| *** dvorkinista has quit IRC | 00:02 | |
| *** hrzbrg has joined #openstack-neutron | 00:02 | |
| *** hrzbrg has quit IRC | 00:05 | |
| *** dvorkinista has joined #openstack-neutron | 00:11 | |
| *** lnxnut has quit IRC | 00:12 | |
| *** suresh12 has quit IRC | 00:12 | |
| *** dvorkinista has quit IRC | 00:16 | |
| *** dvorkinista has joined #openstack-neutron | 00:16 | |
| *** dvorkinista has quit IRC | 00:25 | |
| *** matsuhashi has joined #openstack-neutron | 00:30 | |
| *** thuc has joined #openstack-neutron | 00:32 | |
| *** thuc_ has joined #openstack-neutron | 00:33 | |
| *** dvorkinista has joined #openstack-neutron | 00:34 | |
| *** baoli has joined #openstack-neutron | 00:34 | |
| *** thuc has quit IRC | 00:37 | |
| *** WackoRobie has joined #openstack-neutron | 00:45 | |
| *** thuc_ has quit IRC | 00:45 | |
| *** BuSerD has joined #openstack-neutron | 00:46 | |
| *** marun has joined #openstack-neutron | 00:47 | |
| *** matsuhashi has quit IRC | 00:52 | |
| *** dvorkinista has quit IRC | 00:57 | |
| *** matsuhashi has joined #openstack-neutron | 00:58 | |
| *** zhhuabj has joined #openstack-neutron | 01:02 | |
| *** banix has joined #openstack-neutron | 01:09 | |
| *** sbalukoff1 has joined #openstack-neutron | 01:15 | |
| *** sbalukoff has quit IRC | 01:15 | |
| *** jecarey has joined #openstack-neutron | 01:19 | |
| *** alexpilotti has quit IRC | 01:21 | |
| openstackgerrit | oda-g proposed a change to openstack/neutron: Make help texts more descriptive in Metaplugin https://review.openstack.org/88114 | 01:25 |
|---|---|---|
| *** oda-g has joined #openstack-neutron | 01:28 | |
| *** marun has quit IRC | 01:41 | |
| *** xuhanp has joined #openstack-neutron | 01:52 | |
| *** xianghui has joined #openstack-neutron | 01:56 | |
| *** xianghui has quit IRC | 02:03 | |
| *** xianghui has joined #openstack-neutron | 02:06 | |
| *** Longgeek has joined #openstack-neutron | 02:14 | |
| *** dvorkinista has joined #openstack-neutron | 02:14 | |
| *** dims has quit IRC | 02:29 | |
| *** dims has joined #openstack-neutron | 02:30 | |
| *** dvorkinista has quit IRC | 02:31 | |
| *** dvorkinista has joined #openstack-neutron | 02:31 | |
| *** dims has quit IRC | 02:35 | |
| *** banix has quit IRC | 02:35 | |
| *** dims has joined #openstack-neutron | 02:37 | |
| *** dvorkinista has quit IRC | 02:42 | |
| *** yamamoto_ has joined #openstack-neutron | 02:42 | |
| *** BuSerD has quit IRC | 02:44 | |
| openstackgerrit | shihanzhang proposed a change to openstack/neutron: Disallow changing quota to lower than current usage https://review.openstack.org/87841 | 02:45 |
| *** yamahata has joined #openstack-neutron | 02:46 | |
| *** dvorkinista has joined #openstack-neutron | 02:48 | |
| *** carl_baldwin has joined #openstack-neutron | 02:54 | |
| *** dvorkinista has quit IRC | 02:54 | |
| *** networkstatic has joined #openstack-neutron | 02:56 | |
| xuhanp | amotoki, ping | 02:57 |
| *** ramishra has joined #openstack-neutron | 03:02 | |
| *** ramishra_ has joined #openstack-neutron | 03:08 | |
| *** coolsvap|afk is now known as coolsvap | 03:09 | |
| *** ramishra has quit IRC | 03:10 | |
| *** carl_baldwin has quit IRC | 03:16 | |
| *** Longgeek has quit IRC | 03:17 | |
| *** dvorkinista has joined #openstack-neutron | 03:20 | |
| *** dvorkinista has quit IRC | 03:24 | |
| *** matsuhashi has quit IRC | 03:32 | |
| *** matsuhashi has joined #openstack-neutron | 03:32 | |
| *** matsuhashi has quit IRC | 03:37 | |
| *** carl_baldwin has joined #openstack-neutron | 03:43 | |
| *** baoli has quit IRC | 03:44 | |
| *** suresh12 has joined #openstack-neutron | 03:45 | |
| *** networkstatic has quit IRC | 03:49 | |
| *** dvorkinista has joined #openstack-neutron | 03:58 | |
| *** dvorkinista has quit IRC | 04:01 | |
| openstackgerrit | shihanzhang proposed a change to openstack/neutron: Disallow changing quota to lower than current usage https://review.openstack.org/87841 | 04:03 |
| *** vthapar has joined #openstack-neutron | 04:12 | |
| openstackgerrit | fumihiko kakuma proposed a change to openstack/neutron: OFAgent: Port implementation of l2-population from OVS https://review.openstack.org/87440 | 04:14 |
| *** dvorkinista has joined #openstack-neutron | 04:28 | |
| *** matsuhashi has joined #openstack-neutron | 04:29 | |
| *** dvorkinista has quit IRC | 04:30 | |
| *** chandan_kumar has joined #openstack-neutron | 04:32 | |
| *** iwamoto has joined #openstack-neutron | 04:40 | |
| *** rand738 has quit IRC | 04:46 | |
| *** nati_ueno has joined #openstack-neutron | 04:51 | |
| *** rand738 has joined #openstack-neutron | 04:55 | |
| *** nati_ueno has quit IRC | 05:08 | |
| *** SridharG has joined #openstack-neutron | 05:11 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron-specs: ML2 Mechanism Driver for the Cisco APIC https://review.openstack.org/89305 | 05:25 |
| *** carl_baldwin has quit IRC | 05:28 | |
| *** carl_baldwin has joined #openstack-neutron | 05:39 | |
| *** dvorkinista has joined #openstack-neutron | 05:41 | |
| *** Longgeek has joined #openstack-neutron | 05:42 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron-specs: Allow PNG and JPG images up to 100K in size https://review.openstack.org/89306 | 05:45 |
| *** dvorkinista has quit IRC | 05:45 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron-specs: ML2 Mechanism Driver for the Cisco APIC https://review.openstack.org/89305 | 05:47 |
| openstackgerrit | IWAMOTO Toshihiro proposed a change to openstack/neutron-specs: Add lbaas-lvs-driver https://review.openstack.org/88213 | 05:48 |
| *** networkstatic has joined #openstack-neutron | 05:49 | |
| *** carl_baldwin has quit IRC | 05:49 | |
| *** Longgeek has quit IRC | 05:58 | |
| *** matsuhashi has quit IRC | 05:59 | |
| *** matsuhashi has joined #openstack-neutron | 06:00 | |
| *** yamahata has quit IRC | 06:05 | |
| *** nlahouti has joined #openstack-neutron | 06:09 | |
| *** sungju has joined #openstack-neutron | 06:13 | |
| *** sungju has quit IRC | 06:14 | |
| *** dvorkinista has joined #openstack-neutron | 06:19 | |
| *** suresh12 has quit IRC | 06:26 | |
| *** matsuhashi has quit IRC | 06:32 | |
| *** matsuhashi has joined #openstack-neutron | 06:34 | |
| *** nati_ueno has joined #openstack-neutron | 06:37 | |
| *** dvorkinista has quit IRC | 06:41 | |
| *** SridharG has quit IRC | 06:42 | |
| *** Longgeek has joined #openstack-neutron | 06:50 | |
| *** networkstatic has quit IRC | 06:53 | |
| enikanorov_ | nlahouti: | 06:54 |
| *** Longgeek has quit IRC | 06:55 | |
| nlahouti | enikanorov: Hi | 06:55 |
| nlahouti | thanks for reviewing the BP and your comments. | 06:55 |
| enikanorov_ | np | 06:56 |
| nlahouti | I would like to know more details on how to implement the BP | 06:56 |
| nlahouti | in more general way | 06:56 |
| enikanorov_ | ok, so i think this BP should be mostly around creating a new mechanism in extension framework for loading extensions | 06:57 |
| enikanorov_ | so right now all extensions are stored in /extension folder and are referenced by the plugins via aliases | 06:57 |
| nlahouti | enikanorov:yes that's correct | 06:57 |
| enikanorov_ | if we want particular vendor drivers to implement specific extensions, it make sense to at least keep them in driver's folder | 06:58 |
