22:23 |
| nati_uen_ | is a spec | 22:23 |
| rkukura | I'm OK with per-MechanismDriver, or calling into the bound MechanismDriver each tme its needed rather than caching it in the DB | 22:23 |
| mestery | rkukura: The second option would make it dynamic from ML2 perspective, but each MD could choose to make it static at that level. | 22:23 |
| rkukura | right | 22:24 |
| mestery | nati_uen_: Thoughts on that approach? | 22:24 |
| nati_uen_ | It makes sence | 22:24 |
| nati_uen_ | so in this patch, I'll let each MD to decide vif_security value | 22:24 |
| nati_uen_ | but I'm not going to store it in the db | 22:24 |
| * mestery nods in agreement. | 22:24 | |
| nati_uen_ | rkukura: is this makes sence? | 22:25 |
| rkukura | So should this call from the plugin into the MechanismDriver be specific to vif_security, or should it allow the MechanismDriver to add arbitrary attributes to the port dictionary being built? | 22:25 |
| nati_uen_ | I wanna work on specific to vif_security in this patch. | 22:26 |
| rkukura | nati_uen_, mestery: I'm OK with not storing it in the DB, but we do need to make sure there are sensible values (None?) to use when there is no bound MechanismDriver to call | 22:27 |
| *** SumitNaiksatam has quit IRC | 22:28 | |
| rkukura | nati_uen_: I understand keeping this patch focused on vif_security, but also want to balance this against churn in the ML2 driver API | 22:28 |
| *** SumitNaiksatam_ has joined #openstack-neutron | 22:28 | |
| nati_uen_ | rkukura: ok sure. it is also makes sence | 22:28 |
| nati_uen_ | I'm not professional for ML2 now, so if you have a preferred design for this, I would like to follow it | 22:29 |
| nati_uen_ | or may be we can leave simple impl in this patch, then we can have another patch for refactoring the design on ML2 | 22:29 |
| mestery | nati_uen_: I think simple in your patch is good and we can refactor it later. | 22:30 |
| nati_uen_ | rkukura: is this OK for you too? | 22:31 |
| rkukura | nati_uen_: Seems _update_port_dict_binding() is going to need to call some new function on MechansmDriver, right? | 22:31 |
| rkukura | On the bound one, if there is a bound one. | 22:31 |
| nati_uen_ | rkukura: yes my current plan is let _update_port_dict_binding call driver | 22:32 |
| nati_uen_ | so i'm going to add update_port_dict_binding on each driver | 22:32 |
| nati_uen_ | and inside update_port_dict_binding code, I'll add vif_security dict | 22:32 |
| rkukura | I guess I think its worth thinking about whether that function should be something like MechnanismDriver.get_vif_security(), or something like MechanismDriver.extend_port_dict() | 22:33 |
| nati_uen_ | I'm +1 for MechanismDriver.extend_port_dict() | 22:33 |
| *** thedodd has quit IRC | 22:33 | |
| rkukura | mestery: What do you think of that option? | 22:33 |
| mestery | rkukura: I like that option. | 22:35 |
| nati_uen_ | ok deal :) I'll add new MD method "extend_port_dict(port)" | 22:35 |
| nati_uen_ | then call it from _update_port_dict_binding | 22:35 |
| nati_uen_ | Thank you for your suggestions! | 22:35 |
| rkukura | Sounds good - and the doc string in driver_api.py should specify that its called inside at transaction, I think. | 22:36 |
| nati_uen_ | rkukura: sure. I'll add doc strings also | 22:36 |
| rkukura | nati_uen_: Sounds like a plan! | 22:37 |
| nati_uen_ | russellb: Thanks! | 22:37 |
| rkukura | nati_uen_: Please don't hesitrate to ask for advice if you run into any major gotchas. | 22:37 |
| nati_uen_ | typo | 22:37 |
| nati_uen_ | rkukura: sure | 22:38 |
| *** yamahata_ has quit IRC | 22:38 | |
| nati_uen_ | rkukura: mestery: may I ask few more questions? | 22:41 |
| rkukura | sure | 22:42 |
| mestery | sure | 22:42 |
| nati_uen_ | How I can get Mechdrvier from inside of _update_port_dict_binding ? | 22:42 |
| nati_uen_ | It looks it get PortContext | 22:42 |
| nati_uen_ | May be I should update MechanismManager also | 22:43 |
| *** suresh12 has joined #openstack-neutron | 22:44 | |
| rkukura | It looks like everywhere that calls _update_port_dict_binding() has a PortContext, so maybe you can just pass this to _update_port_dict_binding() as well. | 22:45 |
| mestery | rkukura: Another ML2 question: Should we do the ML2 meeting tomorrow morning? | 22:47 |
| mestery | Will you be around? | 22:47 |
