| *** sdague has joined #openstack-neutron | 00:02 | |
| *** limao has quit IRC | 00:04 | |
| *** kfox1111 has joined #openstack-neutron | 00:05 | |
| kfox1111 | cli question. | 00:05 |
|---|---|---|
| kfox1111 | with a freshly built cloud, | 00:05 |
| kfox1111 | whats the best way to get the 'default' security group id in the 'admin' tenant? | 00:06 |
| kfox1111 | I use to just do a neutron security-group-rule-create default, but in ocata it complains there are more then one. | 00:07 |
| *** pramodrj07 has quit IRC | 00:07 | |
| *** Swami_ has joined #openstack-neutron | 00:08 | |
| dasm | kfox1111: what do you see for: neutron security-group-rule-list | 00:08 |
| kfox1111 | I'm doing it in gate, so I can get that, but wil take me aabout 20-30 min. :/ | 00:08 |
| dasm | ack | 00:09 |
| *** Swami has quit IRC | 00:10 | |
| *** tbachman has quit IRC | 00:11 | |
| *** nicolasbock has quit IRC | 00:12 | |
| *** princenana has joined #openstack-neutron | 00:17 | |
| dasm | armax: kevinbenton: hey. could you look at this patch: https://review.openstack.org/#/c/424906 it solves an issue with broken functional tests | 00:17 |
| *** Trident has joined #openstack-neutron | 00:21 | |
| *** Guest9757 is now known as rook | 00:21 | |
| *** tlian2 has joined #openstack-neutron | 00:21 | |
| *** masaki has joined #openstack-neutron | 00:24 | |
| *** tlian has quit IRC | 00:24 | |
| *** Trident has quit IRC | 00:24 | |
| *** Trident has joined #openstack-neutron | 00:26 | |
| *** yamahata has quit IRC | 00:27 | |
| *** mlavalle has quit IRC | 00:27 | |
| *** nicolasbock has joined #openstack-neutron | 00:28 | |
| *** Trident has quit IRC | 00:28 | |
| *** Trident has joined #openstack-neutron | 00:29 | |
| *** radhikam has quit IRC | 00:30 | |
| *** nicolasbock has quit IRC | 00:33 | |
| *** iyamahat has quit IRC | 00:34 | |
| *** wolverineav has quit IRC | 00:34 | |
| *** wolverineav has joined #openstack-neutron | 00:36 | |
| *** princenana has quit IRC | 00:39 | |
| *** hoangcx has joined #openstack-neutron | 00:39 | |
| *** princenana has joined #openstack-neutron | 00:39 | |
| *** wolverineav has quit IRC | 00:41 | |
| *** panda is now known as panda|zZ | 00:42 | |
| *** thorst_ has joined #openstack-neutron | 00:42 | |
| *** zhurong has joined #openstack-neutron | 00:44 | |
| *** neiljerram has quit IRC | 00:45 | |
| *** limao has joined #openstack-neutron | 00:46 | |
| *** nicolasbock has joined #openstack-neutron | 00:49 | |
| *** amotoki has joined #openstack-neutron | 00:49 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/neutron: Updated from global requirements https://review.openstack.org/423645 | 00:50 |
| *** artom has left #openstack-neutron | 00:52 | |
| *** thorst_ has quit IRC | 00:58 | |
| *** sdague has quit IRC | 01:01 | |
| *** jose-phillips has joined #openstack-neutron | 01:01 | |
| jose-phillips | hey | 01:01 |
| jose-phillips | a quick help | 01:01 |
| *** princenana has quit IRC | 01:01 | |
| jose-phillips | is possible unplug the network card to all instances and plug it back? | 01:01 |
| *** princenana has joined #openstack-neutron | 01:02 | |
| *** namnh has joined #openstack-neutron | 01:06 | |
| *** rajinir has quit IRC | 01:06 | |
| *** tbachman has joined #openstack-neutron | 01:06 | |
| *** tbachman has quit IRC | 01:08 | |
| *** tbachman has joined #openstack-neutron | 01:09 | |
| *** princenana has quit IRC | 01:09 | |
| kfox1111 | dasm: http://logs.openstack.org/50/418550/52/check/gate-kolla-kubernetes-deploy-ubuntu-binary-2-iscsi-nv/7ebe557/console.html#_2017-01-25_00_52_38_991131 | 01:11 |
| *** ventifus has quit IRC | 01:11 | |
| kfox1111 | weirdly, the openstack security group list -f value | 01:11 |
| kfox1111 | doesn't show the tenant id. | 01:11 |
| kfox1111 | but on another one: | 01:11 |
| *** iwamoto has joined #openstack-neutron | 01:12 | |
| kfox1111 | openstack security group list -f value has an extra field of tenant id at the end. | 01:12 |
| *** princenana has joined #openstack-neutron | 01:18 | |
| *** abhiraut has quit IRC | 01:24 | |
| *** yedongcan has joined #openstack-neutron | 01:25 | |
| openstackgerrit | Swaminathan Vasudevan proposed openstack/neutron: (WIP): DVR: Server side patch to schedule unbound port Floating IP https://review.openstack.org/320669 | 01:28 |
| *** princenana has quit IRC | 01:32 | |
| *** tbachman_ has joined #openstack-neutron | 01:33 | |
| *** wolverineav has joined #openstack-neutron | 01:33 | |
| *** tbachman has quit IRC | 01:34 | |
| *** tbachman_ is now known as tbachman | 01:34 | |
| *** yedongcan1 has joined #openstack-neutron | 01:37 | |
| *** yedongcan has quit IRC | 01:41 | |
| *** John341__ has quit IRC | 01:43 | |
| *** Sukhdev has quit IRC | 01:45 | |
| *** John341_ has joined #openstack-neutron | 01:45 | |
| *** iranzo has quit IRC | 01:47 | |
| openstackgerrit | Swaminathan Vasudevan proposed openstack/neutron: (WIP) DVR: Configure unbound floatingip ports to snat_namespace. https://review.openstack.org/323618 | 01:47 |
| *** john-davidge has joined #openstack-neutron | 01:48 | |
| *** kevo has quit IRC | 01:48 | |
| *** jose-phillips has quit IRC | 01:49 | |
| *** sterdnotshaken2 has quit IRC | 01:49 | |
| *** john-davidge has quit IRC | 01:52 | |
| *** gvrangan has quit IRC | 01:55 | |
| *** jose-phillips has joined #openstack-neutron | 01:56 | |
| *** Swami_ has quit IRC | 02:00 | |
| *** jamesden_ has joined #openstack-neutron | 02:04 | |
| amotoki | kfox1111: with OSC, you need to specify --long option at the same time if you want to show non-default fields. | 02:10 |
| amotoki | kfox1111: but unfortunately --long option does not exist...... | 02:12 |
| amotoki | kfox1111: this incompatbility needs to be fixed | 02:12 |
| *** yedongcan has joined #openstack-neutron | 02:13 | |
| *** yedongcan1 has quit IRC | 02:15 | |
| openstackgerrit | IWAMOTO Toshihiro proposed openstack/neutron: ovsfw: Cleanup *_port_filter methods https://review.openstack.org/421697 | 02:17 |
| openstackgerrit | IWAMOTO Toshihiro proposed openstack/neutron: ovsfw: Refresh OFPort when necessary https://review.openstack.org/404564 | 02:17 |
| *** jamesden_ has quit IRC | 02:24 | |
| *** wolverineav has quit IRC | 02:25 | |
| *** wolverineav has joined #openstack-neutron | 02:25 | |
| *** wolverineav has quit IRC | 02:25 | |
| *** rkukura has joined #openstack-neutron | 02:27 | |
| *** armax has quit IRC | 02:31 | |
| *** tbachman has quit IRC | 02:34 | |
| *** fzdarsky_ has joined #openstack-neutron | 02:39 | |
| *** wolverineav has joined #openstack-neutron | 02:41 | |
| *** fzdarsky has quit IRC | 02:42 | |
| *** jamesdenton has joined #openstack-neutron | 02:43 | |
| *** wolverineav has quit IRC | 02:45 | |
| kfox1111 | http://logs.openstack.org/50/418550/53/check/gate-kolla-kubernetes-deploy-centos-binary-t-ceph-multi-nv/1e45843/console.html#_2017-01-25_01_43_21_788200 | 02:46 |
| kfox1111 | dasm: amotoki: got a log. | 02:46 |
| *** sudipto has joined #openstack-neutron | 02:47 | |
| *** sudipto_ has joined #openstack-neutron | 02:47 | |
| kfox1111 | looks like there are two default security groups. | 02:47 |
| kfox1111 | not really sure how to get current tenants one out though. | 02:47 |
| kfox1111 | amotoki: so --long loads more fields? | 02:48 |
| kfox1111 | oh. so it doesn't exist though? | 02:49 |
| kfox1111 | ok..... | 02:49 |
| kfox1111 | so, how do you edit the default securiyt group for your tenant? | 02:49 |
| amotoki | kfox1111: --long option shows more fields. | 02:49 |
| amotoki | kfox1111: what do you mean? I am not sure what you want to do. | 02:50 |
| kfox1111 | amotoki: so, in that log, I'm trying to add some rules to the currently selecgted tenant's default security group via the cli. | 02:51 |
| kfox1111 | I tried just specifying "default" as the security group name, | 02:51 |
| kfox1111 | but it fails due to multiple "default" sg's being found. | 02:51 |
| kfox1111 | so, I need some way to figure out the current tenants "default" security group id. | 02:51 |
| kfox1111 | I tried: | 02:52 |
| kfox1111 | openstack security group list -f value -c ID -c Project -c Name | 02:52 |
| kfox1111 | and on my mitaka cloud, it works. | 02:52 |
| *** s3wong has quit IRC | 02:52 | |
| kfox1111 | but I think its an older OSC client. | 02:52 |
| kfox1111 | cause when I try with a newer osc, it seems to skip Project. | 02:53 |
| amotoki | kfox1111: am running devstack now. | 02:53 |
| kfox1111 | thx. | 02:53 |
| *** TuHV has joined #openstack-neutron | 02:54 | |
| *** TuHV_ has joined #openstack-neutron | 02:54 | |
| openstackgerrit | ZhaoBo proposed openstack/neutron-lib: Add check for quering neutron core resources with standardattr sort-key https://review.openstack.org/424948 | 02:55 |
| *** tlian2 has quit IRC | 02:56 | |
| openstackgerrit | ZhaoBo proposed openstack/neutron: Support query with sort-key from standardattr https://review.openstack.org/424949 | 02:57 |
| *** thorst_ has joined #openstack-neutron | 02:57 | |
| *** thorst_ has quit IRC | 02:57 | |
| *** rkukura has quit IRC | 03:00 | |
| *** radhikam has joined #openstack-neutron | 03:02 | |
| *** wolverineav has joined #openstack-neutron | 03:02 | |
| *** tommylikehu has quit IRC | 03:05 | |
| *** iyamahat has joined #openstack-neutron | 03:05 | |
| *** chandanc_ has joined #openstack-neutron | 03:05 | |
| *** wolverineav has quit IRC | 03:07 | |
| openstackgerrit | IWAMOTO Toshihiro proposed openstack/neutron: Mark of_interface option deprecated https://review.openstack.org/424953 | 03:10 |
| *** mriedem has joined #openstack-neutron | 03:11 | |
| *** iyamahat has quit IRC | 03:11 | |
| *** iyamahat has joined #openstack-neutron | 03:11 | |
| *** wolverineav has joined #openstack-neutron | 03:13 | |
| *** iyamahat_ has joined #openstack-neutron | 03:14 | |
| *** iyamahat_ has quit IRC | 03:15 | |
| *** iyamahat_ has joined #openstack-neutron | 03:16 | |
| *** iyamahat has quit IRC | 03:17 | |
| *** wolverineav has quit IRC | 03:17 | |
| *** jamesdenton has quit IRC | 03:20 | |
| *** pbandark has joined #openstack-neutron | 03:21 | |
| *** ijw has quit IRC | 03:21 | |
| *** ihrachys has joined #openstack-neutron | 03:23 | |
| *** wolverineav has joined #openstack-neutron | 03:23 | |
| *** nmathew has joined #openstack-neutron | 03:23 | |
| *** wolverineav has quit IRC | 03:27 | |
| *** hfu has joined #openstack-neutron | 03:32 | |
| *** eezhova has joined #openstack-neutron | 03:32 | |
| openstackgerrit | YAMAMOTO Takashi proposed openstack/neutron: Add a tempest scenario for floating-ip https://review.openstack.org/424043 | 03:32 |
| openstackgerrit | YAMAMOTO Takashi proposed openstack/neutron: test_floatingip: Add a case for SRC without FIP https://review.openstack.org/424959 | 03:32 |
| openstackgerrit | YAMAMOTO Takashi proposed openstack/neutron: test_floatingip: Add "same server" case https://review.openstack.org/424960 | 03:32 |
| *** udesale has joined #openstack-neutron | 03:33 | |
| *** ihrachys has quit IRC | 03:37 | |
| *** gcb has joined #openstack-neutron | 03:38 | |
| *** ijw has joined #openstack-neutron | 03:42 | |
| *** ijw has quit IRC | 03:46 | |
| *** john-davidge has joined #openstack-neutron | 03:48 | |
| *** anilvenkata has joined #openstack-neutron | 03:50 | |
| *** wolverineav has joined #openstack-neutron | 03:52 | |
| amotoki | kfox1111: this is a result of my side (with the latest OSC/SDK/neutronclient) | 03:52 |
| *** john-davidge has quit IRC | 03:53 | |
| amotoki | kfox1111: I see project information in OSC output regardless of user and/or role. | 03:53 |
| amotoki | kfox1111: http://paste.openstack.org/show/596347/ | 03:56 |
| *** wolverineav has quit IRC | 03:57 | |
| *** psahoo has joined #openstack-neutron | 03:58 | |
| *** mdnadeem has joined #openstack-neutron | 03:59 | |
| *** marst has joined #openstack-neutron | 04:01 | |
| *** wolverineav has joined #openstack-neutron | 04:02 | |
| *** wolverineav has quit IRC | 04:07 | |
| *** armax has joined #openstack-neutron | 04:10 | |
| *** nicolasbock has quit IRC | 04:16 | |
| *** sputnik13 has joined #openstack-neutron | 04:21 | |
| *** wolverineav has joined #openstack-neutron | 04:21 | |
| *** eezhova has quit IRC | 04:22 | |
| *** radhikam has quit IRC | 04:25 | |
| *** tommylikehu_ has joined #openstack-neutron | 04:26 | |
| *** wolverineav has quit IRC | 04:26 | |
| *** jhershbe_ has joined #openstack-neutron | 04:28 | |
| *** dave-mccowan has quit IRC | 04:30 | |
| *** tommylikehu_ has quit IRC | 04:30 | |
| *** armax has quit IRC | 04:30 | |
| *** wolverineav has joined #openstack-neutron | 04:31 | |
| *** mriedem has quit IRC | 04:32 | |
| *** wolverineav has quit IRC | 04:36 | |
| *** Sukhdev has joined #openstack-neutron | 04:37 | |
| *** jhershbe_ has quit IRC | 04:39 | |
| *** jose-phillips has quit IRC | 04:40 | |
| *** wolverineav has joined #openstack-neutron | 04:41 | |
| *** gcb has quit IRC | 04:43 | |
| *** gcb has joined #openstack-neutron | 04:44 | |
| *** ayogi has joined #openstack-neutron | 04:45 | |
| *** wolverineav has quit IRC | 04:46 | |
| *** gkadam has joined #openstack-neutron | 04:46 | |
| *** dewanee has quit IRC | 04:52 | |
| *** armax has joined #openstack-neutron | 04:55 | |
| *** yamahata has joined #openstack-neutron | 04:56 | |
| *** armax has quit IRC | 04:56 | |
| *** thorst_ has joined #openstack-neutron | 04:59 | |
| *** thorst_ has quit IRC | 05:03 | |
| *** jamielennox is now known as jamielennox|away | 05:06 | |
| *** mkolesni has joined #openstack-neutron | 05:06 | |
| openstackgerrit | zhangyanxian proposed openstack/neutron: Fix typo in test_l2_lb_agent.py https://review.openstack.org/423727 | 05:10 |