| enikanorov_ | so existing mechanism is not suitable for such extensions | 06:58 |
| nlahouti | we can do that. | 06:58 |
| enikanorov_ | i think extension framework should be amended with the following functionality: | 06:59 |
| nlahouti | but do you see any issue with proposal to get the the extensions from Ml2Plugin? | 06:59 |
| enikanorov_ | nlahouti: the issue is in how it is done | 06:59 |
| enikanorov_ | nlahouti: aliases is not a suitable way i think | 06:59 |
| enikanorov_ | aliases are good for common extensions that are part of common API | 07:00 |
| nlahouti | But this is how all extensions are implemented. No? | 07:00 |
| enikanorov_ | right | 07:00 |
| enikanorov_ | that's because we yet don't have vendor-specific (driver-specific) extensions | 07:00 |
| enikanorov_ | e.g. we need to create a way of loading 'non-common API' | 07:01 |
| nlahouti | right. you mean the APIs for extensions GET/CREATE/... | 07:02 |
| enikanorov_ | extensions - i mean additional resources or attributes for existing resources | 07:03 |
| enikanorov_ | so, going back to the patch, right now extension loading relies on the plugin to collect aliases | 07:04 |
| nlahouti | Yes. | 07:04 |
| enikanorov_ | i think we need another way of doing that: for drivers that need to expose additional API (attributes or resources), they should return whole resources or extension itself directly | 07:05 |
| *** kevinbenton has quit IRC | 07:05 | |
| enikanorov_ | in other words, instead of returning an alias to a comon extension (alias is only good for common ext), driver (and then plugin) returns extension itself | 07:05 |
| enikanorov_ | and it is then processed by extension framework, e.g. resources/attributes from that extension are embedded into API | 07:06 |
| nlahouti | how would that work in existing extensions framework? | 07:07 |
| nlahouti | Does it mean we can still use the PluginAwareExtensionManager? | 07:08 |
| *** kevinbenton has joined #openstack-neutron | 07:12 | |
| enikanorov_ | nlahouti: yeah, i think it should also ask the plugin for extensions, and not just for aliases | 07:15 |
| nlahouti | What I'm not clear is that, the extension manager is using aliases for loading the extension.- and you are saying instead of returinnig aliases return extension itself. | 07:15 |
| nlahouti | Do you have any example for that? | 07:15 |
| nlahouti | Ok. You are saying the extension manager should change as well. | 07:16 |
| nlahouti | right? | 07:16 |
| *** oda-g has quit IRC | 07:17 | |
| enikanorov_ | no, no examples because it should be new functionality | 07:17 |
| nlahouti | :) | 07:17 |
| enikanorov_ | not change, exactly, it should add the ability to process extension directly | 07:17 |
| nlahouti | I see. | 07:18 |
| enikanorov_ | common extensions are still loaded via aliases | 07:18 |
| enikanorov_ | driver-specific extenions are loaded through the plugin | 07:18 |
| nlahouti | So if I understood correctly, | 07:18 |
| enikanorov_ | e.g. PluginAwareExtensionManager -> plugin.get_extenions() -> drivers.get_extension() | 07:19 |
| enikanorov_ | something like that | 07:19 |
| nlahouti | ok | 07:19 |
| *** sweston has joined #openstack-neutron | 07:19 | |
| nlahouti | If I understood correctly, A plugin that supports a vendor specific extension shoud return the extensions (via get_extenions() implementation) | 07:20 |
| nlahouti | then the extesnion manager should use that and load the extensions | 07:20 |
| *** Longgeek has joined #openstack-neutron | 07:20 | |
| nlahouti | right? | 07:20 |
| enikanorov_ | yes, instead of alias it could be a list of extension objects and on each element it should be possible to call get_resources() | 07:22 |
| nlahouti | ok. | 07:23 |
| nlahouti | How about the APIs for attributes for example create_xxx, list_xxx, ... | 07:24 |
| nlahouti | in the v2_0/base.py? | 07:24 |
| enikanorov_ | not sure i understand | 07:25 |
| enikanorov_ | usually what goes after 'create_' is called resources | 07:25 |
| nlahouti | see, for example for a vendore specific, when we specify an extension for a resource (i.e. network) | 07:25 |
| nlahouti | we want to get for example the list of that extensions | 07:26 |
| nlahouti | I mean attribute | 07:26 |
| nlahouti | somehow the Plugin or mechanism driver has to implement that API. | 07:27 |
| nlahouti | as most of the API for the resource implemented in the neutron base class. | 07:27 |
| enikanorov_ | so, you're saying that for network resource there could be an attribute that is supported by some driver, but not supported by others? | 07:28 |
| nlahouti | yes | 07:28 |
| enikanorov_ | can you give an example? | 07:28 |
| nlahouti | yes. In my case, I added a an extension to the network resource i.e. config_profile. | 07:29 |
| nlahouti | When I want to create a network, I need to get the list of this config_profile. | 07:30 |
| nlahouti | but only my mechanism dirver is aware of that | 07:30 |
| enikanorov_ | btw, i was under impression that when ml2 plugin processes a call, it calls all mechanism drivers, no? | 07:30 |
| nlahouti | sure. | 07:31 |
| *** SridharG has joined #openstack-neutron | 07:31 | |
| nlahouti | but in the v2_0/base.py when REST API call comes, | 07:31 |
| enikanorov_ | so other drivers could just skip unknown attributes (if we're talking about attributes) | 07:31 |
| nlahouti | it cannot find any attribute that matches with extenion in plugin. | 07:31 |
| nlahouti | It throw exception 'attribute error' | 07:32 |
| nlahouti | and also never calls the mechanism driver | 07:32 |
| enikanorov_ | that's because attribute was not embedded into 'network' resource definition | 07:32 |
| nlahouti | as base.py api only knows about Plugin | 07:32 |
| *** nati_ueno has quit IRC | 07:32 | |
| nlahouti | can you elaborate on that? | 07:33 |
| enikanorov_ | when API doesn't know about the attribute, it will not allow such call, and it will never hit the plugin | 07:33 |
| enikanorov_ | nlahouti: some how you didn't update 'network' resource with your extended attribute, that's why API is throwing exception | 07:33 |
| nlahouti | API goes to the supported Plugin but cannot find it in the plugin. | 07:34 |
| nlahouti | API expect the plugin implement the method | 07:34 |
| nlahouti | as they are mandatory in the Neutron Base class. for the resource. | 07:34 |
| enikanorov_ | now i don't understand. we were talking about 'network' resource with additional attribute | 07:35 |
| nlahouti | Yes | 07:36 |
| enikanorov_ | 'network' resource already has all methods in plugins or base plugin | 07:36 |
| nlahouti | sure | 07:36 |
| enikanorov_ | so nothing else needed | 07:36 |
| enikanorov_ | can you show the traceback? | 07:36 |
| *** suresh12 has joined #openstack-neutron | 07:37 | |
| nlahouti | sure. | 07:37 |
| nlahouti | cut&paste here? | 07:37 |
| nlahouti | ===>>>> log error in: 2014-03-28 12:27:02.494 DEBUG routes.middleware [-] Match dict: {'action': u'ind ex', 'controller': <wsgify at 79934352 wrapping <function resource at 0x4c2fa28> >, 'format': u'json'} from (pid=2056) __call__ /usr/lib/python2.7/dist-packages/ routes/middleware.py:103 2014-03-28 12:27:02.495 ERROR NL>>> resource.py: neutron.api.v2.resource [-] ind ex failed 2014-03-28 12:27:02.495 TRACE NL>>> resour | 07:37 |