| rkukura | Well _ml2_extend_port_dict_binding() does not have a PortContext, and I'm not positive whether its inside a tranaction | 22:47 |
| nati_uen_ | so let's _update_port_dict_binding have a mech context? | 22:48 |
| nati_uen_ | ah I got your point now | 22:49 |
| nati_uen_ | _ml2_extend_port_dict_binding can get only port and binding | 22:49 |
| mestery | That makes sense to me. | 22:50 |
| mestery | nati_uen_ rkukura: I need to take off now for a bit. | 22:52 |
| nati_uen_ | mestery: Thanks for your help :) | 22:52 |
| nati_uen_ | tl | 22:52 |
| mestery | rkukura: Let me know your thoughts on ML2 meeting, have 2 action items to cover, both could be done in email if needed. | 22:52 |
| rkukura | I should be around, but not sure about others | 22:52 |
| *** yamahata_ has joined #openstack-neutron | 22:57 | |
| *** suresh12 has quit IRC | 22:58 | |
| nati_uen_ | rkukura: I'm not still not clear to how to implement this. it looks like Mec Manager call all drivers. Is this true? | 22:59 |
| nati_uen_ | How can we call actual mechdriver in update_port_dict_binding? | 23:00 |
| rkukura | Yes, you'd probably need to add a MechanismManager.extend_port_dict() function that takes care of calling the bound MechanismDriver. | 23:00 |
| nati_uen_ | rkukura: but _call_on_drivers looks calling all driver's method. How we can choose proper one? | 23:02 |
| rkukura | nati_uen_: Look at the validate_port_binding() and unbind_port() methods on MechansimManager | 23:03 |
| *** jpich has quit IRC | 23:04 | |
| nati_uen_ | rkukura: thanks! | 23:04 |
| nati_uen_ | can I do this? MechanismManager.extend_port_dict(binding, port_dict) | 23:05 |
| rkukura | I'd probably use MechanismManager.extend_port_dict(context, port_dict) to be more consistent with the other methods. Also, the context needs to be passed to the driver I think | 23:07 |
| nati_uen_ | rkukura: I agree, but how I can get context in _ml2_extend_port_dict_binding ? | 23:07 |
| rkukura | One consideration with all of this is whether port dictionaries and this notion of extending them will still exist with the refactoring markmcclain has been talking about | 23:08 |
| nati_uen_ | yeah, but this patch will be faster than his refactoring | 23:08 |
| nati_uen_ | Currently, neutron + nova security group isn't working.. so I wanna fix this issue as fast as I can | 23:09 |
| rkukura | nati_uen_: Agreed this will be merged first. | 23:09 |
| *** amotoki has quit IRC | 23:09 | |
| rkukura | Can you check to make sure _ml2_extend_port_dict_binding() is called within the transaction? | 23:10 |
| nati_uen_ | sure | 23:10 |
| *** pcm_ has quit IRC | 23:10 | |
| rkukura | And I guess _ml2_extend_port_dict_binding() will need to construct a PortContext | 23:11 |
| nati_uen_ | rkukura: this is outside of transaction https://github.com/openstack/neutron/blob/master/neutron/db/db_base_plugin_v2.py#L916 | 23:11 |
| rkukura | nati_uen_: I need to head out to dinner with my family. | 23:12 |
| nati_uen_ | rkukura: sure Thank you for your help. TL | 23:12 |
| rkukura | It does kind of look like the code avoids a transaction for the get_port() | 23:13 |
| rkukura | Maybe we can just specify that MechanismDriver.extend_port_dict() may be called inside or outside a transaction if that is the case | 23:13 |
| rkukura | nati_uen_: Send me an email or look for me online tomorrow if we need to continue this discussion. | 23:14 |
| nati_uen_ | rkukura: Thanks. I'll do my proposal in my review, I'll ping you when I finish 1st draft | 23:15 |
| *** julim has quit IRC | 23:16 | |
| *** banix has quit IRC | 23:17 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Rebind security groups only when they're updated https://review.openstack.org/58597 | 23:17 |
| *** otherwiseguy has quit IRC | 23:18 | |
| *** yamahata_ has quit IRC | 23:20 | |
| openstackgerrit | Armando Migliaccio proposed a change to openstack/neutron: Handle exceptions on create_dhcp_port https://review.openstack.org/57812 | 23:24 |
| *** reaper has quit IRC | 23:26 | |
| *** armax has left #openstack-neutron | 23:28 | |
| *** otherwiseguy has joined #openstack-neutron | 23:31 | |
| *** aymenfrikha has joined #openstack-neutron | 23:32 | |
| *** gdubreui has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!