| *** wolverineav has joined #openstack-neutron | 05:11 | |
| *** mkolesni has quit IRC | 05:15 | |
| *** wolverineav has quit IRC | 05:15 | |
| *** mkolesni has joined #openstack-neutron | 05:16 | |
| *** udesale__ has joined #openstack-neutron | 05:17 | |
| *** udesale has quit IRC | 05:17 | |
| *** nagarjung has joined #openstack-neutron | 05:18 | |
| *** gongysh has joined #openstack-neutron | 05:18 | |
| *** ratailor has joined #openstack-neutron | 05:18 | |
| *** udesale has joined #openstack-neutron | 05:19 | |
| *** udesale__ has quit IRC | 05:21 | |
| *** eezhova has joined #openstack-neutron | 05:29 | |
| *** s3wong has joined #openstack-neutron | 05:29 | |
| *** janki has joined #openstack-neutron | 05:32 | |
| *** shausy has joined #openstack-neutron | 05:34 | |
| *** wolverineav has joined #openstack-neutron | 05:36 | |
| *** wolverineav has quit IRC | 05:41 | |
| *** radhikam has joined #openstack-neutron | 05:41 | |
| *** radhikam has quit IRC | 05:45 | |
| *** sudswas__ has joined #openstack-neutron | 05:47 | |
| *** john-davidge has joined #openstack-neutron | 05:49 | |
| *** sudipto has quit IRC | 05:50 | |
| *** sudipto_ has quit IRC | 05:50 | |
| *** sudipto has joined #openstack-neutron | 05:50 | |
| *** jchhatbar has joined #openstack-neutron | 05:50 | |
| *** janki has quit IRC | 05:51 | |
| openstackgerrit | Aradhana Singh proposed openstack/neutron: [WIP] Extend Quota API to report usage statistics https://review.openstack.org/383673 | 05:53 |
| *** john-davidge has quit IRC | 05:54 | |
| *** abregman has joined #openstack-neutron | 05:55 | |
| *** wolverineav has joined #openstack-neutron | 05:56 | |
| *** kevo has joined #openstack-neutron | 05:58 | |
| *** wolverineav has quit IRC | 06:00 | |
| *** janki has joined #openstack-neutron | 06:02 | |
| *** jchhatbar has quit IRC | 06:04 | |
| *** wolverineav has joined #openstack-neutron | 06:06 | |
| *** nyechiel_ has joined #openstack-neutron | 06:06 | |
| *** pgadiya has joined #openstack-neutron | 06:10 | |
| *** wolverineav has quit IRC | 06:10 | |
| *** irenab_ has joined #openstack-neutron | 06:11 | |
| *** irenab_ has quit IRC | 06:12 | |
| *** armax has joined #openstack-neutron | 06:13 | |
| *** nyechiel_ has quit IRC | 06:14 | |
| *** armax has quit IRC | 06:16 | |
| *** jpena|off has quit IRC | 06:18 | |
| *** hichihara has joined #openstack-neutron | 06:19 | |
| *** huanxie has joined #openstack-neutron | 06:19 | |
| *** jpena|off has joined #openstack-neutron | 06:20 | |
| *** wolverineav has joined #openstack-neutron | 06:26 | |
| *** mkolesni has quit IRC | 06:26 | |
| *** mkolesni has joined #openstack-neutron | 06:26 | |
| *** jchhatbar has joined #openstack-neutron | 06:26 | |
| *** horms has joined #openstack-neutron | 06:28 | |
| *** janki has quit IRC | 06:29 | |
| *** pgadiya has quit IRC | 06:30 | |
| *** wolverineav has quit IRC | 06:30 | |
| *** dsneddon has quit IRC | 06:31 | |
| *** s3wong has quit IRC | 06:31 | |
| *** anilvenkata has quit IRC | 06:32 | |
| *** ekuris_ has joined #openstack-neutron | 06:34 | |
| *** wolverineav has joined #openstack-neutron | 06:36 | |
| *** chandanc_ has quit IRC | 06:37 | |
| *** Sukhdev has quit IRC | 06:39 | |
| *** wolverineav has quit IRC | 06:40 | |
| *** horms has quit IRC | 06:40 | |
| *** belharar_ has joined #openstack-neutron | 06:42 | |
| *** pgadiya has joined #openstack-neutron | 06:42 | |
| *** s3wong has joined #openstack-neutron | 06:43 | |
| *** anilvenkata has joined #openstack-neutron | 06:44 | |
| *** belharar__ has joined #openstack-neutron | 06:44 | |
| *** gkadam is now known as gkadam|lunch | 06:47 | |
| *** jchhatbar is now known as janki | 06:47 | |
| *** Jeffrey4l_ has quit IRC | 06:48 | |
| *** ratailor is now known as ratailor|afk | 06:49 | |
| *** Jeffrey4l has joined #openstack-neutron | 06:53 | |
| *** trinaths has joined #openstack-neutron | 06:54 | |
| *** lihi has joined #openstack-neutron | 06:54 | |
| *** udesale__ has joined #openstack-neutron | 06:56 | |
| *** hoangcx_ has joined #openstack-neutron | 06:57 | |
| *** s3wong has quit IRC | 06:57 | |
| *** belharar_ has quit IRC | 06:58 | |
| *** belharar__ has quit IRC | 06:58 | |
| *** belharar_ has joined #openstack-neutron | 06:58 | |
| *** thorst_ has joined #openstack-neutron | 06:59 | |
| openstackgerrit | YAMAMOTO Takashi proposed openstack/neutron: test_floatingip: Add a case for SRC without FIP https://review.openstack.org/424959 | 07:00 |
| openstackgerrit | YAMAMOTO Takashi proposed openstack/neutron: test_floatingip: Add "same server" case https://review.openstack.org/424960 | 07:00 |
| openstackgerrit | YAMAMOTO Takashi proposed openstack/neutron: Add a tempest scenario for floating-ip https://review.openstack.org/424043 | 07:00 |
| *** hoangcx has quit IRC | 07:00 | |
| *** nyechiel_ has joined #openstack-neutron | 07:03 | |
| *** belharar_ has quit IRC | 07:04 | |
| *** thorst_ has quit IRC | 07:04 | |
| *** belharar_ has joined #openstack-neutron | 07:04 | |
| *** namnh has quit IRC | 07:06 | |
| kevinbenton | anilvenkata: ping | 07:06 |
| *** aojea has quit IRC | 07:06 | |
| *** sridharg has joined #openstack-neutron | 07:09 | |
| *** nmathew- has joined #openstack-neutron | 07:09 | |
| *** mkolesni has quit IRC | 07:10 | |
| *** mkolesni has joined #openstack-neutron | 07:10 | |
| *** nmathew has quit IRC | 07:11 | |
| *** nmathew has joined #openstack-neutron | 07:12 | |
| *** sudipto_ has joined #openstack-neutron | 07:13 | |
| *** belharar_ has quit IRC | 07:14 | |
| *** sudipto has quit IRC | 07:14 | |
| *** jprovazn has joined #openstack-neutron | 07:14 | |
| *** nmathew- has quit IRC | 07:14 | |
| *** sudswas__ has quit IRC | 07:14 | |
| *** sudipto has joined #openstack-neutron | 07:14 | |
| *** ratailor|afk is now known as ratailor | 07:14 | |
| *** matrohon has joined #openstack-neutron | 07:16 | |
| *** yamamoto has quit IRC | 07:17 | |
| *** anilvenkata has quit IRC | 07:17 | |
| *** aojea has joined #openstack-neutron | 07:19 | |
| *** adriant has quit IRC | 07:19 | |
| *** matrohon has quit IRC | 07:23 | |
| *** lihi has quit IRC | 07:25 | |
| *** jprovazn has quit IRC | 07:26 | |
| *** jprovazn has joined #openstack-neutron | 07:26 | |
| *** lihi has joined #openstack-neutron | 07:27 | |
| *** pcaruana has joined #openstack-neutron | 07:28 | |
| openstackgerrit | Hirofumi Ichihara proposed openstack/neutron: Enhance tag mechanism https://review.openstack.org/413662 | 07:28 |
| openstackgerrit | Ann Taraday proposed openstack/neutron: Check arg type for SegmentTypeDriver functions https://review.openstack.org/425023 | 07:29 |
| *** hfu has quit IRC | 07:29 | |
| *** anilvenkata has joined #openstack-neutron | 07:30 | |
| *** belharar_ has joined #openstack-neutron | 07:30 | |
| anilvenkata | kevinbenton, pong | 07:30 |
| kevinbenton | anilvenkata: currently when an OVS agent starts up, it must set its ports statuses to BUILD and then back to ACTIVE in order to get the l2pop mechanism driver to flood entries to it, right? | 07:31 |
| *** iranzo has joined #openstack-neutron | 07:32 | |
| anilvenkata | kevinbenton, yes | 07:32 |
| *** nmathew- has joined #openstack-neutron | 07:33 | |
| kevinbenton | anilvenkata: ok. i'm working on the push notification stuff and i'm trying to decide how to maintain compatibility with that | 07:34 |
| anilvenkata | kevinbenton, ok | 07:34 |
| kevinbenton | anilvenkata: i may just have to have an explicit RPC call to set status to BUILD | 07:34 |
| kevinbenton | anilvenkata: because the agent no longer calls get_device_details | 07:34 |
| *** tesseract has joined #openstack-neutron | 07:34 | |
| *** john-davidge has joined #openstack-neutron | 07:35 | |
| *** nmathew has quit IRC | 07:35 | |
| anilvenkata | kevinbenton, this is the check https://github.com/openstack/neutron/blob/master/neutron/plugins/ml2/drivers/l2pop/mech_driver.py#L177 | 07:36 |
| *** andreas_s has joined #openstack-neutron | 07:37 | |
| anilvenkata | kevinbenton, may be if we are transitioning from DOWN to ACTIVE also this can work | 07:37 |
| kevinbenton | anilvenkata: thanks. that's what I figured. so it's not that it needs to be BUILD | 07:37 |
| kevinbenton | anilvenkata: right | 07:37 |
| anilvenkata | kevinbenton, so instead of BUILD to ACTIVE | 07:37 |
| kevinbenton | anilvenkata: it just needs to go from something else to ACTIVE | 07:37 |
| anilvenkata | kevinbenton, yes :) | 07:37 |
| kevinbenton | anilvenkata: do other agents depend on this status change? | 07:38 |
| kevinbenton | anilvenkata: like if agentA comes online and does the status change with the ports | 07:39 |
| kevinbenton | anilvenkata: does agentB also receive data and react to it? | 07:39 |
| anilvenkata | kevinbenton, no | 07:39 |
| kevinbenton | anilvenkata: for example, if I do have a way for the new agent to get its forwarding data without flipping the port status, would that break other agents? | 07:40 |
| *** john-davidge has quit IRC | 07:40 | |
| *** teclator has joined #openstack-neutron | 07:41 | |
| anilvenkata | kevinbenton, when agents get port update notification, they will process that port only if that agent is hosting that port right, otherwise they ignore the port | 07:41 |
| kevinbenton | anilvenkata: right, but they setup forwarding entries to the other agents | 07:42 |
| *** jhershbe_ has joined #openstack-neutron | 07:42 | |
| anilvenkata | kevinbenton, forwarding entries will be taken care by l2pop | 07:42 |
| kevinbenton | anilvenkata: does that happen on port create/update with this function? https://github.com/openstack/neutron/blob/master/neutron/plugins/ml2/drivers/l2pop/mech_driver.py#L103 | 07:43 |
| kevinbenton | anilvenkata: or is that also just after the port first goes to ACTIVE here: https://github.com/openstack/neutron/blob/master/neutron/plugins/ml2/drivers/l2pop/mech_driver.py#L178-L179 | 07:44 |
| *** hoangcx has joined #openstack-neutron | 07:46 | |
| anilvenkata | kevinbenton, fixed_ips_changed wont be processed when update_port_status is called | 07:46 |
| *** vikram has joined #openstack-neutron | 07:47 | |
| anilvenkata | kevinbenton, as u know update_port_status wont change fixed_ip annd will change port status | 07:47 |
| openstackgerrit | Ha Van Tu proposed openstack/neutron-fwaas: Netlink solution to improve FWaaS performance https://review.openstack.org/389654 | 07:47 |
| kevinbenton | anilvenkata: right. what i mean is when do other agents get their forwarding entries for a port that belongs to a given agent? | 07:48 |
| *** moshele has joined #openstack-neutron | 07:48 | |
| kevinbenton | anilvenkata: agentA and agentB have ports on networkX | 07:48 |
| kevinbenton | anilvenkata: a new vm is booted on agentC on networkX | 07:48 |
| anilvenkata | kevinbenton, ok, let me explain this case | 07:48 |
| kevinbenton | anilvenkata: at which point to agentA and agentB setup their tunnels to agentC | 07:48 |
| *** hoangcx_ has quit IRC | 07:48 | |
| *** korzen has joined #openstack-neutron | 07:49 | |
| anilvenkata | kevinbenton, agentC calls 1) get_device_details and then 2) update_port_status(ACTIVE), | 07:49 |
| *** Miouge has joined #openstack-neutron | 07:49 | |
| anilvenkata | kevinbenton, for this new port | 07:49 |
| anilvenkata | kevinbenton, so when agentC calls update_port_status(ACTIVE), L178-179 will send this new port FDB to agentA and agentB | 07:50 |
| anilvenkata | kevinbenton, this is existing behaviour | 07:50 |
| kevinbenton | anilvenkata: ok, so the other agents get the entries on the update to active | 07:51 |
| *** Miouge has quit IRC | 07:51 | |
| *** panda|zZ is now known as panda | 07:51 | |
| anilvenkata | kevinbenton, one morething, before agentC calls get_device_details(which sets status to BUILD), port status will be in ACTIVE state(and not DOWN) | 07:51 |
| *** Miouge has joined #openstack-neutron | 07:52 | |
| kevinbenton | anilvenkata: on an initial port create it should be DOWN | 07:52 |
| kevinbenton | anilvenkata: (so this floods to all hosts then https://github.com/openstack/neutron/blob/master/neutron/plugins/ml2/drivers/l2pop/mech_driver.py#L300-L301) | 07:52 |
| kevinbenton | right? | 07:52 |
| anilvenkata | kevinbenton, L300-301 will be executed only on transition from BUILD to ACTIVE | 07:53 |
| *** lennyb has joined #openstack-neutron | 07:53 | |
| kevinbenton | anilvenkata: DOWN to ACTIVE | 07:54 |
| kevinbenton | anilvenkata: should also work | 07:54 |
| kevinbenton | anilvenkata: right? | 07:54 |
| *** obondarev has joined #openstack-neutron | 07:54 | |
| anilvenkata | kevinbenton, yes, but as I said | 07:54 |
| *** mkolesni has quit IRC | 07:54 | |
| anilvenkata | kevinbenton, jeust need to confirm initial port status before the agent calls update_port_status(active) is down | 07:55 |
| *** mkolesni has joined #openstack-neutron | 07:55 | |
| kevinbenton | anilvenkata: right | 07:55 |
| kevinbenton | anilvenkata: it's DOWN for new ports in ML2 | 07:56 |
| anilvenkata | kevinbenton, so then no changes needed in l2pop :) | 07:56 |
| kevinbenton | anilvenkata: well what i want to avoid is having an agent reboot require flipping the status on the ports | 07:57 |
| kevinbenton | anilvenkata: so if they were already active before reboot, i will need an interface to retrieve the forwarding table for an agent | 07:57 |
| *** lennyb has quit IRC | 07:57 | |
| kevinbenton | anilvenkata: so they don't have to go ACTIVE->BUILD->ACTIVE each time | 07:57 |
| anilvenkata | kevinbenton, yes, no rewiring | 07:58 |
| kevinbenton | anilvenkata: right now an agent restart increments the revision_number and changes the updated_at in the API | 07:58 |
| kevinbenton | anilvenkata: it would be nice if it had no impact | 07:58 |