| nlahouti | Is it readable? | 07:37 |
| enikanorov_ | paste.openstack.org | 07:38 |
| nlahouti | I just paste there | 07:39 |
| nlahouti | can you read it? | 07:40 |
| *** suresh12 has quit IRC | 07:41 | |
| enikanorov_ | i guess it was not a full traceback, the link above is more convenient for this | 07:41 |
| nlahouti | where you able to see in paste.openstack.org? | 07:42 |
| enikanorov_ | you should give a link to paste - look at the address bar! :) | 07:43 |
| *** lukego has joined #openstack-neutron | 07:43 | |
| nlahouti | :) sorry about that never used it | 07:43 |
| nlahouti | Here it is | 07:43 |
| nlahouti | http://paste.openstack.org/?reply_to=76485 | 07:43 |
| nlahouti | nice feature. | 07:44 |
| nlahouti | I need to leave in 5 min. here is late. | 07:44 |
| nlahouti | Can we discuss it tomorrow morning, if you are available. | 07:44 |
| enikanorov_ | morning which timezone? | 07:45 |
| nlahouti | PT | 07:45 |
| nlahouti | I'm in California | 07:45 |
| enikanorov_ | i see, yeah. probably i'll be online. looking at your traceback it seems that you were accessing config_profiles resource, not an attribute | 07:46 |
| nlahouti | yes. | 07:47 |
| enikanorov_ | ok, let's continue in the morning then (will be evening for me) | 07:49 |
| nlahouti | Thanks a lot for your time | 07:50 |
| nlahouti | will talk to you tomorrow. | 07:50 |
| *** dvorkinista has joined #openstack-neutron | 07:51 | |
| *** sweston has quit IRC | 07:53 | |
| *** yamamoto_ has quit IRC | 07:54 | |
| *** dvorkinista has quit IRC | 07:56 | |
| *** alexpilotti has joined #openstack-neutron | 08:06 | |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: Fix migration that breaks Grenade jobs https://review.openstack.org/87935 | 08:06 |
| *** ramishra_ has quit IRC | 08:06 | |
| *** kevinbenton has quit IRC | 08:17 | |
| *** kevinbenton has joined #openstack-neutron | 08:23 | |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: Set correct nullable parameter for columns https://review.openstack.org/82089 | 08:27 |
| *** hrzbrg has joined #openstack-neutron | 08:28 | |
| *** alexpilotti has quit IRC | 08:43 | |
| *** ramishra has joined #openstack-neutron | 08:49 | |
| *** chandan_kumar has quit IRC | 08:50 | |
| *** overlayer has joined #openstack-neutron | 08:52 | |
| *** overlayer has quit IRC | 08:52 | |
| *** overlayer has joined #openstack-neutron | 08:55 | |
| *** bashok has joined #openstack-neutron | 08:56 | |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: Fix incorrect change of Enum type https://review.openstack.org/84730 | 09:05 |
| *** chandan_kumar has joined #openstack-neutron | 09:08 | |
| *** iwamoto has quit IRC | 09:14 | |
| *** bvandenh has joined #openstack-neutron | 09:16 | |
| *** salv-orlando has quit IRC | 09:25 | |
| *** sweston has joined #openstack-neutron | 09:26 | |
| *** leseb has joined #openstack-neutron | 09:40 | |
| *** matsuhashi has quit IRC | 09:48 | |
| *** xuhanp has quit IRC | 09:50 | |
| *** xuhanp has joined #openstack-neutron | 09:51 | |
| *** matsuhashi has joined #openstack-neutron | 09:54 | |
| *** baojg_ has joined #openstack-neutron | 09:57 | |
| *** pcm_ has joined #openstack-neutron | 10:01 | |
| *** pcm_ has quit IRC | 10:01 | |
| *** pcm_ has joined #openstack-neutron | 10:02 | |
| *** Longgeek has quit IRC | 10:03 | |
| *** coolsvap is now known as coolsvap|afk | 10:03 | |
| *** Longgeek has joined #openstack-neutron | 10:12 | |
| *** alex_klimov has joined #openstack-neutron | 10:15 | |
| *** baojg_ has quit IRC | 10:18 | |
| *** baojg has joined #openstack-neutron | 10:18 | |
| *** xuhanp has quit IRC | 10:24 | |
| *** Longgeek has quit IRC | 10:29 | |
| *** b3nt_pin has joined #openstack-neutron | 10:42 | |
| *** chandan_kumar has quit IRC | 10:43 | |
| *** overlayer has quit IRC | 10:43 | |
| *** leseb has quit IRC | 10:44 | |
| *** leseb has joined #openstack-neutron | 10:44 | |
| *** baojg_ has joined #openstack-neutron | 10:46 | |
| *** leseb has quit IRC | 10:49 | |
| *** baojg has quit IRC | 10:50 | |
| *** chandan_kumar has joined #openstack-neutron | 10:51 | |
| *** lukego has quit IRC | 10:54 | |
| *** baojg_ has quit IRC | 11:03 | |
| *** baojg has joined #openstack-neutron | 11:03 | |
| *** Longgeek has joined #openstack-neutron | 11:05 | |
| *** baojg has quit IRC | 11:07 | |
| *** Longgeek has quit IRC | 11:10 | |
| *** leseb has joined #openstack-neutron | 11:15 | |
| *** lukego has joined #openstack-neutron | 11:18 | |
| *** leseb has quit IRC | 11:19 | |
| *** matsuhashi has quit IRC | 11:22 | |
| *** phil_h has joined #openstack-neutron | 11:29 | |
| *** xianghui has quit IRC | 11:30 | |
| *** matsuhashi has joined #openstack-neutron | 11:30 | |
| *** b3nt_pin has quit IRC | 11:32 | |
| *** b3nt_pin has joined #openstack-neutron | 11:36 | |
| *** matsuhashi has quit IRC | 11:48 | |
| *** matsuhashi has joined #openstack-neutron | 11:49 | |
| *** baoli has joined #openstack-neutron | 11:53 | |
| *** baoli has quit IRC | 11:54 | |
| *** Vicky_Pawar has joined #openstack-neutron | 12:05 | |
| *** bashok_ has joined #openstack-neutron | 12:09 | |
| *** bashok has quit IRC | 12:09 | |
| *** WackoRobie has quit IRC | 12:10 | |
| *** erecio has joined #openstack-neutron | 12:11 | |
| *** Vicky_Pawar has quit IRC | 12:15 | |
| *** b3nt_pin is now known as beagles | 12:15 | |
| *** leseb has joined #openstack-neutron | 12:15 | |
| *** zhipeng has joined #openstack-neutron | 12:19 | |
| *** leseb has quit IRC | 12:20 | |
| *** Longgeek has joined #openstack-neutron | 12:22 | |
| *** ilyashakhat has quit IRC | 12:22 | |
| *** zhipeng has quit IRC | 12:37 | |
| *** ramishra has quit IRC | 12:40 | |
| *** matsuhashi has quit IRC | 12:42 | |
| *** matsuhashi has joined #openstack-neutron | 12:44 | |
| *** matsuhashi has quit IRC | 12:51 | |
| *** baoli has joined #openstack-neutron | 12:56 | |
| *** WackoRobie has joined #openstack-neutron | 13:02 | |
| *** jecarey has quit IRC | 13:05 | |
| *** BillTheKat has joined #openstack-neutron | 13:05 | |
| *** ramishra has joined #openstack-neutron | 13:06 | |
| *** overlayer has joined #openstack-neutron | 13:06 | |
| *** matsuhashi has joined #openstack-neutron | 13:06 | |
| *** Jianyong has joined #openstack-neutron | 13:09 | |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: Fix migration that breaks Grenade jobs https://review.openstack.org/87935 | 13:09 |
| *** kbringard has joined #openstack-neutron | 13:09 | |
| *** matsuhashi has quit IRC | 13:12 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron-specs: ML2 Mechanism Driver for the Cisco APIC https://review.openstack.org/89305 | 13:13 |
| openstackgerrit | Ann Kamyshnikova proposed a change to openstack/neutron: Fix migration that breaks Grenade jobs https://review.openstack.org/87935 | 13:15 |
| *** leseb has joined #openstack-neutron | 13:16 | |
| *** nlahouti has quit IRC | 13:17 | |
| *** jecarey has joined #openstack-neutron | 13:20 | |
| *** leseb has quit IRC | 13:21 | |
| *** matsuhashi has joined #openstack-neutron | 13:21 | |
| *** tongli has joined #openstack-neutron | 13:24 | |
| *** overlayer has quit IRC | 13:28 | |
| *** leseb has joined #openstack-neutron | 13:32 | |
| *** jgrimm has quit IRC | 13:39 | |
| *** ramishra has quit IRC | 13:39 | |
| *** mwagner_lap has quit IRC | 13:40 | |
| *** banix has joined #openstack-neutron | 13:43 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron: Cisco APIC ML2 mechanism driver, part 1 https://review.openstack.org/73355 | 13:48 |
| *** SridharG has quit IRC | 13:53 | |
| *** julim has joined #openstack-neutron | 13:53 | |
| *** julim_ has joined #openstack-neutron | 13:55 | |
| *** nati_ueno has joined #openstack-neutron | 13:58 | |
| *** julim has quit IRC | 13:58 | |
| *** jobewan has joined #openstack-neutron | 14:00 | |
| *** mlavalle has joined #openstack-neutron | 14:02 | |
| *** tongli has quit IRC | 14:03 | |
| *** dvorkinista has joined #openstack-neutron | 14:03 | |
| *** djoreilly has joined #openstack-neutron | 14:03 | |
| *** nati_ueno has quit IRC | 14:04 | |
| *** jaypipes has joined #openstack-neutron | 14:09 | |
| *** markmcclain has joined #openstack-neutron | 14:10 | |
| *** rwsu has joined #openstack-neutron | 14:11 | |
| *** thedodd has joined #openstack-neutron | 14:11 | |
| *** doddstack has joined #openstack-neutron | 14:14 | |
| *** doddstack has quit IRC | 14:14 | |
| *** chandan_kumar has quit IRC | 14:14 | |
| *** doddstack has joined #openstack-neutron | 14:14 | |
| *** doddstack has quit IRC | 14:14 | |
| *** doddstack has joined #openstack-neutron | 14:15 | |
| openstackgerrit | rcurran proposed a change to openstack/neutron: ML2 Cisco Nexus MD: Improve Unit Test Coverage https://review.openstack.org/88695 | 14:15 |
| *** doddstack has quit IRC | 14:16 | |
| *** matsuhashi has quit IRC | 14:17 | |
| *** thedodd has quit IRC | 14:17 | |
| *** xuhanp has joined #openstack-neutron | 14:18 | |
| *** ramishra has joined #openstack-neutron | 14:21 | |
| *** dims has quit IRC | 14:27 | |
| *** Longgeek has quit IRC | 14:27 | |
| *** dims has joined #openstack-neutron | 14:29 | |
| openstackgerrit | rcurran proposed a change to openstack/neutron: ML2 Cisco Nexus MD: Improve Unit Test Coverage https://review.openstack.org/88695 | 14:34 |
| *** jgrimm has joined #openstack-neutron | 14:35 | |
| *** chandan_kumar has joined #openstack-neutron | 14:35 | |
| *** clev has left #openstack-neutron | 14:39 | |
| *** bashok_ has quit IRC | 14:41 | |
| *** dvorkinista has quit IRC | 14:43 | |
| *** alagalah has quit IRC | 14:47 | |
| *** TrevorV has joined #openstack-neutron | 14:53 | |
| *** Jianyong has quit IRC | 14:55 | |
| *** vthapar has quit IRC | 14:57 | |
| *** carlp has joined #openstack-neutron | 15:00 | |
| *** SridharG has joined #openstack-neutron | 15:05 | |
| *** alex_klimov has quit IRC | 15:07 | |
| *** sbalukoff1 has quit IRC | 15:11 | |
| *** carl_baldwin has joined #openstack-neutron | 15:15 | |
| openstackgerrit | Paul Michali proposed a change to openstack/neutron-specs: Publish L3 service capabilites to client https://review.openstack.org/89366 | 15:22 |
| *** mwagner_lap has joined #openstack-neutron | 15:23 | |
| *** otherwiseguy has joined #openstack-neutron | 15:26 | |
| *** chandan_kumar has quit IRC | 15:29 | |
| *** jorgem has joined #openstack-neutron | 15:31 | |
| *** leseb has quit IRC | 15:35 | |
| *** reaper has joined #openstack-neutron | 15:35 | |
| *** dvorkinista has joined #openstack-neutron | 15:38 | |
| *** jgrimm_ has joined #openstack-neutron | 15:38 | |
| *** sballe has joined #openstack-neutron | 15:39 | |
| *** jgrimm has quit IRC | 15:39 | |
| *** reaper has quit IRC | 15:42 | |
| *** lukego has quit IRC | 15:44 | |
| *** lukego has joined #openstack-neutron | 15:47 | |
| *** otherwiseguy has quit IRC | 15:47 | |
| sc68cal | mestery: Do we have some logic in the tox config for neutron-specs to pick up new additions to requirements.txt? | 15:49 |
| mestery | sc68cal: Not sure, let me check, but FYI, on the phone right now. | 15:49 |
| sc68cal | I had to nuke my .tox dir to pick up the nwdiag deps and get them installed correctly | 15:49 |
| *** otherwiseguy has joined #openstack-neutron | 15:50 | |
| *** alagalah has joined #openstack-neutron | 15:54 | |
| *** dvorkinista has quit IRC | 15:55 | |
| *** dvorkinista has joined #openstack-neutron | 15:56 | |
| *** carl_baldwin has quit IRC | 15:56 | |
| *** phil_h has quit IRC | 15:56 | |
| *** alagalah has quit IRC | 15:59 | |
| *** alagalah has joined #openstack-neutron | 16:00 | |
| *** otherwiseguy has quit IRC | 16:05 | |
| *** nlahouti has joined #openstack-neutron | 16:06 | |
| nlahouti | enikanorov_: Hi Can we continue our conversation, if you have time? | 16:07 |
| *** blogan has joined #openstack-neutron | 16:09 | |
| *** xuhanp has quit IRC | 16:10 | |
| *** marun has joined #openstack-neutron | 16:12 | |
| *** csd has quit IRC | 16:14 | |
| *** networkstatic has joined #openstack-neutron | 16:16 | |
| *** lukego has quit IRC | 16:19 | |
| *** coolsvap|afk is now known as coolsvap | 16:20 | |
| *** nati_ueno has joined #openstack-neutron | 16:21 | |
| *** otherwiseguy has joined #openstack-neutron | 16:21 | |
| *** lukego has joined #openstack-neutron | 16:28 | |
| *** mriedem has joined #openstack-neutron | 16:33 | |
| *** sbfox has joined #openstack-neutron | 16:33 | |
| mriedem | question on amqp setup if anyone knows, looking at the config-reference doc it says to use control_exchange=openstack for the compute node which is the default exchange in oslo.messaging, | 16:34 |
| mriedem | but the network config says to use 'neutron' for the control_exchange: http://docs.openstack.org/trunk/config-reference/content/networking-configuring-rpc.html#networking-common-messaging-configuration | 16:34 |
| mriedem | is that intentional? | 16:34 |
| mriedem | i'm using qpid and have been seeing it trip up a lot lately, then messages get lost between nova and neutron for callbacks | 16:35 |
| *** leseb has joined #openstack-neutron | 16:36 | |
| mriedem | or maybe it's a security policy thing http://docs.openstack.org/security-guide/content/ch038_transport-security.html | 16:38 |
| *** leseb has quit IRC | 16:40 | |
| *** sweston has quit IRC | 16:43 | |
| openstackgerrit | Paul Michali proposed a change to openstack/neutron-specs: BP: Provider Validation for L3 Plugins https://review.openstack.org/88406 | 16:45 |
| *** leseb has joined #openstack-neutron | 16:45 | |
| *** HenryG_ has quit IRC | 16:47 | |
| *** HenryG has joined #openstack-neutron | 16:47 | |
| *** harlowja_away is now known as harlowja | 16:48 | |
| *** amcrn has joined #openstack-neutron | 16:49 | |
| *** leseb has quit IRC | 16:49 | |
| *** BuSerD has joined #openstack-neutron | 16:54 | |
| *** dvorkinista has quit IRC | 16:57 | |
| *** manishg has joined #openstack-neutron | 16:59 | |
| *** dvorkinista has joined #openstack-neutron | 17:02 | |
| *** sbfox has quit IRC | 17:09 | |
| *** alagalah has quit IRC | 17:12 | |
| *** chuckC has quit IRC | 17:13 | |
| *** alagalah has joined #openstack-neutron | 17:13 | |
| *** zzelle has joined #openstack-neutron | 17:15 | |
| openstackgerrit | A change was merged to openstack/neutron: NSX: sync thread catches wrong exceptions on not found https://review.openstack.org/88450 | 17:16 |
| enikanorov_ | nlahouti: in 10 minutes | 17:17 |
| *** alagalah_ has joined #openstack-neutron | 17:19 | |