| kevinbenton | anilvenkata: also better for performance | 07:58 |
| *** jlinkes has joined #openstack-neutron | 07:58 | |
| anilvenkata | kevinbenton, yes dataplane will be also faster as we wont recreate ovs flows for this port on other agents | 07:59 |
| kevinbenton | anilvenkata: ah, there might be an issue though. we need to be careful to detect an IP address change of the agent on restart | 08:00 |
| *** yamamoto has joined #openstack-neutron | 08:00 | |
| kevinbenton | anilvenkata: let me think about it for a bit. i'll tag you on a patch for it | 08:00 |
| anilvenkata | kevinbenton, yes | 08:00 |
| anilvenkata | kevinbenton, sure, thanks Kevin | 08:00 |
| kevinbenton | anilvenkata: thanks for your help | 08:00 |
| anilvenkata | kevinbenton, welcome :) | 08:01 |
| *** yamamoto has quit IRC | 08:02 | |
| anilvenkata | ajo, ping | 08:03 |
| *** yamamoto has joined #openstack-neutron | 08:03 | |
| *** horms has joined #openstack-neutron | 08:03 | |
| *** vijaykc4 has joined #openstack-neutron | 08:06 | |
| *** gkadam|lunch is now known as gkadam | 08:06 | |
| openstackgerrit | Thomas Morin proposed openstack/neutron: Fix OVSBridge.delete_flows when called with no args https://review.openstack.org/423153 | 08:08 |
| *** TuHV has quit IRC | 08:09 | |
| *** obondarev has quit IRC | 08:09 | |
| *** lennyb_ has joined #openstack-neutron | 08:09 | |
| *** lennyb_ has quit IRC | 08:09 | |
| *** lennyb has joined #openstack-neutron | 08:10 | |
| openstackgerrit | Merged openstack/neutron-fwaas: Optimize _make_firewall_dict_with_rules db queries https://review.openstack.org/424361 | 08:10 |
| Miouge | kevinbenton: I’m exactly in the situation you described in https://bugs.launchpad.net/neutron/+bug/1486039/comments/2 any advice? | 08:12 |
| openstack | Launchpad bug 1486039 in neutron "Setting a policy to a network will limit the router/dhcp/net-device ports, that's not expected" [High,Fix released] - Assigned to Miguel Angel Ajo (mangelajo) | 08:12 |
| kevinbenton | Miouge: so you want to limit router ports? | 08:14 |
| *** gcheresh has joined #openstack-neutron | 08:14 | |
| Miouge | That’s right, I was trying to set the qos-policy on the external network to limit all tenants to say 1Gbps | 08:15 |
| Miouge | (to address noisy neighbour effect when 1 tenant runs a backup job or something) | 08:16 |
| *** belharar_ has quit IRC | 08:16 | |
| kevinbenton | Miouge: yeah, based on the fix that went in (https://git.openstack.org/cgit/openstack/neutron/tree/neutron/objects/qos/rule.py?id=fe236bdaadb949661a0bfb9b62ddbe432b4cf5f1#n75) it sounds like you will have to apply it to each router port | 08:16 |
| kevinbenton | Miouge: so you would need to script setting the qos policy manually on each port :( | 08:17 |
| *** horms has quit IRC | 08:20 | |
| Miouge | kevinbenton: I ran some tests in a Mitaka environment the qos-policy is effective only on VM ports. It does not apply if I specify a router port (private or external side). I am considering to patch the rule.py to allow this use case :D | 08:21 |
| Miouge | I understand that we don’t want the private side of the router to get the qos-policy when applied to a network, but it can be pretty neat for the external side. I’m not really kean on making a script re-implement it | 08:22 |
| *** jlibosva has joined #openstack-neutron | 08:22 | |
| openstackgerrit | Thomas Morin proposed openstack/neutron: OpenFlowSwitchMixin: do not override delete_flows https://review.openstack.org/380329 | 08:24 |
| *** ralonsoh has joined #openstack-neutron | 08:24 | |
| kevinbenton | Miouge: perhaps you can file an upstream patch as well that has it apply if the device owner is network:router_gateway | 08:24 |
| kevinbenton | Miouge: seems like that would be an easy way to solve the use case without much complexity | 08:24 |
| Miouge | That’s what I’m thinking :) | 08:25 |
| Miouge | I’ll give it a short then | 08:25 |
| Miouge | kevinbenton: on a similar topic, any idea why the bw-limit are only egress? | 08:26 |
| *** tommylikehu_ has joined #openstack-neutron | 08:27 | |
| kevinbenton | Miouge: IIRC that was a limitation of the implimentation | 08:28 |
| kevinbenton | Miouge: couldn't apply policies on the way out of an interface or something like that | 08:28 |
| kevinbenton | Miouge: once ajo is online he can shed some light on that | 08:28 |
| kevinbenton | Miouge: like no queue to shape on the egress from the bridge's perspective | 08:29 |
| *** Jeffrey4l_ has joined #openstack-neutron | 08:30 | |
| *** vijaykc4 has quit IRC | 08:31 | |
| Miouge | I guess that the igress traffic shaping won’t do much for UDP but for TCP the sender should adjust if retransmits are requested. I’ll stick around then | 08:32 |
| *** tommylikehu_ has quit IRC | 08:32 | |
| *** abregman has quit IRC | 08:32 | |
| *** Matias has quit IRC | 08:33 | |
| *** amarao has joined #openstack-neutron | 08:33 | |
| *** Jeffrey4l has quit IRC | 08:34 | |
| *** moshele has quit IRC | 08:34 | |
| ralonsoh | kevinbenton: ping | 08:34 |
| *** moshele has joined #openstack-neutron | 08:34 | |
| ralonsoh | kevinbenton: about https://review.openstack.org/#/c/357865/ | 08:35 |
| *** bfernando has joined #openstack-neutron | 08:35 | |
| *** Matias has joined #openstack-neutron | 08:36 | |
| *** john-davidge has joined #openstack-neutron | 08:36 | |
| kevinbenton | ralonsoh: pong | 08:36 |
| kevinbenton | ralonsoh: just read response | 08:37 |
| kevinbenton | ralonsoh: sounds good | 08:37 |
| ralonsoh | kevinbenton: perfect! thank you | 08:37 |
| *** horms has joined #openstack-neutron | 08:38 | |
| *** mkolesni has quit IRC | 08:39 | |
| *** mkolesni has joined #openstack-neutron | 08:39 | |
| reedip | kevinbenton : ping | 08:39 |
| kevinbenton | reedip: pong | 08:39 |
| reedip | kevinbenton :just wanted to know your opinion about https://bugs.launchpad.net/neutron/+bug/1437496 | 08:40 |
| openstack | Launchpad bug 1437496 in neutron "port-update --fixed-ips doesn't work for routers" [Undecided,Incomplete] | 08:40 |
| *** Matias has quit IRC | 08:41 | |
| kevinbenton | reedip: i would confirm that it's still an issue. i know there have been quite a few changes over the last year to notifications to the routers (both server side and l3 agent side) | 08:42 |
| *** hfu has joined #openstack-neutron | 08:42 | |
| *** ArchiFleKs has quit IRC | 08:42 | |
| reedip | kevinbenton : okay, then I will start working on it if possible :) | 08:42 |
| reedip | thanks for your quick response | 08:43 |
| kevinbenton | reedip: no prob | 08:43 |
| *** abregman has joined #openstack-neutron | 08:47 | |
| *** horms has quit IRC | 08:47 | |
| *** amotoki has quit IRC | 08:47 | |
| *** chandanc_ has joined #openstack-neutron | 08:49 | |
| *** pgadiya has quit IRC | 08:54 | |
| *** iranzo has quit IRC | 08:54 | |
| *** Matias has joined #openstack-neutron | 08:55 | |
| *** vikram has quit IRC | 08:56 | |
| openstackgerrit | Thomas Morin proposed openstack/neutron: OVS: merge the required OpenFlow version rather than replace https://review.openstack.org/371455 | 08:57 |
| *** abregman has quit IRC | 08:58 | |
| *** hfu has quit IRC | 08:59 | |
| *** rossella_ has joined #openstack-neutron | 08:59 | |
| *** hoangcx has quit IRC | 09:00 | |
| *** zzzeek has quit IRC | 09:00 | |
| *** thorst_ has joined #openstack-neutron | 09:00 | |
| *** abregman has joined #openstack-neutron | 09:00 | |
| *** zzzeek has joined #openstack-neutron | 09:00 | |
| *** masaki has quit IRC | 09:00 | |
| *** claudiub|2 is now known as claudiub | 09:02 | |
| *** horms has joined #openstack-neutron | 09:02 | |
| *** hichihara has quit IRC | 09:02 | |
| *** tmorin has quit IRC | 09:03 | |
| *** tmorin has joined #openstack-neutron | 09:04 | |
| *** ArchiFleKs has joined #openstack-neutron | 09:04 | |
| *** thorst_ has quit IRC | 09:05 | |
| *** vijaykc4 has joined #openstack-neutron | 09:08 | |
| *** hfu has joined #openstack-neutron | 09:08 | |
| *** kevo has quit IRC | 09:09 | |
| *** pgadiya has joined #openstack-neutron | 09:09 | |
| *** vijaykc4 has quit IRC | 09:09 | |
| *** vijaykc4 has joined #openstack-neutron | 09:10 | |
| *** iwamoto has quit IRC | 09:10 | |
| *** jpena|off is now known as jpena | 09:11 | |
| *** rvba` is now known as rvba | 09:13 | |
| *** mickeys has quit IRC | 09:13 | |
| *** zhurong has quit IRC | 09:14 | |
| *** pmannidi has quit IRC | 09:14 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: ovs fw migration from iptables https://review.openstack.org/413082 | 09:15 |
| jschwarz | jlibosva, rossella_, as promised, with tests: ^ | 09:15 |
| *** efoley has joined #openstack-neutron | 09:21 | |
| openstackgerrit | Thomas Morin proposed openstack/neutron: Add OVS agent bridge extensions https://review.openstack.org/425074 | 09:22 |
| *** efoley_ has joined #openstack-neutron | 09:23 | |
| openstackgerrit | Cedric Brandily proposed openstack/neutron-fwaas: Optimize _make_firewall_group_dict_with_rules https://review.openstack.org/424875 | 09:24 |
| *** jpena is now known as jpena|off | 09:25 | |
| *** hfu_ has joined #openstack-neutron | 09:26 | |
| *** oreillyd_ has joined #openstack-neutron | 09:26 | |
| *** mhickey has joined #openstack-neutron | 09:27 | |
| *** efoley has quit IRC | 09:27 | |
| *** lucas-afk is now known as lucasagomes | 09:27 | |
| *** hfu has quit IRC | 09:29 | |
| *** liusheng has quit IRC | 09:29 | |
| *** limao has quit IRC | 09:31 | |
| *** jpena|off is now known as jpena | 09:31 | |
| *** davidsha has joined #openstack-neutron | 09:33 | |
| *** horms has quit IRC | 09:35 | |
| openstackgerrit | Nguyen Phuong An proposed openstack/neutron-specs: (Operator-only) Logging API for security groups https://review.openstack.org/203509 | 09:35 |
| *** mdnadeem has quit IRC | 09:37 | |
| *** yamamoto has quit IRC | 09:38 | |
| *** horms has joined #openstack-neutron | 09:39 | |
| *** mdnadeem has joined #openstack-neutron | 09:39 | |
| *** annp has joined #openstack-neutron | 09:45 | |
| *** udesale__ has quit IRC | 09:47 | |
| *** rossella_ has quit IRC | 09:47 | |
| *** udesale has quit IRC | 09:47 | |
| *** rossella_ has joined #openstack-neutron | 09:47 | |
| *** nplanel has joined #openstack-neutron | 09:48 | |
| ajo | kevinbenton Miouge if you apply the policy to the router ports it should work as well | 09:49 |
| ajo | I mean, if you have the internal ports of the router , and you attach a policy (with egress limit of 1Mbps) to each of it's internal ports | 09:50 |
| ajo | each internal tenant network will only be able to consume 1Mbps download each | 09:50 |
| ajo | each/each ':D | 09:50 |
| *** fzdarsky_ is now known as fzdarsky|afk | 09:50 | |
| ajo | Miouge what plugin / mechdriver are you using? | 09:51 |
| ajo | We eventually have to find ways of attaching policies to routers as a whole | 09:51 |
| kevinbenton | ajo: right, but what was the reason we didn't have the ability to limit ingress (from VM perspective) on it's port? | 09:51 |
| Miouge | ajo: I tried last night, did not see the qos-policy applied, but i’ll double check | 09:51 |
| ajo | kevinbenton we have an on flight patch for that that we can land after the qos driver refactor | 09:52 |
| *** ushkalim has joined #openstack-neutron | 09:52 | |
| ajo | kevinbenton it was just that we started with egress, and we will have ingress next. | 09:52 |
| ajo | manpower :D | 09:52 |
| ajo | I guess | 09:52 |
| kevinbenton | ajo: ack. for some reason i thought we had a limitation with OVS | 09:52 |
| Miouge | ajo: I use the openvswitch mech driver | 09:53 |
| ajo | https://review.openstack.org/#/c/303626/ | 09:53 |
| *** abregman has quit IRC | 09:53 | |
| ajo | kevinbenton the ingress of the VM is more expensive, we need to use queues, etc, instead of just policing (dropping packets) | 09:53 |
| kevinbenton | ajo: why can't we police? | 09:53 |
| ajo | Miouge, then, applying egress limit to router ports directly may help you, if that works for you | 09:53 |
| ajo | kevinbenton it's a linux kernel limitation AFAIK | 09:54 |
| kevinbenton | ajo: ah, that must have been what i was thinking of then. i remember it wasn't as simple as "do the same thing but the other way" :) | 09:54 |
| ajo | the same way, we can't use queues for egress (nftables can) | 09:54 |
| *** fzdarsky|afk has quit IRC | 09:55 | |
| ajo | for egress, if you want queues, you need to assemble an ifb ( ralonsoh did already something like that for Linuxbridge minimum bandwidth support, since policer won't help) | 09:55 |
| Miouge | I’ll keep an eye on the direction patch :p | 09:55 |
| ajo | Miouge thanks, I hope we can land it in Pike | 09:55 |
| reedip | ajo : ping | 09:56 |
| ajo | it didn't land in Ocata mostly because of me, I slowed down the whole waterfall of features with my refactor :( | 09:56 |
| ajo | reedip pong | 09:56 |
| reedip | ajo : hi, we have QoS metting today at 1500 UTC, right? | 09:56 |
| Miouge | I’m not sure how effective it will be for noisyneighbors, because at the end of the day policing on the hypervisor won’t help free up the transit link? | 09:57 |
| ajo | reedip it's strange, we changed it in repo, but the calendar didn't update | 09:57 |
| ajo | kevinbenton do you know if there's something wrong with the eavesdrop.openstack.org calendar generator? | 09:57 |
| ralonsoh | no, next Tuesday | 09:57 |
| ralonsoh | at 15:00 UTC | 09:57 |
| reedip | ajo : I would like to discuss the ECN point in the meeting, if possible | 09:57 |
| reedip | damn :( | 09:57 |
| kevinbenton | ajo: not sure. i haven't heard anything about it having issues | 09:57 |
| ajo | hmm | 09:57 |
| kevinbenton | ajo: i need to confirm a rolling upgrades question | 09:57 |
| reedip | thanks ralonsoh. | 09:57 |
| ajo | kevinbenton tellme | 09:58 |
| kevinbenton | ajo: we can depend on all of the servers being the newest version for an agent, right? | 09:58 |
| kevinbenton | ajo: i.e. an Ocata agent doesn't have to work with a newton server | 09:58 |
| ajo | That's correct | 09:59 |
| reedip | ralonsoh : I have asked ajo for his review help, but would appreciate if you can also help me with the ECN implementation. That is , if you have some bandwidth ? | 09:59 |
| ajo | we don't support upgrading any agent before all servers are on the latest version | 09:59 |
| *** vijaykc4 has quit IRC | 09:59 | |
| ralonsoh | reedip: can you send me the bug number? | 09:59 |
| ajo | kevinbenton but that brings me one corner case I don't know if I thought of. | 09:59 |
| ajo | ah | 10:00 |
| ajo | no, it's ok :D | 10:00 |
| kevinbenton | ajo: ok, then i can count on modern servers upconverting to latest OVO | 10:00 |
| ajo | I was thinking of , you upgrade server 1, 2, but still not 3, an then agent A is newer... | 10:00 |
| ajo | that case is not supported | 10:00 |
| kevinbenton | ajo: right | 10:00 |
| kevinbenton | ajo: that was the case i was worried about | 10:00 |
| ajo | exactly | 10:00 |
| *** pgadiya has quit IRC | 10:00 | |
| reedip | ralonsoh : Bug ID : https://bugs.launchpad.net/neutron/+bug/1505627 , and the Etherpad where I am drafting things up : https://etherpad.openstack.org/p/QoS_ECN | 10:00 |
| openstack | Launchpad bug 1505627 in neutron "[RFE] QoS Explicit Congestion Notification (ECN) Support" [Wishlist,Incomplete] - Assigned to Reedip (reedip-banerjee) | 10:00 |
| kevinbenton | ajo: old, dirty, rusted out RPC interfaces | 10:00 |
| *** ygbo has joined #openstack-neutron | 10:00 | |
| ajo | yes but we don't support that | 10:00 |
| ajo | hehehe | 10:00 |
| ralonsoh | reedip: give me some time to ask my manager | 10:01 |
| ajo | kevinbenton I saw your bulk_pull rpc | 10:01 |
| ralonsoh | reedip: I'll reply you today | 10:01 |
| reedip | ralonsoh : sure , no issues. Thanks :) | 10:01 |
| ajo | I was wondering if we could somehow coalesce the filters option to the other pull, so we just have one method | 10:01 |
| *** TMM has joined #openstack-neutron | 10:01 | |
| ajo | then I thought it wasn't possible, but I can't remember why | 10:01 |
| kevinbenton | ajo: i think the pull response expects one object | 10:02 |
| ajo | kevinbenton no, it expects a list | 10:02 |
| ajo | since we can now pass a list of IDs as far as I remember | 10:02 |
| ajo | we changed that | 10:02 |
| ajo | I believe | 10:03 |
| ajo | may be I'm wrong now | 10:03 |
| kevinbenton | ajo: L101 https://review.openstack.org/#/c/412613/5/neutron/api/rpc/handlers/resources_rpc.py | 10:03 |
| ajo | hmm | 10:03 |
| *** annp has quit IRC | 10:03 | |
| ajo | ah | 10:03 |
| ajo | sorry it was the push | 10:03 |
| kevinbenton | ajo: that reminds me | 10:03 |
| kevinbenton | ajo: why do we have the call to the producer registry in between? | 10:04 |
| kevinbenton | ajo: why doesn't pull just do a get_object directly? | 10:04 |
| ajo | kevinbenton , let me read the code, which call? | 10:04 |
| kevinbenton | ajo: L131 https://review.openstack.org/#/c/412613/5/neutron/api/rpc/handlers/resources_rpc.py | 10:04 |
| ajo | ahh | 10:04 |
| ajo | the idea is that services can register to provide those | 10:05 |
| ajo | and be decoupled | 10:05 |
| ajo | so, an external service could join the party, and provide himself as a producer for some type of OVO out of tree | 10:05 |
| kevinbenton | ajo: but it's OVO. why does a service need to register to provide that? | 10:05 |
| ajo | still, we'd need to provide the dynamic registration | 10:05 |
| ajo | but that was the plan | 10:05 |
| kevinbenton | ajo: but wouldn't _resource_to_class return whatever was dynamically registered as a type? | 10:06 |
| ajo | yeah, we could may be use a refactor there now, our neutron ovos have changed a lot since we started :D | 10:06 |
| kevinbenton | ajo: yeah, it just seemed like a lot of extra complexity to put something in between | 10:07 |
| ajo | we can probably simplify that a lot | 10:07 |
| kevinbenton | ajo: if we could get rid of it, that would make it a lot easier to follow :) | 10:07 |
| ajo | ok, I will give it a clean up along with the dynamic registration of ovos for push/pull | 10:07 |
| ajo | and try to remember if ain't missing anything | 10:08 |
| kevinbenton | ajo: while i have you, can you do a couple of quick OVO reviews? | 10:08 |
| ajo | kevinbenton sure thing | 10:08 |
| kevinbenton | ajo: https://review.openstack.org/#/c/417672/ https://review.openstack.org/#/c/410417/ | 10:08 |
| ajo | acl | 10:08 |
| ajo | ack | 10:08 |
| kevinbenton | ajo: both pretty easy | 10:08 |
| Miouge | ajo: tried again to apply the qos-policy on the private port of the router, no effect on bandwidth: http://paste.openstack.org/show/596402/ | 10:09 |
| ajo | Miouge can you file a bug for that? | 10:10 |
| ajo | that should work | 10:10 |
| ajo | ralonsoh ^ | 10:10 |
| ajo | Miouge please add the #qos tag to the bug | 10:10 |
| Miouge | Sure! | 10:10 |
| *** iyamahat_ has quit IRC | 10:10 | |
| kevinbenton | Miouge: do you also have that qos policy ID set for the network? | 10:11 |
| kevinbenton | Miouge: based on that conditional we saw, i think that's the only way it would work | 10:11 |
| Miouge | Nope qos-policy is set only for the router port, I removed it from the private and external networks | 10:12 |
| *** pgadiya has joined #openstack-neutron | 10:12 | |
| ralonsoh | Miouge: which backend do you have? | 10:12 |
| kevinbenton | Miouge: https://git.openstack.org/cgit/openstack/neutron/tree/neutron/objects/qos/rule.py?id=fe236bdaadb949661a0bfb9b62ddbe432b4cf5f1#n70 | 10:12 |
| kevinbenton | Miouge: the only way that will return True is if is_network_rule is false | 10:13 |
| Miouge | ralonsoh: ovs mech driver, vlan networks | 10:13 |
| *** sambetts|afk is now known as sambetts | 10:13 | |
| ralonsoh | Miouge: but in OVS qos applies only in the interfaces of the VMs | 10:14 |
| kevinbenton | Miouge: and for is_network_rule to be false, I think network qos policy has to match port qos policy | 10:15 |
| *** horms_ has joined #openstack-neutron | 10:15 | |
| *** puck is now known as Guest85154 | 10:15 | |
| *** neiljerram has joined #openstack-neutron | 10:16 | |
| kevinbenton | Miouge: unless there some some logic somewhere else skipping network owned ports for QoS | 10:16 |
| *** hfu_ has quit IRC | 10:16 | |
| kevinbenton | ralonsoh: where is the logic that only applies to VMs? | 10:16 |
| Miouge | kevinbenton: you are right, “neutron net-update --qos-policy bw-limiter 3a91f3bf-9d81-4f2a-90cc-6d7d3544f00e” and the qos works | 10:17 |
| *** openstackgerrit has quit IRC | 10:17 | |
| ralonsoh | Miouge: is this an external router? | 10:18 |
| Miouge | ralonsoh: it’s a neutron router with external connectivity | 10:18 |
| Miouge | (10.0.0.0/24 on the private side and public IP ont the public side) | 10:19 |
| *** amotoki has joined #openstack-neutron | 10:19 | |
| *** obondarev has joined #openstack-neutron | 10:19 | |
| *** nmathew has joined #openstack-neutron | 10:19 | |
| Miouge | Initialy, I wanted to use qos on the whole public network to avoid VMs/people from using the whole external bandwidth | 10:20 |
| jlibosva | kevinbenton: hi, would you be so kind? https://review.openstack.org/#/c/418867/7 | 10:20 |
| ralonsoh | Miouge: so the qos should apply on this port 3a91f3bf-9d81-4f2a-90cc-6d7d3544f00e | 10:20 |
| *** horms has quit IRC | 10:20 | |
| *** horms_ is now known as horms | 10:20 | |
| ralonsoh | Miouge: now it's working, isn't it? | 10:20 |
| *** mdnadeem has quit IRC | 10:21 | |
| Miouge | ralonsoh: 3a91… is the private port of the router. Now it is working if I set qos-policy also on the private network | 10:21 |
| kevinbenton | ralonsoh: it's because of this logic https://git.openstack.org/cgit/openstack/neutron/tree/neutron/objects/qos/rule.py?id=fe236bdaadb949661a0bfb9b62ddbe432b4cf5f1#n70 | 10:21 |
| Miouge | (i’m testing the same thing on the external/public network qos-policy on port and on networ) | 10:21 |
| kevinbenton | ralonsoh: the only way you can get a policy to apply to a network owned port is to have the port policy and network policy match | 10:21 |
| *** ociuhandu has joined #openstack-neutron | 10:22 | |
| ralonsoh | Miouge: no no | 10:22 |
| ralonsoh | Miouge: I'll check this | 10:22 |
| *** nmathew- has quit IRC | 10:23 | |
| ralonsoh | Miouge: but you don't need to have both. The port policy takes precedence over the network policy | 10:23 |
| ralonsoh | and you don't need both | 10:23 |
| *** mvk has quit IRC | 10:23 | |
| ajo | kevinbenton asked a little question: https://review.openstack.org/#/c/410417/4/neutron/objects/rbac_db.py | 10:23 |
| *** openstackgerrit has joined #openstack-neutron | 10:24 | |
| openstackgerrit | Carlos Goncalves proposed openstack/neutron-lib: Adds API definition for data plane status extension https://review.openstack.org/424868 | 10:24 |
| *** nmathew- has joined #openstack-neutron | 10:24 | |
| *** nmathew- has quit IRC | 10:24 | |
| Miouge | the logic is equivalent to: “(self.qos_policy_id == port[qos_consts.QOS_POLICY_ID] or not(any(port['device_owner'].startswith(prefix) for prefix in constants.DEVICE_OWNER_PREFIXES)))” I agree setting it on the port should be enough, but empirical evidence shows otherwise | 10:24 |
| *** yamamoto has joined #openstack-neutron | 10:25 | |
| kevinbenton | ralonsoh, Miouge: the logic is a not of an AND | 10:25 |
| *** vijaykc4 has joined #openstack-neutron | 10:25 | |
| *** yamamoto has quit IRC | 10:26 | |
| Miouge | not(a AND b) <=> not(a) or not(b) no? | 10:26 |
| openstackgerrit | Carlos Goncalves proposed openstack/neutron-lib: Adds API definition for data plane status extension https://review.openstack.org/424868 | 10:27 |
| kevinbenton | Miouge: right, i didn't notice you swapped the policy to == in there | 10:28 |
| kevinbenton | Miouge: so yeah, the behavior matches your expansion | 10:28 |
| kevinbenton | Miouge: router ports will never pass the second half because they have 'network' device_owners | 10:28 |
| kevinbenton | Miouge: so the only way to get them to pass the first half is to make the policy IDs match | 10:29 |
| *** nyechiel_ has quit IRC | 10:29 | |
| ralonsoh | Miouge: "'network' device_owners" | 10:29 |
| *** mdnadeem has joined #openstack-neutron | 10:29 | |
| ralonsoh | Miouge: that's the point | 10:29 |
| openstackgerrit | Bertrand Lallau proposed openstack/neutron-fwaas: Fix scale issue in case of services_sync_needed https://review.openstack.org/424551 | 10:30 |
| kevinbenton | ralonsoh: yeah, i think you missed the earlier conversation | 10:30 |
| ralonsoh | Miouge: the router doesn't belongs to you | 10:30 |
| ralonsoh | Miouge: ooook | 10:30 |
| kevinbenton | ralonsoh: the router does belong in your network | 10:30 |
| kevinbenton | ralonsoh: so you should be able to put policies on it | 10:30 |
| Miouge | The router belongs to my tenant | 10:30 |
| kevinbenton | ralonsoh: but it happening by default was confusing people | 10:31 |
| kevinbenton | ralonsoh: https://bugs.launchpad.net/neutron/+bug/1486039/ | 10:31 |
| openstack | Launchpad bug 1486039 in neutron "Setting a policy to a network will limit the router/dhcp/net-device ports, that's not expected" [High,Fix released] - Assigned to Miguel Angel Ajo (mangelajo) | 10:31 |
| *** dewanee has joined #openstack-neutron | 10:31 | |
| ralonsoh | kevinbenton: which should be the solution? To improve the documentation? | 10:32 |
| *** aarefiev_afk is now known as aarefiev | 10:32 | |
| ralonsoh | kevinbenton: because to change the behaviour shouldn't be the answer to this... | 10:32 |
| kevinbenton | ralonsoh: i'm not sure. we did rule out a use case by changing the behavior with the bug fix for 1486039 | 10:32 |
| kevinbenton | ralonsoh: particularily this use case https://bugs.launchpad.net/neutron/+bug/1486039/comments/2 | 10:33 |
| openstack | Launchpad bug 1486039 in neutron "Setting a policy to a network will limit the router/dhcp/net-device ports, that's not expected" [High,Fix released] - Assigned to Miguel Angel Ajo (mangelajo) | 10:33 |
| kevinbenton | ralonsoh: which is what Miouge was trying to achieve | 10:33 |
| *** abregman has joined #openstack-neutron | 10:34 | |
| kevinbenton | ralonsoh: we could potentially alter that conditional so it doesn't apply to router gateway ports | 10:34 |
| ralonsoh | kevinbenton: and only to router ports? | 10:35 |
| kevinbenton | ralonsoh: possibly | 10:35 |
| Miouge | kevinbenton: I think that would be best: exclude qos-policy to private router port, but keep it on gateway router port | 10:35 |
| *** vijaykc4 has quit IRC | 10:36 | |
| kevinbenton | Miouge, ralonsoh: we also may be able to fix that conditional | 10:37 |
| kevinbenton | so if a router port has a qos policy explicitly set, it is used | 10:37 |
| Miouge | I am still confused about the situation: why do I need to set the qos-policy on both the port and the network for it to affect the router port. Is that what you mean with fix the conditional kevinbenton ? | 10:37 |
| kevinbenton | Miouge: yes | 10:37 |
| ralonsoh | but this is the actual behaviour | 10:38 |
| ralonsoh | current behaviour | 10:38 |
| kevinbenton | ralonsoh: no it's not | 10:38 |
| kevinbenton | ralonsoh: the current behavior forces you to set the policy to match the network policy | 10:39 |
| kevinbenton | ralonsoh: and then it will apply | 10:39 |
| kevinbenton | ralonsoh: because of this line | 10:39 |
| kevinbenton | is_network_rule = self.qos_policy_id != port[qos_consts.QOS_POLICY_ID] | 10:39 |