| *** alagalah has quit IRC | 17:19 | |
| *** chandan_kumar has joined #openstack-neutron | 17:21 | |
| *** suresh12 has joined #openstack-neutron | 17:26 | |
| *** sweston has joined #openstack-neutron | 17:32 | |
| enikanorov_ | nlahouti: hey | 17:35 |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron-specs: ML2 Mechanism Driver for the Cisco APIC https://review.openstack.org/89305 | 17:36 |
| *** terrylhowe has joined #openstack-neutron | 17:40 | |
| *** terrylhowe has left #openstack-neutron | 17:40 | |
| *** morganfainberg_Z is now known as morganfainberg | 17:42 | |
| *** leseb has joined #openstack-neutron | 17:46 | |
| *** leseb has quit IRC | 17:50 | |
| *** phil_h has joined #openstack-neutron | 17:53 | |
| *** chuckC has joined #openstack-neutron | 17:55 | |
| *** sacharya has joined #openstack-neutron | 17:55 | |
| *** tongli has joined #openstack-neutron | 17:58 | |
| *** ramishra has quit IRC | 18:02 | |
| *** sbalukoff has joined #openstack-neutron | 18:02 | |
| *** sbfox has joined #openstack-neutron | 18:03 | |
| *** markmcclain has quit IRC | 18:03 | |
| *** coolsvap is now known as coolsvap|afk | 18:18 | |
| *** sbfox has quit IRC | 18:20 | |
| nlahouti | enikanorov: Hi enikanorov. | 18:20 |
| nlahouti | enikanorov: sorry missed ur message. I was in meeting. | 18:21 |
| *** ale____ has joined #openstack-neutron | 18:22 | |
| *** ale____ has quit IRC | 18:27 | |
| *** ale____ has joined #openstack-neutron | 18:27 | |
| openstackgerrit | Sean M. Collins proposed a change to openstack/neutron-specs: Add ipv6-provider-nets-slaac https://review.openstack.org/88043 | 18:28 |
| *** otherwiseguy has quit IRC | 18:29 | |
| *** ale_____ has joined #openstack-neutron | 18:29 | |
| *** SridharG has quit IRC | 18:29 | |
| *** harmw has quit IRC | 18:30 | |
| openstackgerrit | Sean M. Collins proposed a change to openstack/neutron-specs: Add ipv6-provider-nets-slaac https://review.openstack.org/88043 | 18:31 |
| *** harmw has joined #openstack-neutron | 18:31 | |
| *** sbfox has joined #openstack-neutron | 18:31 | |
| *** SridharG has joined #openstack-neutron | 18:31 | |
| *** ale____ has quit IRC | 18:33 | |
| adam_g | hi--im trying to fix some of the FWAAS tempest tests so that they pass against environments other than the default devstack. hitting an issue where firewalls never leave PENDING_CREATE. which network resources are required to have been created on the tenant network prior to firewall creatin? i have a network, subnet, router, and router interface but still PENDING_CREATE | 18:36 |
| *** markmcclain has joined #openstack-neutron | 18:45 | |
| *** nplanel has quit IRC | 18:55 | |
| *** chandan_kumar has quit IRC | 18:55 | |
| *** chandan_kumar has joined #openstack-neutron | 18:56 | |
| *** nati_ueno has quit IRC | 18:57 | |
| *** nati_ueno has joined #openstack-neutron | 18:58 | |
| *** chuckC has quit IRC | 19:01 | |
| *** nati_ueno has quit IRC | 19:05 | |
| nlahouti | enikanorov: Hi enikanorov will you be available? | 19:06 |
| *** nati_ueno has joined #openstack-neutron | 19:07 | |
| *** TrevorV has quit IRC | 19:10 | |
| *** TrevorV has joined #openstack-neutron | 19:12 | |
| *** otherwiseguy has joined #openstack-neutron | 19:18 | |
| *** amotoki_ has joined #openstack-neutron | 19:21 | |
| *** tchaypo has left #openstack-neutron | 19:23 | |
| *** dims has quit IRC | 19:25 | |
| *** dims has joined #openstack-neutron | 19:27 | |
| enikanorov_ | nlahouti: yes | 19:29 |
| enikanorov_ | nlahouti: so, continuing our discussion, traceback shows that you're accessing new resource, that was not embedded into API layer (at least it seems so) | 19:30 |
| *** erecio has quit IRC | 19:33 | |
| *** erecio has joined #openstack-neutron | 19:33 | |
| *** chandan_kumar has quit IRC | 19:34 | |
| *** TrevorV has quit IRC | 19:39 | |
| rkukura | SumitNaiksatam: Do you have any changes about to go into your sumit/pm branch, or other branches? | 19:40 |
| *** leseb has joined #openstack-neutron | 19:40 | |
| *** networkstatic is now known as networkstatic_zZ | 19:43 | |
| *** otherwiseguy has quit IRC | 19:44 | |
| *** zzelle_ has joined #openstack-neutron | 19:44 | |
| zzelle_ | markmcclain, hi | 19:45 |
| markmcclain | zzelle: hi | 19:45 |
| *** leseb has quit IRC | 19:46 | |
| zzelle_ | could you remove your -2 (liked to icehouse FF) on https://review.openstack.org/71904 ? | 19:46 |
| zzelle_ | s/liked/linked/ | 19:46 |
| *** TrevorV has joined #openstack-neutron | 19:46 | |
| zzelle_ | markmcclain, should i also change the BP approver in launchpad ? | 19:47 |
| markmcclain | zzelle: you need to refile the blueprint into the neutron-specs repo | 19:48 |
| zzelle_ | markmcclain, that is these week job :) | 19:49 |
| markmcclain | the -2 will be removed once the spec is approved and the doc is linked to the launchpad bp | 19:49 |
| zzelle_ | markmcclain, fine | 19:50 |
| *** lukego has quit IRC | 19:52 | |
| *** baoli has quit IRC | 19:52 | |
| markmcclain | zzelle: thanks, I know this it is a bit of a pain, but this is the only way we can ensure the specs repo contains the proper docs | 19:53 |
| *** thedodd has joined #openstack-neutron | 19:55 | |
| *** zzelle_ has quit IRC | 19:56 | |
| *** lukego has joined #openstack-neutron | 19:58 | |
| *** amcrn has quit IRC | 19:58 | |
| *** mwagner_lap has quit IRC | 19:59 | |
| *** zzelle has quit IRC | 19:59 | |
| *** ctooley has joined #openstack-neutron | 20:00 | |
| *** TrevorV has quit IRC | 20:03 | |
| *** ale_____ has quit IRC | 20:04 | |
| *** hrzbrg has quit IRC | 20:06 | |
| *** hrzbrg has joined #openstack-neutron | 20:07 | |
| openstackgerrit | A change was merged to openstack/neutron: Register LBaaS resources to quotas engine https://review.openstack.org/86624 | 20:07 |
| *** ale____ has joined #openstack-neutron | 20:07 | |
| *** alagalah_ has quit IRC | 20:08 | |
| *** alagalah has joined #openstack-neutron | 20:08 | |
| *** hrzbrg has quit IRC | 20:08 | |
| *** nati_ueno has quit IRC | 20:08 | |
| *** tongli has quit IRC | 20:09 | |
| *** erecio has quit IRC | 20:09 | |
| *** alagalah has quit IRC | 20:09 | |
| *** kfox1111 has joined #openstack-neutron | 20:09 | |
| *** alagalah has joined #openstack-neutron | 20:10 | |
| *** erecio has joined #openstack-neutron | 20:10 | |
| *** erecio has quit IRC | 20:10 | |
| kfox1111 | is there a way to have floating ip's for a subnet and not use seperate subnets? | 20:10 |
| kfox1111 | ie, have a router with no gateway, and have it assign floating ip's from the only subnet attached to the router? | 20:10 |
| *** erecio has joined #openstack-neutron | 20:11 | |
| *** leseb has joined #openstack-neutron | 20:13 | |
| *** ale_____ has joined #openstack-neutron | 20:13 | |
| *** catohornet1 has joined #openstack-neutron | 20:14 | |
| *** erecio has quit IRC | 20:14 | |
| *** erecio has joined #openstack-neutron | 20:14 | |
| *** erecio has quit IRC | 20:15 | |
| *** erecio has joined #openstack-neutron | 20:15 | |
| *** ale____ has quit IRC | 20:15 | |
| kevinbenton | mestery: ping | 20:17 |
| *** chuckC has joined #openstack-neutron | 20:17 | |
| mestery | kevinbenton: pong | 20:17 |
| *** nati_ueno has joined #openstack-neutron | 20:19 | |