| ralonsoh | ok, so your proposal is to force to set both, port and network for router | 10:39 |
| kevinbenton | no, that's what you have to do now to get it to work | 10:40 |
| kevinbenton | my proposal is if the port has any policy set directly on it, apply it | 10:40 |
| ralonsoh | ok | 10:40 |
| ralonsoh | I'll open a bug for this today | 10:40 |
| kevinbenton | so there are two bugs | 10:41 |
| kevinbenton | one is that we should probably allow *network* level policies apply to network:router_gateway | 10:41 |
| kevinbenton | to allow the use case of throttling all tenants attaching to an external network | 10:41 |
| kevinbenton | and second bug is that if any network has a *port* level policy, always enforce it | 10:42 |
| kevinbenton | ralonsoh: does that make sense? | 10:42 |
| Miouge | Sounds good to me | 10:42 |
| ralonsoh | the second is not necessary | 10:43 |
| ralonsoh | the first one solves the problem with the router | 10:43 |
| ralonsoh | the second one is the current behaviour | 10:43 |
| kevinbenton | ralonsoh: no, they are different use cases | 10:43 |
| kevinbenton | ralonsoh: and the second one is not the current behavior | 10:43 |
| kevinbenton | ralonsoh: the current behavior is that the router policy has to be explicitly set to the same policy as the network policy, or it won't be applied | 10:44 |
| ralonsoh | kevinbenton: ok | 10:44 |
| *** ushkalim has quit IRC | 10:44 | |
| *** obondarev has quit IRC | 10:44 | |
| kevinbenton | ralonsoh: and the use case the second one enables is putting an egress rule on an internal router interface to limit all ingress traffic to a tenant | 10:44 |
| Miouge | (you may have a router with 2 private networks, and want to qos only one of them so you can’t do it with router_gateway) | 10:44 |
| *** huanxie has quit IRC | 10:45 | |
| kevinbenton | well and remember these are egress rules | 10:45 |
| kevinbenton | so egress on an external gateway is egress from the tenant | 10:45 |
| kevinbenton | and egress on an internal interface is ingress to the tenant | 10:45 |
| kevinbenton | so they let you do different directions of shaping | 10:46 |
| *** boden has joined #openstack-neutron | 10:46 | |
| ralonsoh | kevinbenton: I'll send you the links of the bugs today | 10:46 |
| *** ushkalim has joined #openstack-neutron | 10:46 | |
| kevinbenton | ralonsoh: sounds good | 10:46 |
| *** boden has quit IRC | 10:47 | |
| *** dinghh has joined #openstack-neutron | 10:47 | |
| *** dinghh has left #openstack-neutron | 10:48 | |
| kevinbenton | is_network_rule = self.qos_policy_id != port[qos_consts.QOS_POLICY_ID] might just become is_network_rule = port[qos_consts.QOS_POLICY_ID] is None and self.qos_policy_id != port[qos_consts.QOS_POLICY_ID] | 10:48 |
| Miouge | I’m happy to help if I can submit a patch | 10:49 |
| kevinbenton | Miouge: sure, if you want to push up the patch checking for "not device_owner == 'network:router_gateway'" | 10:50 |
| kevinbenton | both of these i think are effectively one line changes | 10:50 |
| *** yedongcan has left #openstack-neutron | 10:50 | |
| kevinbenton | jlibosva: ping | 10:50 |
| *** nyechiel_ has joined #openstack-neutron | 10:51 | |
| kevinbenton | er, sorry. that first conditional can just become "is_network_rule = port[qos_consts.QOS_POLICY_ID] is None" | 10:53 |
| *** iyamahat_ has joined #openstack-neutron | 10:53 | |
| *** gcheresh_ has joined #openstack-neutron | 10:54 | |
| ajo | kevinbenton, but "kevinbenton:one is that we should probably allow *network* level policies apply to network:router_gateway" doen't make sense in some cases | 10:55 |
| ajo | people ended up applying an "egress" limit to a network, expecting VMs to be limited | 10:55 |
| ajo | and then discovering that routers where limitting his egress on internal ports (hence limiting the ingress of the whole network through the port) | 10:56 |
| ajo | that's what we ended doing that | 10:56 |
| ajo | , and ... the policies not working directly on ports, is a bug | 10:56 |
| *** mvk has joined #openstack-neutron | 10:56 | |
| *** yamahata has quit IRC | 10:56 | |
| kevinbenton | ajo: but that bug was about issues with internal router ports | 10:57 |
| ajo | so routers are special devices when it comes to direction, we need to polish that in regards of QoS, and, may be give them special handling. | 10:57 |
| ajo | (still more special) | 10:57 |
| ajo | more love | 10:57 |
| ajo | kevinbenton yes | 10:57 |
| kevinbenton | ajo: so i can see how that might surprise a tenant | 10:57 |
| kevinbenton | ajo: since to the tenant is seems 'ingressy' | 10:57 |
| kevinbenton | ajo: however, we're talking about the gateway port | 10:57 |
| ajo | yes, they were expecting to limit all their ports "egress" | 10:58 |
| ajo | but found their global ingress limited too | 10:58 |
| kevinbenton | which is egress from the tenant's perspected | 10:58 |
| dalvarez | ajo, thanks for https://review.openstack.org/#/c/418120/ maybe some other eyes please? | 10:58 |
| *** gcheresh has quit IRC | 10:58 | |
| ajo | dalvarez ack | 10:58 |
| kevinbenton | ajo: so if an admin sets an egress policy on an external network, it would make sense to me to limit the egress of the router gateway ports | 10:59 |
| ajo | kevinbenton so you mean reverting directionality for network policies on router ports? | 10:59 |
| ajo | reversing | 10:59 |
| kevinbenton | ajo: no, i mean just apply them router gateway ports | 10:59 |
| ajo | kevinbenton sorry, I don't follow :] | 10:59 |
| ajo | on the external (gateway port). ? | 10:59 |
| kevinbenton | ajo: i'm an admin, i want to limit all tenants connecting to the external network with 1mbps egress | 11:00 |
| ajo | aha | 11:00 |
| kevinbenton | ajo: so i set a policy on the network | 11:00 |
| kevinbenton | that should apply IMO | 11:00 |
| ajo | kevinbenton the external network? | 11:00 |
| kevinbenton | yes | 11:00 |
| ajo | yes, that'd make sense for the external network | 11:00 |
| kevinbenton | basically anything the network:router_gateway port is attaching to would make sense | 11:00 |
| openstackgerrit | Artur Korzeniewski proposed openstack/neutron: Integration of (Distributed) Port Binding OVO https://review.openstack.org/407868 | 11:01 |
| kevinbenton | because from the tenant's perspective that is also egress traffic | 11:01 |
| ajo | ahh | 11:01 |
| ajo | ok, router_gateway | 11:01 |
| kevinbenton | yeah | 11:01 |
| ajo | sorry, it seems like my coffee intake wasn't enough this morning | 11:01 |
| *** thorst_ has joined #openstack-neutron | 11:01 | |
| *** martink2_ has quit IRC | 11:01 | |
| ajo | kevinbenton++ | 11:01 |
| ajo | thanks fro the patience | 11:01 |
| ajo | for | 11:01 |
| ajo | so, yes, that would make sense | 11:01 |
| ajo | that would be simpler than letting people attach policies to routers | 11:01 |
| ajo | such model could have other advantages, like granularity | 11:02 |
| ajo | but, if you want to do it as a whole, on your external network, I don't see a reason not to provide that | 11:02 |
| *** chandanc_ has quit IRC | 11:02 | |
| *** mkolesni has quit IRC | 11:02 | |
| ajo | Miouge could you file an RFE for what kevinbenton was saying ^ | 11:02 |
| Miouge | Yes, the admin sets the policy on the external network and it will automatically be applied to all routers with a network:router_gateway on it :) | 11:02 |
| kevinbenton | ajo: yeah, and that's exactly the use case Miouge is targeting | 11:02 |
| *** mkolesni has joined #openstack-neutron | 11:02 | |
| ajo | ok, got it | 11:03 |
| kevinbenton | ajo: i'm not sure we need an RFE? it's a one line change to sort of undo part of that bug fix | 11:03 |
| Miouge | RFE? is that Request for Evidence Oo? | 11:03 |
| kevinbenton | request for enhancement | 11:03 |
| ajo | it's an enhancement, isn't it? well, it's so simple that it could just be a bug | 11:03 |
| ajo | that's true | 11:03 |
| *** horms has quit IRC | 11:03 | |
| kevinbenton | ajo: right, but it's an enhancement needed because the bug fix was too restrictive :) | 11:03 |
| kevinbenton | ajo: it used to work :P | 11:04 |
| ajo | we need to enhance testing coverage on that | 11:04 |
| ajo | in fact, the bugfix will have more testing and doc than LOCs | 11:04 |
| ajo | :D | 11:04 |
| kevinbenton | ajo: ralonsoh is filing two bugs | 11:05 |
| *** thorst_ has quit IRC | 11:05 | |
| ralonsoh | Miouge is going to fill the first one now | 11:06 |
| Miouge | Sure I can do that | 11:06 |
| kevinbenton | ajo: switching gears to my patch | 11:06 |
| kevinbenton | obj_base.NeutronObject.clean_obj_from_primitive does not take a context | 11:06 |
| kevinbenton | so any OVO objects created from that (i.e. everything on the agent) do not have contexts | 11:07 |
| kevinbenton | which will result in an exception here because context is None | 11:07 |
| kevinbenton | https://review.openstack.org/#/c/410417/4/neutron/objects/rbac_db.py | 11:07 |
| *** iyamahat_ has quit IRC | 11:07 | |
| kevinbenton | ajo: ^^ | 11:07 |
| ajo | kevinbenton ack | 11:07 |
| *** vijaykc4 has joined #openstack-neutron | 11:08 | |
| kevinbenton | ajo: we could potentially stash the context on the object as well, but i would rather not if that's its only purpose for now | 11:08 |
| *** nicolasbock has joined #openstack-neutron | 11:09 | |
| ajo | kevinbenton ok, so that's on the agent side | 11:09 |
| ajo | there we really don't care if the network is shared or not, I guess | 11:09 |
| ajo | so defaulting to False if no context makes sense | 11:09 |
| kevinbenton | ajo: right | 11:10 |
| kevinbenton | ajo: shared is based on who is asking :) | 11:10 |
| kevinbenton | ajo: and on the agent, nobody is asking | 11:10 |
| ajo | right | 11:10 |
| ajo | I was thinking of lost context on the server side, and I wondered why | 11:11 |
| kevinbenton | ajo: i added another comment to clarify a bit | 11:11 |
| ajo | may be it would be good to explain that on the commit | 11:11 |
| ajo | yeah | 11:11 |
| ajo | or that | 11:11 |
| ajo | thanks kevinbenton :D | 11:11 |
| kevinbenton | ajo: i can actually respin to put a comment inline if you want | 11:11 |
| *** panda is now known as panda|afk | 11:12 | |
| kevinbenton | ajo: just so people don't wonder why | 11:12 |
| kevinbenton | ajo: doesn't have any +2's yet that i would lose :) | 11:12 |
| ajo | kevinbenton and we need a recheck anyway :D | 11:12 |
| kevinbenton | ajo: ok, new patch coming up | 11:12 |
| ajo | thanks :D | 11:12 |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Use push-notificates for OVSPluginAPI https://review.openstack.org/410385 | 11:14 |
| *** hfu has joined #openstack-neutron | 11:15 | |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Fixes to allow OVO deserializion of ports/networks https://review.openstack.org/410417 | 11:15 |
| kevinbenton | ajo: done ^^ | 11:15 |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Add missing port UPDATE event to ML2 https://review.openstack.org/399964 | 11:16 |
| kevinbenton | ajo: this one is also related to push notifications as well ^^. Took me a long time to realize why my L2 agent wasn't getting a bound port :) | 11:17 |
| ajo | kevinbenton yes I was reading :D | 11:17 |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Add bulk pull OVO interface https://review.openstack.org/412613 | 11:17 |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Agent-side receiver cache for ML2 OVO https://review.openstack.org/404265 | 11:18 |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Use push-notificates for OVSPluginAPI https://review.openstack.org/410385 | 11:18 |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Decompose SG RPC API DB methods https://review.openstack.org/410418 | 11:18 |
| *** yamamoto has joined #openstack-neutron | 11:18 | |
| *** yamamoto has quit IRC | 11:18 | |
| kevinbenton | ajo: you're also intimately familiar with that code ^^ ;) | 11:19 |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Use push notification for security groups https://review.openstack.org/410422 | 11:19 |
| kevinbenton | I know you were dying to review a bunch of patches :D | 11:19 |
| ajo | kevinbenton it doesn't hurt me | 11:19 |
| ajo | :) | 11:19 |
| *** lihi has quit IRC | 11:21 | |
| ajo | kevinbenton a tiny nit on this one: | 11:26 |
| ajo | https://review.openstack.org/#/c/399964/7 | 11:26 |
| Miouge | ralonsoh: I submited the first bug LP#1659265 let me know if I missed anything | 11:26 |
| *** ociuhandu has quit IRC | 11:32 | |
| *** yamamoto has joined #openstack-neutron | 11:33 | |
| *** yamamoto has quit IRC | 11:34 | |
| kevinbenton | ajo: ack, update coming up | 11:37 |
| openstackgerrit | vikram.choudhary proposed openstack/neutron-dynamic-routing: Renamed [BGP] config section to [bgp] https://review.openstack.org/424532 | 11:39 |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Add missing port UPDATE event to ML2 https://review.openstack.org/399964 | 11:39 |
| kevinbenton | ajo: ^^ | 11:39 |
| jlibosva | kevinbenton: pong (/me was at lunch) | 11:40 |
| *** wolverineav has joined #openstack-neutron | 11:40 | |
| kevinbenton | jlibosva: one question on your patch. doesn't seem like we need the generate subnet call? | 11:40 |
| ralonsoh | Moiuge: take a look at this https://www.rdoproject.org/networking/networking-in-too-much-detail/. In this case, the port affected should be L | 11:41 |
| ralonsoh | Miouge: I agree with the bug description | 11:41 |
| *** anilvenkata has quit IRC | 11:41 | |
| jlibosva | kevinbenton: the guest inside would have overlapping addresses | 11:41 |
| jlibosva | kevinbenton: vlan interface needs to be on a different subnet, otherwise egress traffic from instance on guest OS level would pick the trunk interface | 11:42 |
| *** panda|afk is now known as panda | 11:44 | |
| kevinbenton | jlibosva: but don't we get to pick both cidrs for the test? | 11:44 |
| kevinbenton | jlibosva: aren't both a private network? | 11:44 |
| jlibosva | kevinbenton: they are. it's not Neutron related but related to routing of guest OS | 11:44 |
| jlibosva | kevinbenton: let's say you have interface A with address 1 and interface B with address 2, while 1 and 2 are from the same subnet but different networks | 11:44 |
| kevinbenton | jlibosva: no, i mean if we have control over the CIDRs, we can always use static ones for the whole test set | 11:45 |
| kevinbenton | jlibosva: parent net has 10.0.0.0 | 11:45 |
| kevinbenton | jlibosva: sub network has 192.168.0.0 | 11:45 |
| kevinbenton | jlibosva: every run | 11:45 |
| jlibosva | kevinbenton: you mean to introduce constant cidrs not using the range from tempest conf? | 11:47 |
| kevinbenton | jlibosva: right, the range from the tempest conf was i thought for public networks | 11:47 |
| kevinbenton | jlibosva: what we do in the privacy of our own tenant is between us and our family :) | 11:47 |
| jlibosva | kevinbenton: ok, got it now. sorry for not getting it for the first time :) | 11:47 |
| *** bfernando has quit IRC | 11:48 | |
| kevinbenton | jlibosva: no prob. i can see why you thought i was suggesting overlap between subinterface and parent | 11:48 |
| *** ratailor has quit IRC | 11:49 | |
| jlibosva | kevinbenton: wouldn't creating static cidr introduce a potential threat that public network will have the same subnet as the parent subnet? I see we use fips there | 11:53 |
| jlibosva | I mean, if parent network would have 10.0.0.0/24 and we use 10.0.0.0/24 for parent subnet too | 11:53 |
| *** gkadam has quit IRC | 11:55 | |
| kevinbenton | jlibosva: oh, maybe that's why it's there | 11:58 |
| kevinbenton | jlibosva: yeah, better keep using it to be safe | 11:58 |
| *** hfu has quit IRC | 12:00 | |
| *** hfu has joined #openstack-neutron | 12:00 | |
| *** anilvenkata has joined #openstack-neutron | 12:01 | |
| *** hfu has quit IRC | 12:01 | |
| *** hfu has joined #openstack-neutron | 12:01 | |
| *** lucasagomes is now known as lucas-hungry | 12:01 | |
| *** hfu has quit IRC | 12:01 | |
| *** hfu has joined #openstack-neutron | 12:02 | |
| *** hfu has quit IRC | 12:02 | |
| jlibosva | kevinbenton: thanks :) | 12:03 |
| *** hfu has joined #openstack-neutron | 12:03 | |
| *** hfu has quit IRC | 12:03 | |
| *** hfu has joined #openstack-neutron | 12:03 | |
| *** hfu has quit IRC | 12:04 | |
| *** moshele has quit IRC | 12:05 | |
| *** moshele has joined #openstack-neutron | 12:05 | |
| *** haplo37_ has quit IRC | 12:08 | |
| *** anilvenkata has quit IRC | 12:09 | |
| *** donghao has joined #openstack-neutron | 12:09 | |
| *** donghao has quit IRC | 12:09 | |
| *** lihi has joined #openstack-neutron | 12:09 | |
| *** fzdarsky has joined #openstack-neutron | 12:10 | |
| *** itzikb_ has quit IRC | 12:10 | |
| *** haplo37_ has joined #openstack-neutron | 12:10 | |
| *** donghao has joined #openstack-neutron | 12:10 | |
| *** reedip_ has joined #openstack-neutron | 12:11 | |
| *** trinaths has left #openstack-neutron | 12:14 | |
| *** itzikb_ has joined #openstack-neutron | 12:14 | |
| *** wolverineav has quit IRC | 12:14 | |
| *** pgadiya has quit IRC | 12:16 | |
| *** anilvenkata has joined #openstack-neutron | 12:21 | |
| *** lihi has quit IRC | 12:22 | |
| *** wolverineav has joined #openstack-neutron | 12:24 | |
| *** lihi has joined #openstack-neutron | 12:25 | |
| *** psahoo has quit IRC | 12:27 | |
| *** wolverineav has quit IRC | 12:28 | |
| *** catintheroof has joined #openstack-neutron | 12:28 | |
| *** tommylikehu_ has joined #openstack-neutron | 12:29 | |
| *** tommylikehu_ has quit IRC | 12:29 | |
| openstackgerrit | Yushiro FURUKAWA proposed openstack/neutron-fwaas: Fix validation with 'protocol' for firewall_rule https://review.openstack.org/423229 | 12:30 |
| *** tommylikehu_ has joined #openstack-neutron | 12:30 | |
| *** rmart04 has joined #openstack-neutron | 12:33 | |
| *** wolverineav has joined #openstack-neutron | 12:34 | |
| *** neiljerram has quit IRC | 12:36 | |
| jschwarz | kevinbenton, hey, responded to https://review.openstack.org/#/c/413082/10 | 12:37 |
| *** neiljerram has joined #openstack-neutron | 12:39 | |
| *** sudipto has quit IRC | 12:39 | |
| *** sudipto_ has quit IRC | 12:39 | |
| *** wolverineav has quit IRC | 12:39 | |
| *** neiljerram has quit IRC | 12:43 | |
| *** ayogi has quit IRC | 12:44 | |
| *** thorst_ has joined #openstack-neutron | 12:44 | |
| *** neiljerram has joined #openstack-neutron | 12:45 | |
| -openstackstatus- NOTICE: Gerrit is going to be restarted due to slow performance | 12:46 | |
| *** nmathew- has joined #openstack-neutron | 12:48 | |
| -openstackstatus- NOTICE: Gerrit has been successfully restarted | 12:50 | |
| *** nmathew has quit IRC | 12:51 | |
| *** wolverineav has joined #openstack-neutron | 12:53 | |
| *** eezhova has quit IRC | 12:55 | |
| *** nmathew- has quit IRC | 12:55 | |
| *** chandanc_ has joined #openstack-neutron | 12:57 | |
| *** moshele has quit IRC | 12:57 | |
| openstackgerrit | Genadi Chereshnya proposed openstack/neutron: Tempest: Fixing L3 agent hosting router for DVR setup https://review.openstack.org/421155 | 12:57 |
| *** wolverineav has quit IRC | 12:58 | |
| *** lucas-hungry is now known as lucasagomes | 13:00 | |
| *** eezhova has joined #openstack-neutron | 13:01 | |
| *** chandanc_ has quit IRC | 13:02 | |
| openstackgerrit | Jakub Libosvar proposed openstack/neutron: DNM: Try to increase ssh timeout for ubuntu images https://review.openstack.org/425165 | 13:03 |
| *** tbachman has joined #openstack-neutron | 13:05 | |
| *** nagarjung has quit IRC | 13:05 | |
| *** hfu has joined #openstack-neutron | 13:05 | |
| anilvenkata | jlibosva, ajo jschwarz https://github.com/openstack/neutron-dynamic-routing/tree/master/devstack is neutron-dynamic-routing a separate process or part of q-svc, q-l3, q-agent? | 13:09 |
| *** korzen has quit IRC | 13:10 | |
| *** hfu has quit IRC | 13:10 | |
| *** ranjithd has joined #openstack-neutron | 13:10 | |
| *** tbachman has quit IRC | 13:12 | |
| *** tbachman_ has joined #openstack-neutron | 13:12 | |
| ralonsoh | kevinbenton: hi, about enforcing the qos policy | 13:14 |
| ralonsoh | kevinbenton: I've been looking at the code and is happening now | 13:15 |
| ralonsoh | kevinbenton: for VMs ports | 13:15 |
| ralonsoh | kevinbenton: we shouldn't change anything | 13:15 |
| kevinbenton | ralonsoh: what? | 13:21 |
| *** vijaykc4 has quit IRC | 13:21 | |
| kevinbenton | ralonsoh: this isn't about VM ports | 13:21 |
| kevinbenton | ralonsoh: its about the logic blocking the application to router ports | 13:21 |
| kevinbenton | ralonsoh: i'm not saying we should change anything about VM ports. | 13:22 |
| kevinbenton | ralonsoh: the two changes are: 1) honor qos policies set explicitly on router interfaces | 13:23 |
| openstackgerrit | Miguel Angel Ajo proposed openstack/neutron: Transition qos notification driver into qos driver https://review.openstack.org/396651 | 13:23 |
| kevinbenton | ralonsoh: 2) honor the network level qos policy for router_gateway ports | 13:23 |
| *** wolverineav has joined #openstack-neutron | 13:23 | |
| *** jchhatbar has joined #openstack-neutron | 13:24 | |
| *** hishh has joined #openstack-neutron | 13:24 | |
| *** bfernando has joined #openstack-neutron | 13:25 | |
| *** bfernando has quit IRC | 13:25 | |
| ralonsoh | kevinbenton: ok, but we will have problems with the directionality of the qos rules | 13:26 |
| *** janki has quit IRC | 13:26 | |
| *** ociuhandu has joined #openstack-neutron | 13:26 | |
| kevinbenton | ralonsoh: for router_gateway ports, the directionality works as expected | 13:26 |
| kevinbenton | ralonsoh: because they egress to the external network where the policy is set | 13:27 |
| *** tongli has joined #openstack-neutron | 13:27 | |
| kevinbenton | ralonsoh: for the internal ports, if someone is setting a policy explicitly on the interface, we can assume they understand that egress will be to their own VMs | 13:27 |
| ralonsoh | kevinbenton: ok, just to be sure | 13:28 |
| *** wolverineav has quit IRC | 13:28 | |
| kevinbenton | ralonsoh: it won't be a silent surprise like it was before when it automatically applied based on network policy | 13:28 |
| *** wolverineav has joined #openstack-neutron | 13:29 | |
| ajo | jlibosva were you looking at https://review.openstack.org/#/c/344997/9 ? | 13:30 |
| ajo | jlibosva I didn't +W because I thought you could be doing a review atm :) | 13:30 |
| *** shausy has quit IRC | 13:30 | |
| jlibosva | ajo: I am :) | 13:30 |
| ajo | jlibosva ok, then I'll hold it, I didn't spot anything on my side | 13:30 |
| ajo | but I certainly never spot it all | 13:31 |
| jlibosva | ajo: I spotted bad indentation but that's a int :) | 13:31 |
| jlibosva | nit | 13:31 |
| ajo | jlibosva ok, may be we can fix that on the fly, and submit it to gate | 13:31 |
| ajo | no reason to make DPDKers life harder :D | 13:32 |
| ajo | sean-k-mooney ^ | 13:32 |
| tbachman_ | kevinbenton: “artisanal NFV workloads” | 13:32 |
| *** moshele has joined #openstack-neutron | 13:33 | |
| * tbachman_ is laughing pretty hard | 13:33 | |
| *** tbachman_ is now known as tbachman | 13:33 | |
| tbachman | that’s a gem | 13:33 |
| *** udesale has joined #openstack-neutron | 13:33 | |
| ajo | lol, what are you talking about tbachman | 13:33 |
| kevinbenton | :) | 13:33 |
| tbachman | openstack-dev | 13:33 |
| tbachman | PTL candidacy thread | 13:33 |
| kevinbenton | http://lists.openstack.org/pipermail/openstack-dev/2017-January/110924.html | 13:34 |
| ajo | ralonsoh I was confused this morning too, apparently we have a bug, when directly applied to router ports (not through network) it won't work | 13:34 |
| ajo | and also router_gateway ports are the external ports of routers | 13:34 |
| ajo | in that case it would make sense to apply the rules too (egress is egress, ingress is ingress at that point) | 13:35 |
| jlibosva | ajo: I didn't even comment on that, let's get it in :) | 13:35 |
| ajo | jlibosva awesome | 13:36 |
| kevinbenton | ajo, ralonsoh: one thing we need to be sure of is that we don't forget the external port on DVR compute nodes | 13:36 |
| kevinbenton | i don't remember at the moment if they are 'network:router_gateway' as well | 13:37 |
| sean-k-mooney | ajo: jlibosva i was just looking for the comment. since it passed pep8 i assumend the indentaiton was fine | 13:37 |
| kevinbenton | or if they have a different device owner | 13:37 |
| ajo | kevinbenton ack, makes sense | 13:37 |
| sean-k-mooney | also thanks for taking the time to review | 13:37 |
| ajo | sean-k-mooney pep8 doesn't catch some nits | 13:37 |
| kevinbenton | ajo: that's by design. how else would we get to -1 things? :P | 13:38 |
| jlibosva | sean-k-mooney: it's not a pep8 violation :) I was thinking about https://review.openstack.org/#/c/344997/9/neutron/plugins/ml2/drivers/openvswitch/mech_driver/mech_openvswitch.py@87 and following line :) It's easier to read where the condition ends if the indentation isn't the same as the content of the block after 'if'. but maybe that's just me :) | 13:38 |
| ajo | kevinbenton :P :) | 13:38 |
| ajo | kevinbenton style is critical :) | 13:39 |
| sean-k-mooney | jlibosva: ah ok | 13:39 |
| jlibosva | code must look beautifully, the functional aspect is just a bonus | 13:40 |
| ajo | long if clauses are always hard to read :D | 13:40 |
| ajo | sometimes I split the and/or'ed clauses into vars | 13:40 |
| kevinbenton | jlibosva: there actually is a pep8 check for that i think | 13:40 |
| ajo | with human names :D | 13:40 |
| *** jlinkes has quit IRC | 13:40 | |
| ajo | but that's extremist nitpicking :D | 13:40 |
| kevinbenton | jlibosva: I remember it yelling at me long ago | 13:40 |
| kevinbenton | jlibosva: we may have shut it off | 13:40 |
| jlibosva | kevinbenton: if there is, it's not working anymore | 13:40 |
| jlibosva | apparently | 13:40 |
| sean-k-mooney | i think if i double indented it might have been a pep issue but a new line may have helped. | 13:40 |
| sean-k-mooney | see this is why i am happy clang fromat is becoming more popular in the c++ comunities | 13:41 |
| sean-k-mooney | you create one file and it formats your code | 13:41 |
| jlibosva | ajo: human names? you mean you name variables e.g. "eric = this and that", thomas = "that or this" ? :D | 13:42 |
| sean-k-mooney | no arguing be cause you can make it a git commit hook and its done on every commit | 13:42 |
| ajo | jlibosva lol | 13:42 |
| ajo | rofl | 13:42 |
| kevinbenton | jlibosva: marry_poppings = j << 1 & filter_mask | 13:42 |
| ajo | and then | 13:42 |
| *** eezhova has quit IRC | 13:42 | |
| ajo | if eric and thomas: .... | 13:42 |
| kevinbenton | jlibosva: marry_poppins * | 13:42 |
| jlibosva | :] | 13:42 |
| sean-k-mooney | i had the lovely pleasue in my previous job of cleaning up a phd students code where in his theais work he worte for loops of boob in boobs ... | 13:43 |
| kevinbenton | jlibosva, ajo: https://github.com/openstack/neutron/blame/master/tox.ini#L136 | 13:44 |
| sean-k-mooney | just because he was granted a doctorate doesnt mean his mental age was over 12 | 13:44 |
| kevinbenton | sean-k-mooney: ha! | 13:44 |
| kevinbenton | sean-k-mooney: it's unfortunate people with the guts to do that don't have more creativity | 13:46 |
| kevinbenton | sean-k-mooney: i remember reading some thread somewhere about an engineer getting scolded for having customer facing messages that mentioned "Killing orphans" | 13:47 |