| *** mriedem has left #openstack-neutron | 20:22 | |
| sweston | nati_ueno: ping | 20:22 |
| kevinbenton | mestery: oh, nevermind. i had a question about a blueprint but it was for devstack | 20:22 |
| nati_ueno | sweston: sorry buzy. can we talk later? | 20:22 |
| nati_ueno | sweston: could you give me your email? | 20:22 |
| sweston | nati_ueno: Yes, I just had some questions regardintg the bug I was working on, but it can wait. sweston@brocade.com | 20:23 |
| nati_ueno | sweston: sure | 20:23 |
| *** djoreilly has quit IRC | 20:24 | |
| kfox1111 | Is there no way to do a floating ip without having it on a router's gateway interface? | 20:24 |
| sweston | nati_ueno: ok, thank you ;-) | 20:24 |
| sweston | kfox1111: so, a floating ip is any address on an external network | 20:24 |
| sweston | kfox1111: you just need to mark the router as external, and any addresses assigned on that network will be floating | 20:25 |
| ctooley | I'm running Icehouse. Neutron seems to not be able to send notifications to nova. I'm getting ERROR neutron.notifiers.nova [-] Failed to notify nova on events: [{'status': 'completed', 'tag': u'da8f1f36-df7a-4d78-aa14-4dbfd263a0b2', 'name': 'network-vif-unplugged', 'server_uuid': u'652de88a-fc31-4071-9ff8-f6213efd1225'}] | 20:25 |
| kfox1111 | its also a nice abstraction to allow one to be allocated, and moved between vm's. | 20:25 |
| kfox1111 | hmm.. ok. I'll give that a try. thanks. | 20:26 |
| sweston | kfox1111: sure | 20:26 |
| kfox1111 | hmm... if I mark it as external, can I have regular vm's allocated directly on it too? | 20:27 |
| ctooley | This looks like the exception: BadRequest: object of type 'NoneType' has no len() (HTTP 400) | 20:27 |
| *** nati_uen_ has joined #openstack-neutron | 20:27 | |
| kfox1111 | or is that an either or situation? | 20:27 |
| kfox1111 | I guess that would be easy enough to test... | 20:28 |
| sweston | kfox1111: if you boot the vm up on that network as the primary nic netid, then I suppose you could call those regular vms? :-) | 20:28 |
| *** nati_ue__ has joined #openstack-neutron | 20:28 | |
| *** lukego has quit IRC | 20:29 | |
| kfox1111 | hmm.. at least in horizon, I can't select the external flagged network to launch the vm with. | 20:29 |
| sweston | kfox1111: the only difference with a external network, is that network must be realised outside of neutron. A floating ip address is meant to be shared among the tenants in your cloud | 20:29 |
| *** markmcclain has quit IRC | 20:29 | |
| openstackgerrit | Michael Smith proposed a change to openstack/neutron: Modify L3 Agent for Distributed Routers https://review.openstack.org/89413 | 20:30 |
| *** nati_ueno has quit IRC | 20:30 | |
| sweston | kfox1111: probably because you created the network as another tenant. external networks are frequently created with special access privilidges, only to be allocated and managed by a cloud admin | 20:30 |
| kfox1111 | in this case, I'd prefer it to be private.... but since the whole cloud is private, at least for the moment, that should be ok. | 20:30 |
| *** lukego has joined #openstack-neutron | 20:30 | |
| kfox1111 | ah. | 20:30 |
| *** amotoki_ has quit IRC | 20:32 | |
| *** lukego has quit IRC | 20:32 | |
| *** nati_uen_ has quit IRC | 20:32 | |
| kfox1111 | hmm... another question. Can you disable setting dhcp to issue a default gatway, but keep everything else? I'm thinking of having two external networks, each with floating ip's, and some vm's I want to allocate would be on both networks. It works now but route -n shows two default gateways. | 20:32 |
| *** singhs has joined #openstack-neutron | 20:33 | |
| sweston | kfox1111: you can (and should) disable dhcp on your external networks completely. If you are showing multiple external routes, then you probably have something else wrong ... but if you want to customize dhcp options, you may use the dns masq configuration options in your neutron.conf file | 20:36 |
| *** lukego has joined #openstack-neutron | 20:41 | |
| *** carl_baldwin_m has joined #openstack-neutron | 20:44 | |
| *** ctooley has left #openstack-neutron | 20:46 | |
| *** jecarey has quit IRC | 20:49 | |
| *** nati_ue__ has quit IRC | 20:49 | |
| *** jorgem has quit IRC | 20:53 | |
| *** erecio has quit IRC | 20:54 | |
| *** SumitNaiksatam has quit IRC | 20:55 | |
| *** otherwiseguy has joined #openstack-neutron | 20:55 | |
| *** WackoRobie has quit IRC | 20:56 | |
| *** SumitNaiksatam has joined #openstack-neutron | 20:56 | |
| *** SumitNaiksatam has quit IRC | 20:56 | |
| nlahouti | enikanorov: By API layer you mean the NeutronPluginBaseV2 ? | 20:57 |
| *** jecarey has joined #openstack-neutron | 20:58 | |
| *** amotoki_ has joined #openstack-neutron | 20:58 | |
| enikanorov_ | nlahouti: no, i mean the resource was not added by extension manager to a set of supported resources | 21:00 |
| enikanorov_ | and after that it doesn't have an implementing controller | 21:01 |
| enikanorov_ | in your case it seems that driver itself should be such controller, and not the plugin | 21:01 |
| *** carl_baldwin_m has quit IRC | 21:01 | |
| nlahouti | enikanorov: Yes I added it to the driver. | 21:01 |
| *** carl_baldwin_m has joined #openstack-neutron | 21:01 | |
| *** Sukhdev has joined #openstack-neutron | 21:01 | |
| nlahouti | enikanorov: I mean the mechainsm driver | 21:02 |
| *** singhs_ has joined #openstack-neutron | 21:04 | |
| *** manishg has quit IRC | 21:04 | |
| *** singhs has quit IRC | 21:04 | |
| *** singhs_ is now known as singhs | 21:04 | |
| openstackgerrit | Henry Gessau proposed a change to openstack/neutron-specs: ML2 Mechanism Driver for the Cisco APIC https://review.openstack.org/89305 | 21:05 |
| *** manishg has joined #openstack-neutron | 21:06 | |
| *** prad_ has joined #openstack-neutron | 21:08 | |
| *** nati_ueno has joined #openstack-neutron | 21:09 | |
| *** markmcclain has joined #openstack-neutron | 21:09 | |
| enikanorov_ | nlahouti: if so, then you wouldn't need additional code in base.py | 21:10 |
| *** julim_ has quit IRC | 21:13 | |
| *** carl_baldwin_m has quit IRC | 21:13 | |
| nlahouti | enikanorov: Do you mean when I have it in mechanism driver I don't need changes in base.py? | 21:18 |
| *** ale has joined #openstack-neutron | 21:19 | |
| *** BillTheKat has quit IRC | 21:19 | |
| *** SumitNaiksatam has joined #openstack-neutron | 21:21 | |
| enikanorov_ | nlahouti: look at how resource is created in any extension. it provides a controller for the resource (a class which has create_* update_* etc methods), usually it's a plugin, but it could be an instance of your driver class | 21:22 |
| *** ale_____ has quit IRC | 21:23 | |
| *** WackoRobie has joined #openstack-neutron | 21:27 | |
| *** chuckC has quit IRC | 21:27 | |
| *** blogan has quit IRC | 21:29 | |
| *** nati_ueno has quit IRC | 21:31 | |
| *** nati_ueno has joined #openstack-neutron | 21:31 | |
| *** WackoRobie has quit IRC | 21:32 | |
| sballe | enikanorov_, I have a quick question about Neutron LbaaS and the haproxy implementation. | 21:33 |
| sballe | Does the haproxy lbaas driver allow for haproxy to be setup on a separate VM? I am playign with Devstack so my ability to do so is limited | 21:34 |
| *** alexpilotti has joined #openstack-neutron | 21:34 | |
| *** openstackstatus has quit IRC | 21:35 | |
| *** WackoRobie has joined #openstack-neutron | 21:36 | |
| enikanorov_ | sballe: sure | 21:36 |
| *** openstackstatus has joined #openstack-neutron | 21:36 | |
| enikanorov_ | sballe: no, haproxy driver only works with lbaas agent which manages haproxy processes on the node | 21:37 |
| enikanorov_ | and not inside the vm | 21:37 |
| nlahouti | enikanorov: Sure I'll look again as I created the resources for the extensions and provided the controller. | 21:37 |
| enikanorov_ | so running it in devstack is totally fine | 21:37 |
| sballe | enikanorov_, so I can setup HA on a VM withing a tenant's private network. Cool! Where do the HA proxy VMs land? on a new VM within the tenants' network? | 21:39 |
| sballe | s/HA/LB | 21:39 |
| sballe | s/HA/HAproxy | 21:40 |
| enikanorov_ | sballe: haproxy is spawned on the node as a process, not in the VM | 21:40 |
| sballe | so on the Network node? | 21:40 |
| *** mlavalle has quit IRC | 21:41 | |
| enikanorov_ | sballe: on the node where lbaas agent runs. you can place the agent wherever you want :) | 21:41 |
| enikanorov_ | it can be compute | 21:41 |
| *** WackoRob_ has joined #openstack-neutron | 21:42 | |
| rm_work | sballe: in the current configuration, *all* HAProxy servers for every LB for every customer are all run on a single node, as processes, not in their own VMs :( | 21:42 |
| rm_work | I think you may have gotten that, but just making sure that's 100% clear | 21:42 |
| sballe | I did. :) | 21:43 |
| rm_work | so yeah, for a private cloud it is not too bad :P | 21:43 |
| rm_work | for public cloud, it is a bit… painful :( | 21:43 |
| sballe | rm_work, Even for a private cloud it is painful | 21:44 |
| rm_work | single-tenant cloud using a single server for all HAProxy configs isn't THAT bad I think? | 21:44 |
| rm_work | since most of the issues we have with that method are related to process isolation or scalability | 21:44 |
| enikanorov_ | rm_work: not correct | 21:44 |
| enikanorov_ | not a single node | 21:44 |
| sballe | Our private cloud use case is still multi-tenant | 21:44 |
| enikanorov_ | you can place lbaas agents on multiple nodes | 21:45 |
| rm_work | sballe: ah | 21:45 |
| enikanorov_ | that would distribute spawnd LBs across nodes | 21:45 |
| *** nati_ueno has quit IRC | 21:45 | |
| rm_work | enikanorov_: how does it schedule across multiple nodes? random? round-robin? some intelligence? | 21:45 |
| sballe | yeah that was my next question | 21:45 |
| *** WackoRobie has quit IRC | 21:46 | |
| enikanorov_ | rm_work: random | 21:46 |
| enikanorov_ | rm_work: it's configurable though, but changing the algorithm requires a piece of code, the default is random | 21:46 |
| rm_work | hmm k | 21:47 |
| *** WackoRob_ has quit IRC | 21:47 | |
| sballe | enikanorov_, so when you deploy your cloud you allocate X nodes for LBaaS? Since we are running Libra in VMs the concept is a little different | 21:47 |
| *** manishg has quit IRC | 21:47 | |
| sballe | I am assumign the X nodes will host many tenants' LBaaS? or as the LBaaS per tenant? I am assuming the former | 21:47 |
| enikanorov_ | sballe: that could be one strategy | 21:48 |
| enikanorov_ | yes, the former | 21:48 |
| sballe | What's another strategy? | 21:48 |
| rm_work | so, could you write a scheduler that would actually go to a different node per tenant? | 21:48 |
| rm_work | that would be… not a great answer to our problem, but it's interesting] | 21:49 |
| enikanorov_ | rm_work: the node doesn't matter as long as l2 connectivity for haproxy is provided | 21:50 |
| *** manishg has joined #openstack-neutron | 21:50 | |
| sballe | enikanorov_, so can you spin up a new VM on the tenat's private network and have ha-proxy/LB in there | 21:50 |
| sballe | enikanorov_, Anywhere this is documented? I am interested in udnerstanding this use case a little more | 21:52 |
| *** ale___ has joined #openstack-neutron | 21:52 | |
| *** ale has quit IRC | 21:52 | |
| *** suresh12 has quit IRC | 21:53 | |
| *** suresh12 has joined #openstack-neutron | 21:56 | |
| *** xmltok_ has joined #openstack-neutron | 21:58 | |
| enikanorov_ | sballe: right now there is no driver that supports haproxy in VM | 21:59 |
| sballe | so it is only HPproxy as a process on the physical node? | 21:59 |
| *** rkukura has left #openstack-neutron | 21:59 | |
| enikanorov_ | sballe: yes | 21:59 |
| *** kbringard has quit IRC | 22:00 | |
| xmltok_ | should icehouse+ml2+openvswitch support vlan tagged networking? | 22:00 |
| *** otherwiseguy has quit IRC | 22:00 | |
| sballe | enikanorov_, What is the typical deployment? lbaas-agent on the network nodes? on separate nodes? | 22:01 |
| amotoki_ | let's continue on -neutron. time is over.... | 22:01 |
| sballe | enikanorov_, what you have seen so far. | 22:01 |
| amotoki_ | wrong channell..... :-( | 22:01 |
| amotoki_ | xmltok_: what do you mean "vlan tagged"? | 22:02 |
| *** oda-g has joined #openstack-neutron | 22:02 | |
| amotoki_ | xmltok_: vlan tagged frame from VM? | 22:02 |
| *** banix has quit IRC | 22:02 | |
| enikanorov_ | sballe: typical is having lbaas agent on several network nodes | 22:03 |
| xmltok_ | amotoki: my neutron subnets need to be segregated using vlan tags. so i am using the provider:segmentation_id parameter | 22:03 |
| xmltok_ | network, not subnet | 22:03 |
| xmltok_ | it worked fine on my linuxbridge configuration in havana, but im building out new with icehouse and i want to get on the right plugins | 22:03 |
| amotoki_ | xmltok_: it should work. there is no spec change in icehouse. | 22:04 |
| sballe | enikanorov_, yeah that's what I thought. in this use case the network nodes could be bogged down by the LBaaS traffic | 22:04 |
| xmltok_ | amotoki_: i didnt use OVS in havana | 22:04 |
| sballe | enikanorov_, but good to know that you can load-balance the LB across several nodes | 22:04 |
| *** nati_ueno has joined #openstack-neutron | 22:05 | |
| enikanorov_ | sballe: with haproxy driver it is inevitable, even if you have separate nodes for lbaas | 22:05 |
| amotoki_ | xmltok_: ML2 with linux bridge should work. vlan segmetation should work both for linxu bridge and ovs. | 22:05 |
| *** networkstatic_zZ has quit IRC | 22:05 | |
| xmltok_ | i may try ML2 with linux bridge and see if that works for me, i'd like to use OVS but i suppose its not a requirement and it looks like its adding a lot of variables | 22:05 |
| enikanorov_ | sballe: because external traffic can only hit haproxies through floating ip, and hence is NATted from network nodes | 22:05 |
| sballe | enikanorov_, oh I see good point | 22:06 |
| *** pcm_ has quit IRC | 22:06 | |
| amotoki_ | xmltok_: hmm..... when I tested it during RC1, ml2 with ovs (vlan) worked though it was just a setup with two nodes. | 22:08 |
| xmltok_ | amotoki_: that gives me enough hope that i should be able to pull this off. right now i have dhcp going and i see its netns but i cant ping the ip from a remote node or locally | 22:09 |
| xmltok_ | i dont know how to verify that it has the right vlan with OVS though | 22:09 |
| amotoki_ | xmltok_: are you ping-ing from inside netns? If not, no ping is reachable to dhcp netns. | 22:10 |