| *** belharar_ has joined #openstack-neutron | 13:47 | |
| jlibosva | sean-k-mooney: lol | 13:48 |
| ajo | XDDDDDDDD | 13:48 |
| *** tommylikehu_ has quit IRC | 13:49 | |
| *** jpena is now known as jpena|lunch | 13:50 | |
| jschwarz | kevinbenton, replied again on https://review.openstack.org/#/c/413082/10 | 13:51 |
| *** pck has quit IRC | 13:52 | |
| *** pck has joined #openstack-neutron | 13:53 | |
| *** mdnadeem has quit IRC | 13:53 | |
| kevinbenton | jschwarz: so nova doesn't ensure the XML is correct on n-cpu start? | 13:53 |
| jschwarz | kevinbenton, yep | 13:53 |
| jschwarz | kevinbenton, it just syncs up the changes "for future possible references" | 13:53 |
| kevinbenton | jschwarz: and what happens on a nova reboot if the XML still references a bridge that no longer exists but its tap device is fine? | 13:55 |
| *** mdnadeem has joined #openstack-neutron | 13:55 | |
| *** sridharg has quit IRC | 13:55 | |
| jschwarz | kevinbenton, I don't *think* that it does anything | 13:56 |
| jschwarz | it only uses the domxml on certain events (like "start this VM") | 13:56 |
| *** ihrachys has joined #openstack-neutron | 13:56 | |
| jschwarz | kevinbenton, alas, that was my impression... | 13:57 |
| jschwarz | jlibosva or ajo might know more about this | 13:57 |
| *** baoli has joined #openstack-neutron | 13:58 | |
| openstackgerrit | Carlos Goncalves proposed openstack/neutron-lib: Adds API definition for data plane status extension https://review.openstack.org/424868 | 13:58 |
| *** bobmel has quit IRC | 13:58 | |
| *** marst has quit IRC | 13:58 | |
| jlibosva | I recall playing with that last February, so my memories are blurry. But I think on shutdown libvirt was unhappy that instance domxml contains a bridge that is not present on hypervisor. I think it tried to detach tap from that bridge or something. | 13:59 |
| *** panda is now known as notpython-panda | 13:59 | |
| openstackgerrit | Alex Stafeyev proposed openstack/neutron: Added Descriptions to scenario/base.py and test_basic.py functions https://review.openstack.org/425187 | 13:59 |
| *** baoli_ has joined #openstack-neutron | 14:00 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: ovs fw migration from iptables https://review.openstack.org/413082 | 14:03 |
| reedip_ | kevinbenton : hi, Is this being worked upon ?? I think you guys were discussing this earlier ? https://bugs.launchpad.net/neutron/+bug/1659265 | 14:03 |
| openstack | Launchpad bug 1659265 in neutron "Network level QoS policies should apply to network:router_gateway" [Undecided,New] | 14:03 |
| *** baoli has quit IRC | 14:03 | |
| *** notpython-panda is now known as panda | 14:05 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: ovs fw migration from iptables https://review.openstack.org/413082 | 14:06 |
| *** rkukura has joined #openstack-neutron | 14:07 | |
| *** pbandark has quit IRC | 14:11 | |
| *** amuller has joined #openstack-neutron | 14:11 | |
| *** tlian has joined #openstack-neutron | 14:12 | |
| *** zzzeek has quit IRC | 14:12 | |
| *** zzzeek has joined #openstack-neutron | 14:14 | |
| *** cleong has joined #openstack-neutron | 14:15 | |
| *** sdague has joined #openstack-neutron | 14:15 | |
| *** matrohon has joined #openstack-neutron | 14:15 | |
| *** ataraday_ has joined #openstack-neutron | 14:17 | |
| haleyb | kevinbenton: maybe you were thinking of floatingip_agent_gateway ^^ ? | 14:17 |
| *** dave-mccowan has joined #openstack-neutron | 14:17 | |
| *** jamesdenton has joined #openstack-neutron | 14:17 | |
| *** belharar_ has quit IRC | 14:18 | |
| kevinbenton | reedip: yes, i believe someone already is | 14:19 |
| kevinbenton | haleyb: yeah, that's the one :) | 14:19 |
| haleyb | kevinbenton: i don't exactly know the context, i just noticed 'dvr' :) | 14:19 |
| *** hfu has joined #openstack-neutron | 14:20 | |
| kevinbenton | haleyb: applying QoS policies to the ports of the router attaching to the external network | 14:20 |
| *** zkassab has joined #openstack-neutron | 14:20 | |
| haleyb | kevinbenton: ah, maybe related to the two qos patches i reviewed yesterday from liu ? | 14:20 |
| kevinbenton | haleyb: nah, this is just something that came up today i think | 14:21 |
| kevinbenton | haleyb: what were the two patches? | 14:21 |
| kevinbenton | haleyb: https://bugs.launchpad.net/neutron/+bug/1659265 | 14:21 |
| openstack | Launchpad bug 1659265 in neutron "Network level QoS policies should apply to network:router_gateway" [Undecided,New] | 14:21 |
| *** korzen has joined #openstack-neutron | 14:22 | |
| haleyb | kevinbenton: https://review.openstack.org/424466 was one, has a link to the other | 14:22 |
| openstackgerrit | Artur Korzeniewski proposed openstack/neutron: objects: get, update and delete converted to Subnet OVO usage https://review.openstack.org/321001 | 14:22 |
| anilvenkata | amuller, busy? | 14:22 |
| haleyb | kevinbenton: and i think the linked one to ^^ review is regarding router gateways, it was late when i was reviewing... | 14:23 |
| *** zkassab has quit IRC | 14:23 | |
| *** belharar_ has joined #openstack-neutron | 14:24 | |
| reedip_ | kevinbenton : ok :) | 14:24 |
| *** zkassab has joined #openstack-neutron | 14:25 | |
| kevinbenton | haleyb: ah, no, those are quite a bit different | 14:25 |
| kevinbenton | haleyb: this is addressing some warts of a fix that was added some time ago to skip network owned ports | 14:25 |
| kevinbenton | haleyb: the fixes for these bugs will just be a couple of line changes | 14:25 |
| haleyb | so they missed the snat port for example? | 14:26 |
| kevinbenton | haleyb: https://git.openstack.org/cgit/openstack/neutron/tree/neutron/objects/qos/rule.py?id=fe236bdaadb949661a0bfb9b62ddbe432b4cf5f1#n70 | 14:26 |
| openstackgerrit | Alex Stafeyev proposed openstack/neutron: Added Descriptions to scenario/base.py and test_basic.py functions https://review.openstack.org/425187 | 14:26 |
| kevinbenton | haleyb: no, they just blanket banned all network ports | 14:26 |
| kevinbenton | haleyb: unless the port policy ID matches the network | 14:26 |
| kevinbenton | haleyb: but that makes it so you can't have different policies per port | 14:27 |
| kevinbenton | haleyb: and it also means you can't set a qos policy for the entire external network as an operator and have it apply to router GW ports | 14:27 |
| amuller | anilvenkata: in meetings yeah | 14:27 |
| kevinbenton | haleyb: the original fix went in because people were getting confused when it was applied automatically to internal router interfaces on tenant networks | 14:28 |
| anilvenkata | amuller, ok | 14:28 |
| *** jperry has joined #openstack-neutron | 14:28 | |
| kevinbenton | haleyb: because from the tenant's perspective it was rate limiting their ingress traffic even though it was an egress policy | 14:28 |
| haleyb | kevinbenton: in the dvr floating ip case i would have thought the rule on the tenant port would have applied, but then i don't now how the qos policing is configured | 14:28 |
| kevinbenton | haleyb: nah, it's just skipping any network owned ports right now | 14:28 |
| haleyb | i will just look for the patches :) | 14:29 |
| kevinbenton | ack | 14:29 |
| Miouge | kevinbenton: is there a bug report for the second thing (port+network policy)? | 14:29 |
| kevinbenton | Miouge: i thought ralonsoh was going to file it? | 14:30 |
| ralonsoh | yes | 14:30 |
| ralonsoh | I'm on it | 14:30 |
| kevinbenton | thx | 14:30 |
| *** gouthamr has joined #openstack-neutron | 14:31 | |
| *** liangy has joined #openstack-neutron | 14:31 | |
| ihrachys | dasm: yes, grafana ideal is at 0; but if it's check queue, like in case of functional tests, you should expect some level of failures due to legit issues in patches posted for review | 14:32 |
| hishh | Hi, I'm using OS Newton. neutron-debug displays | 14:33 |
| hishh | ERROR neutron.common.utils [-] Alias or class name is not set | 14:33 |
| hishh | ERROR neutron.agent.common.utils [-] Error loading interface driver 'None' | 14:33 |
| hishh | What can I do to locate / fix this error? | 14:33 |
| openstackgerrit | Bertrand Lallau proposed openstack/neutron-fwaas: Fix scale issue in case of services_sync_needed https://review.openstack.org/424551 | 14:34 |
| openstackgerrit | Sindhu Devale proposed openstack/neutron: Refactoring agent linux&ovsdb config https://review.openstack.org/347867 | 14:35 |
| *** Swami has joined #openstack-neutron | 14:36 | |
| *** udesale has quit IRC | 14:36 | |
| haleyb | hishh: which agent? there should be an interface_driver line in both the l3 and dhcp agent ini files | 14:37 |
| *** xinliang has quit IRC | 14:40 | |
| openstackgerrit | Artur Korzeniewski proposed openstack/neutron: objects: get, update and delete converted to Subnet OVO usage https://review.openstack.org/321001 | 14:41 |
| *** liangy has quit IRC | 14:41 | |
| hishh | haleyb: l3: interface_driver = neutron.agent.linux.interface.BridgeInterfaceDriver | 14:41 |
| hishh | dhcp: dhcp_driver = neutron.agent.linux.dhcp.Dnsmasq | 14:41 |
| *** liangy has joined #openstack-neutron | 14:41 | |
| haleyb | hishh: which agent is not starting? | 14:42 |
| *** tongli has quit IRC | 14:42 | |
| *** gcb has quit IRC | 14:42 | |
| hishh | haleyb: good question, how do I check that? | 14:43 |
| haleyb | hishh: where did the ERROR come from? what file? | 14:43 |
| dasm | ihrachys: ack | 14:44 |
| hishh | haleyb: I just run: neutron-debug and see this messages | 14:44 |
| electrocucaracha | ihrachys: ping | 14:46 |
| jlibosva | hishh: haleyb I think neutron-debug is not maintained/tested. | 14:46 |
| ihrachys | electrocucaracha: pong | 14:46 |
| ihrachys | hishh: you may need to pass the config file that contains the interface_driver option to the tool via --config-file directive | 14:47 |
| *** korzen has quit IRC | 14:47 | |
| ihrachys | hishh: unless it's in neutron.conf itself | 14:47 |
| electrocucaracha | ihrachys: hey man, I've seen that many patches failed because they couldn't start 2 workers | 14:47 |
| electrocucaracha | ihrachys: http://logs.openstack.org/09/381209/19/check/gate-neutron-dsvm-functional-ubuntu-xenial/4c1e437/testr_results.html.gz | 14:47 |
| ihrachys | electrocucaracha: is it reported? | 14:47 |
| electrocucaracha | ihrachys: I didn't see any bug for that | 14:48 |
| hishh | ihrachys: ok thanks | 14:48 |
| *** bobmel has joined #openstack-neutron | 14:48 | |
| electrocucaracha | ihrachys: I just wanna confirm that it's a valid bug, it seems like it is | 14:49 |
| openstackgerrit | Swaminathan Vasudevan proposed openstack/neutron: DVR: Add static routes to FIP namespace https://review.openstack.org/308068 | 14:49 |
| ihrachys | electrocucaracha: seems like it, yes. I see it shows up yesterday in logstash | 14:50 |
| *** matrohon_ has joined #openstack-neutron | 14:50 | |
| hishh | haleyb: with both config files (l3 and dhcp) I get: "WARNING stevedore.named [-] Could not load neutron.agent.linux.interface.BridgeInterfaceDriver" | 14:51 |
| haleyb | hishh: that is just a warning, not the ERROR, could be somewhere else | 14:51 |
| *** bobmel_ has joined #openstack-neutron | 14:52 | |
| electrocucaracha | ihrachys: got it, I'm going to report it | 14:52 |
| ihrachys | haleyb: hishh: yea, I think kevinbenton had a patch lately removing those warnings | 14:53 |
| ihrachys | here it is https://review.openstack.org/#/c/418016/ | 14:53 |
| *** xinliang has joined #openstack-neutron | 14:53 | |
| *** matrohon has quit IRC | 14:53 | |
| *** bobmel has quit IRC | 14:53 | |
| *** kbringard has joined #openstack-neutron | 14:54 | |
| *** matrohon__ has joined #openstack-neutron | 14:55 | |
| *** mvk has quit IRC | 14:55 | |
| kevinbenton | ihrachys: that reminds me, what do you think about bringin that back to newton? | 14:55 |
| openstackgerrit | Maxime Guyot proposed openstack/neutron: Apply QoS policy on network:router_gateway https://review.openstack.org/425218 | 14:55 |
| jschwarz | kevinbenton, so I'm trying to understand if you're against the approach we took in https://review.openstack.org/#/c/413082/ or not :P | 14:56 |
| kevinbenton | jschwarz: i'm not a big fan :/ | 14:56 |
| kevinbenton | jschwarz: waiting to see if it grows on me | 14:56 |
| *** jpena|lunch is now known as jpena | 14:56 | |
| kevinbenton | jschwarz: i understand that it solves an issue that needs to be fixed | 14:57 |
| *** marst has joined #openstack-neutron | 14:57 | |
| kevinbenton | jschwarz: but i really don't like doing nova things | 14:57 |
| jschwarz | kevinbenton, aye, I can relate | 14:57 |
| *** matrohon_ has quit IRC | 14:57 | |
| jschwarz | kevinbenton, the alternative is put this on nova's side, but then they'll say that they don't like doing neutron things | 14:58 |
| kevinbenton | jschwarz: they already do this | 14:58 |
| kevinbenton | jschwarz: it's vif pluggin | 14:58 |
| jschwarz | uhm | 14:58 |
| electrocucaracha | ihrachys: https://bugs.launchpad.net/neutron/+bug/1659321 | 14:58 |
| openstack | Launchpad bug 1659321 in neutron "Failed to start 2 RPC workers" [Undecided,New] | 14:58 |
| jlibosva | electrocucaracha: it already has a bug and fix in gate | 14:59 |
| jschwarz | kevinbenton, well we can start doing Nova stuff and this will be the first ;-) | 14:59 |
| jlibosva | https://review.openstack.org/#/c/424906/ | 15:00 |
| jlibosva | ihrachys: ^^ | 15:00 |
| electrocucaracha | jlibosva: ohh,. thanks | 15:00 |
| *** udesale has joined #openstack-neutron | 15:00 | |
| *** catinthe_ has joined #openstack-neutron | 15:00 | |
| ihrachys | jlibosva: ack | 15:00 |
| *** mlavalle has joined #openstack-neutron | 15:00 | |