| xmltok_ | if i ping out from the netns space i dont see icmps going through my OVS br-ex bridge, so its probably set up wrong | 22:10 |
| openstackgerrit | oda-g proposed a change to openstack/neutron: Make help texts more descriptive in Metaplugin https://review.openstack.org/88114 | 22:11 |
| xmltok_ | amotoki_: yeah, I am running this from 10.64.34.112: # ip netns exec qdhcp-d9edc84f-ecab-4d23-89f1-de8bfc72ba53 ping 10.64.34.112 and i get nothing back | 22:11 |
| amotoki_ | xmltok_: Is 10.64.34.112 assigned to what? dnsmasq? | 22:12 |
| *** leseb has quit IRC | 22:12 | |
| xmltok_ | .112 is the hypervisor, dnsmasq is on a different vlan with ip 10.64.40.101 | 22:12 |
| amotoki_ | xmltok_: vm? router? I am not sure your topology. | 22:12 |
| xmltok_ | is that making sense? | 22:13 |
| enikanorov_ | gtg, see everyone later, bye | 22:15 |
| *** samuelbercovici has joined #openstack-neutron | 22:16 | |
| *** nati_ueno has quit IRC | 22:23 | |
| *** mlavalle has joined #openstack-neutron | 22:25 | |
| *** jobewan has quit IRC | 22:31 | |
| sweston | markmcclain: ping | 22:32 |
| markmcclain | sweston: pong | 22:33 |
| sweston | markmcclain: Hi!! So there was a patch pushed on the nova side related to the bug I am working on, https://bugs.launchpad.net/neutron/+bug/1163569 | 22:34 |
| *** alexpilotti has quit IRC | 22:35 | |
| sweston | markmcclain: now it looks like the vm is being attached to a bridge, and I believe this will change how the vif is wired in the loadbalancer vip? | 22:35 |
| *** nati_ueno has joined #openstack-neutron | 22:37 | |
| *** jgrimm_ has quit IRC | 22:37 | |
| sweston | markmcclain: or maybe I don't care what's going on on the nova side, I should just ensure the bridge is created, and then attach to it? | 22:41 |
| *** networkstatic has joined #openstack-neutron | 22:42 | |
| *** thedodd has quit IRC | 22:42 | |
| markmcclain | for devices created by Neutron we plug directly into the bridge | 22:43 |
| markmcclain | and I in general we'd have to change the plugging code to inspect the use before creating the the device | 22:43 |
| markmcclain | might make sense to wait on the proposals to implement security groups in OVS | 22:44 |
| sweston | markmcclain: so here, a patch pushed to fix the nova vif ... https://review.openstack.org/#/c/83190/ | 22:45 |
| sweston | markmcclain: kinda changes how I wrote my wiring code for this bug ... | 22:45 |
| *** prad_ has quit IRC | 22:45 | |
| markmcclain | right it creates virtual hops | 22:46 |
| markmcclain | the neutron code adds the port directly to OVS | 22:46 |
| markmcclain | so we'd have to change how we plug | 22:46 |
| *** sacharya has quit IRC | 22:46 | |
| sweston | markmcclain: yup, so now I plug into ovs, and then attach ovs to the linux bridge created by nova, correct? | 22:47 |
| markmcclain | except VIPs never touch anything with nova | 22:48 |
| *** WackoRobie has joined #openstack-neutron | 22:49 | |
| sweston | markmcclain: ok, so then the attachment via the member pool is through another bridge? | 22:49 |
| markmcclain | the internal pool for contacting the pool | 22:50 |
| markmcclain | is through a device that is directly created in OVS | 22:51 |
| openstackgerrit | A change was merged to openstack/neutron: allow delete_port to work when there are multiple floating ips https://review.openstack.org/85432 | 22:51 |
| sweston | markmcclain: right now it is, yes ... but in order to implement this bug fix I have to create a linux bridge to connect it, but I would use a different bridge than the port that nova uses to attach to the integration bridge? | 22:52 |
| markmcclain | ok.. just wondering now that there is a team working on OVS based SG implementation | 22:53 |
| *** samuelbercovici has quit IRC | 22:54 | |
| markmcclain | if the hybrid approach for VIPs in Juno makes sense | 22:54 |
| markmcclain | we will not be able to backport to Icehouse | 22:54 |
| sweston | markmcclain: I hope not ... 'cause I've been working on this patch pretty seriously ... | 22:54 |
| markmcclain | ok understood, but with that native flow based security groups, hybrid sg implementations are likely to be on borrowed time (even for the Nova) | 22:55 |
| markmcclain | might ahve to discuss at the summit | 22:56 |
| markmcclain | or via the mailing list w/larger community | 22:56 |
| sweston | markmcclain: so ... I'm not sure what to do, should I continue working on this patch or not?? | 22:57 |
| sweston | markmcclain: what is the alternative to hybrid sgs? | 22:57 |
| markmcclain | the alternative is native OVS implementations | 22:57 |
| markmcclain | I know there are teams working on it, so seems like we need to sync with those teams | 22:58 |
| sweston | markmcclain: via flow rules? | 22:58 |
| markmcclain | mestery can connect you with the right people | 22:58 |
| sweston | markmcclain: ok, I can send him an email then and ask. | 22:59 |
| sweston | markmcclain: thank you :-) | 22:59 |
| markmcclain | sweston: you're welcome.. sorry I don't have a clearer answer | 23:00 |
| sweston | markmcclain: that's okay, I always appreciate any help you have to offer. | 23:01 |
| *** Sukhdev has quit IRC | 23:01 | |
| *** SridharG has quit IRC | 23:02 | |
| adam_g | what are the router and interface requirements for firewall creation? i can't seem to get one out of PENDING_CREATE, though the tenant has active net/subnet/router/router interface | 23:05 |
| *** BillTheKat has joined #openstack-neutron | 23:06 | |
| *** sweston has quit IRC | 23:06 | |
| *** dvorkinista has quit IRC | 23:07 | |
| openstackgerrit | Kevin Benton proposed a change to openstack/neutron-specs: Add spec for neutron-external-ports blueprint https://review.openstack.org/87825 | 23:07 |
| *** BuSerD has quit IRC | 23:09 | |
| *** BillTheKat has quit IRC | 23:10 | |
| *** SumitNaiksatam has quit IRC | 23:12 | |
| *** beagles has quit IRC | 23:13 | |
| *** markmcclain has quit IRC | 23:17 | |
| *** SumitNaiksatam has joined #openstack-neutron | 23:18 | |
| *** singhs has quit IRC | 23:33 | |
| juice | is anyone around here at the moment? | 23:41 |
| *** SumitNaiksatam has quit IRC | 23:42 | |
| *** dvorkinista has joined #openstack-neutron | 23:45 | |
| *** SumitNaiksatam has joined #openstack-neutron | 23:49 | |
| *** openstackgerrit has quit IRC | 23:50 | |
| * jamespd is around | 23:50 | |
| *** openstackgerrit has joined #openstack-neutron | 23:50 | |
| juice | hi jamespd | 23:51 |
| jamespd | hi juice | 23:51 |
| juice | quick question possibly a bug and to be upfront I went ahead and cross posted in the mail list | 23:51 |
| juice | to be brief: one nova instance; 2 NICs; only one NIC gets configured | 23:52 |
| juice | should nova attach-interface automatically configure the 2nd interface on the Instance | 23:52 |
| juice | it seems odd that once this command is done, one has to go on the instance, configure the interface and bring it up manually | 23:53 |
| *** manishg has quit IRC | 23:53 | |
| jamespd | hmmm. | 23:54 |
| * jamespd goes to test that | 23:57 | |
| *** chuckC has joined #openstack-neutron | 23:57 | |
| jamespd | juice: Is the second interface useable if you configure it manually? | 23:58 |
| juice | yes | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!