| *** sridharg has joined #openstack-neutron | 15:01 | |
| amuller | jschwarz: kevinbenton: Would Nova be ok with sending RPC to neutron-server..? | 15:01 |
| amuller | or how would nova update the neutron DB? | 15:01 |
| *** catintheroof has quit IRC | 15:02 | |
| kevinbenton | amuller: they call port update on neutron as part of normal VM interface operations | 15:02 |
| amuller | jschwarz: I thought there was a field that we could not update via the API | 15:03 |
| jschwarz | yep, just what I was about to say | 15:03 |
| kevinbenton | jschwarz, amuller: nova wouldn't update that field | 15:03 |
| *** jhershbe_ has quit IRC | 15:03 | |
| kevinbenton | it would just trigger re-binding of the prot | 15:03 |
| kevinbenton | port* | 15:03 |
| kevinbenton | the OVS mech driver on binding would see that its agent doesn't need hybrid plugging anymore | 15:03 |
| kevinbenton | and set that flag to false on the port | 15:04 |
| jschwarz | kevinbenton, that has the disadvantage of potentially taking a long time to run | 15:04 |
| jschwarz | kevinbenton, the time it takes for the nova api to process the requests and then the neutron-server to send it the the ovs agent, and then the ovs agent to reach that port... | 15:05 |
| *** fnaval has joined #openstack-neutron | 15:05 | |
| *** mkolesni has quit IRC | 15:05 | |
| jschwarz | kevinbenton, during that time, will the VM still have connectivity? | 15:05 |
| kevinbenton | jschwarz: it depends on how the rebinding happens | 15:06 |
| hishh | How can I locate this error (msg in neutron-server.log while creating a kubernetes cluster with magnum): | 15:06 |
| hishh | INFO neutron.api.v2.resource [req-47305c33-c5d6-4894-bf99-3a7c7e93a33d 8446e842ba864101a9422277d30b1a69 a70a03a59c5d4631 | 15:06 |
| hishh | 844f95fb091b33b9 - - -] show failed (client error): The resource could not be found. | 15:06 |
| kevinbenton | jschwarz: if we get it so its one operation without an intermediary unbound state, then yes | 15:06 |
| kevinbenton | jschwarz: it will have connectivity the whole time | 15:06 |
| kevinbenton | jschwarz: presumably the nova side would be written to not rip out the old stuff until it sees that it needs to make a change because the binding details changed | 15:08 |
| kevinbenton | jschwarz: so the part doing the switching would still be quick | 15:08 |
| kevinbenton | jschwarz: but nova is well frozen for this cycle | 15:08 |
| jschwarz | kevinbenton, aye, that sounds like a proper cross-project development process | 15:09 |
| jschwarz | amuller, ^ | 15:09 |
| kevinbenton | jschwarz: well it's just a bunch of work that needs to be done in nova | 15:09 |
| kevinbenton | jschwarz: not much to do on the neutron side if anything at all | 15:09 |
| amuller | it'd be a shame for this work to be bumped to the next cycle | 15:09 |
| *** jaosorior has joined #openstack-neutron | 15:10 | |
| openstackgerrit | Maxime Guyot proposed openstack/neutron: Apply QoS policy on network:router_gateway https://review.openstack.org/425218 | 15:10 |
| *** hfu has quit IRC | 15:11 | |
| kevinbenton | jschwarz: are you absolutely sure we have to mess with this XML? | 15:11 |
| jschwarz | kevinbenton, it looks like it | 15:11 |
| sean-k-mooney | as far as i am aware the main reason nova does the port update when launcing a vm is to set the hostid in the portbindgs | 15:12 |
| kevinbenton | jschwarz: what happens if we don't though, i didn't get that | 15:12 |
| *** sdague has quit IRC | 15:12 | |
| *** belharar__ has joined #openstack-neutron | 15:12 | |
| *** belharar_ has quit IRC | 15:12 | |
| kevinbenton | sean-k-mooney: right, i'm hoping nova can add something to ask neutron to just rebind the port to the same host | 15:12 |
| kevinbenton | sean-k-mooney: to see if it needs to rewire the port | 15:12 |
| jschwarz | kevinbenton, basically, stuff that depend on domain.xml being right, but that nova doesn't update, will not work correctly | 15:13 |
| jschwarz | kevinbenton, stuff like 'nova reboot' afaik | 15:13 |
| jschwarz | granted, I haven't dug deep into this to understand the exact implications | 15:13 |
| kevinbenton | jschwarz: is there something we can tell nova to do that will force it to flush changes to the XML? | 15:13 |
| sean-k-mooney | kevinbenton: when you say needs to rewire the port you mean nova recall plug on os-vif | 15:14 |
| sean-k-mooney | well wehre we have swaped to os-vif that is | 15:14 |
| jschwarz | kevinbenton, I asked around some nova guys - doesn't look like it | 15:14 |
| jschwarz | kevinbenton, there is the periodic job that we can just call when n-cpu starts, but that's a patch for nova and I haven't looked at yet | 15:14 |
| kevinbenton | sean-k-mooney: sort of. this case might be too narrow to convince nova to pick up anyway | 15:15 |
| kevinbenton | sean-k-mooney: basically unplug and replug | 15:15 |
| sean-k-mooney | kevinbenton: what is the usecase you are trying to enable? | 15:15 |
| kevinbenton | sean-k-mooney: but without removing the interface from the guest itself | 15:15 |
| kevinbenton | sean-k-mooney: this particular one is removing the linux filtering bridge from OVS | 15:16 |
| kevinbenton | sean-k-mooney: for migration to native OVS firewall | 15:16 |
| sean-k-mooney | ah i see | 15:16 |
| sean-k-mooney | well if the nova-compute agent is restarted it calls plug on all interfaces to ensure the connectivity is still correct | 15:17 |
| *** janzian has joined #openstack-neutron | 15:17 | |
| jschwarz | sean-k-mooney, does it do this based on Neutron information, or based on Nova information? | 15:17 |
| *** jlibosva has quit IRC | 15:17 | |
| *** manheim has joined #openstack-neutron | 15:17 | |
| kevinbenton | sean-k-mooney: so it fetches fresh data from neutron and then calls plug? | 15:17 |
| openstackgerrit | Yushiro FURUKAWA proposed openstack/neutron-fwaas: Enable to filter correctly with 'public' https://review.openstack.org/424534 | 15:17 |
| sean-k-mooney | in the case of ovs that dose a del port followed by an add port in an atomic transaction | 15:17 |
| *** jlibosva has joined #openstack-neutron | 15:17 | |
| jschwarz | sean-k-mooney, because, from what I can tell, Nova doesn't sync this information immediately, it takes it around 60 seconds to run a periodic function that updates it | 15:17 |
| sean-k-mooney | jschwarz: korzen_ i belive it gets it from neutron | 15:18 |
| jschwarz | sean-k-mooney, that's not what I saw last week when I dug really deep into this :( | 15:18 |
| sean-k-mooney | sorry kevinbenton not korzen_ | 15:18 |
| sean-k-mooney | so i am basing this on https://review.openstack.org/#/c/408779/ | 15:19 |
| *** mhickey has quit IRC | 15:19 | |
| sean-k-mooney | here we extended os vif so that it would fix the mtus if you had changed them | 15:19 |
| *** manheim has quit IRC | 15:19 | |
| sean-k-mooney | this was for the case where you upgraded and the mtu changed which i belive was the case for trippleo or packstack | 15:20 |
| *** belharar__ has quit IRC | 15:20 | |
| *** manheim has joined #openstack-neutron | 15:20 | |
| *** belharar_ has joined #openstack-neutron | 15:20 | |
| sean-k-mooney | the mtu in that case is coming from neutron | 15:20 |
| jschwarz | sean-k-mooney, sorry, but doesn't that only retrieves the mtu? | 15:21 |
| sean-k-mooney | it retives them and then updates the mtu on the veth pair when using hybrid plug | 15:21 |
| sean-k-mooney | but that is done by nova when it creates teh os-vif object | 15:21 |
| kevinbenton | sean-k-mooney: right. i think the issue is that nova only periodicically retrieves this network data | 15:22 |
| jschwarz | sean-k-mooney, the periodic I saw was https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L5805 | 15:22 |
| sean-k-mooney | kevinbenton: correct but it does it every time the nova-compute agent is restarted | 15:22 |
| jschwarz | and that's the only thing I saw updating the cache | 15:22 |
| kevinbenton | jschwarz: if that's done on restart, your patch could become as simple as updating the state in the DB via RPC and then letting os-vif do the rest on nova compute restart | 15:23 |
| kevinbenton | jschwarz: seems too good to be true :) | 15:23 |
| jschwarz | kevinbenton, that's a nice "if", but yes, I agree | 15:24 |
| kevinbenton | sean-k-mooney: where is the logic that unplugs/plugs? | 15:24 |
| *** manheim has quit IRC | 15:24 | |
| openstackgerrit | Carlos Goncalves proposed openstack/neutron: Port data plane status extension https://review.openstack.org/424340 | 15:25 |
| sean-k-mooney | kevinbenton: that is the rinkel. non hybrid plug is done via libvirt not os-vif | 15:25 |
| sean-k-mooney | kevinbenton: i have wanted to move it to os-vif for a while | 15:25 |
| kevinbenton | ah | 15:25 |
| jschwarz | kevinbenton, how about we merge the patch as-is, and work in Pike to add support for this on Nova's side (practically eliminating most of the code when Nova supports it)? | 15:26 |
| sean-k-mooney | basically libvirt supprot adding the tap to an ovs bridge but it cant handel vhost-user or hybrid plug | 15:26 |
| kevinbenton | jschwarz: sure | 15:26 |
| sean-k-mooney | in os-vif its currently a noop for the non hybird plug when not useing vhost-user | 15:26 |
| kevinbenton | jschwarz: i'm sure we'll get it done right after microversioning and refactoring ML2 :) | 15:27 |
| jschwarz | kevinbenton, ;-) | 15:27 |
| kevinbenton | sean-k-mooney: what's the pushback on moving that to os-vif? | 15:27 |
| sean-k-mooney | kevinbenton: jschwarz well if you want some help in this areay on the os-vif/nova side i can help | 15:27 |
| *** rajinir has joined #openstack-neutron | 15:27 | |
| sean-k-mooney | kevinbenton: danail berange perfered to let libvirt handel it and there was no usecase that need it so we just left it as is | 15:28 |
| kevinbenton | sean-k-mooney: well if we can get it to where os-vif will unplug the old and plug the new at startup, i think that will give us an easy path to do this | 15:28 |
| sean-k-mooney | kevinbenton: my only concern is cleanup of the linux bridge | 15:29 |
| kevinbenton | sean-k-mooney: doesn't the unplug call clean it? | 15:29 |
| *** baoli_ has quit IRC | 15:29 | |
| sean-k-mooney | it would be a different branch based on the presents of hybrid plug=ture | 15:29 |
| sean-k-mooney | but it does not have to be | 15:29 |
| kevinbenton | sean-k-mooney: ah, right. it would be trying to unplug based on the new data | 15:30 |
| sean-k-mooney | kevinbenton: exactly so it would call the unplu for non hybrid plug which is jsut ovs-vsctl del-port | 15:30 |
| *** nyechiel_ has quit IRC | 15:30 | |
| sean-k-mooney | but we could converge the branch and have it delete the bridge and vif pair if present assuming os-vif was extended to handel the non hybrid plug case | 15:31 |
| *** markvoelker has quit IRC | 15:31 | |
| *** jlibosva has quit IRC | 15:31 | |
| *** markvoelker has joined #openstack-neutron | 15:32 | |
| kevinbenton | jschwarz, sean-k-mooney: maybe making use of the new multiple bindings thing would be the correct long term way forward | 15:32 |
| *** jlibosva has joined #openstack-neutron | 15:32 | |
| kevinbenton | so we have an active hybrid binding | 15:33 |
| kevinbenton | create the native inactive binding | 15:33 |
| kevinbenton | switch the native to active | 15:33 |
| jschwarz | kevinbenton, isn't that was andreas_s talked about on the patch? | 15:33 |
| kevinbenton | jschwarz: right | 15:33 |
| jschwarz | kevinbenton, that's also for... Pike, iirc? | 15:33 |
| kevinbenton | jschwarz: yeah, that's why it's "long term" | 15:34 |
| sean-k-mooney | kevinbenton: currently though as it stand to go from iptabels to ovs firewall requires nova to regenerate the libvirt xml which would basicaly require a live resize/migrate to the same flavour/host | 15:34 |
| jschwarz | kevinbenton, ah :) | 15:34 |
| jschwarz | sean-k-mooney, well actually | 15:34 |
| kevinbenton | sean-k-mooney: wait until you see the new nova support inside of neutron! | 15:34 |
| jschwarz | sean-k-mooney, once we update the neutron Port and pass the hybrid_plug=False to nova, it seems to generate the correct xml on its own | 15:34 |
| jschwarz | kevinbenton, we have nova-network, why not have neutron-compute?! | 15:35 |
| sean-k-mooney | jschwarz: well yes it will but nova only generage the xml on server boot,resize or migrate | 15:35 |
| *** sridharg has quit IRC | 15:35 | |
| jschwarz | sean-k-mooney, does it need to actually update the xml if the VM is left running? | 15:36 |
| *** markvoelker has quit IRC | 15:36 | |
| jschwarz | i.e. if I don't resize/migrate/boot the VM, does libvirt care if its' active domain xml is wrong? | 15:36 |
| sean-k-mooney | to be a good citizen with libvirt yes. | 15:36 |
| jschwarz | sean-k-mooney, Make Libvirt Great Again? | 15:36 |
| sean-k-mooney | haha i belive libvirt allows some online updates to the xml also. | 15:37 |
| jschwarz | if there is, I didn't find it | 15:37 |
| jschwarz | all I found was virsh define, but that applies to the inactive xml | 15:37 |
| sean-k-mooney | but it would result with a change in the network interface definition. | 15:38 |
| sean-k-mooney | jschwarz: there is a virsh edit | 15:38 |
| *** rmart04 has quit IRC | 15:38 | |
| jschwarz | sean-k-mooney, virsh edit is interactive (vi, etc) | 15:38 |
| jschwarz | we need something to feed on a file | 15:38 |
| *** manheim has joined #openstack-neutron | 15:38 | |
| *** jhershbe has joined #openstack-neutron | 15:39 | |
| jschwarz | kevinbenton, https://www.youtube.com/watch?v=f_4-rCROcsM | 15:40 |
| jschwarz | XD | 15:40 |
| *** jschwarz is now known as jschwarz|brb | 15:41 | |
| *** ekuris_ has quit IRC | 15:41 | |
| sean-k-mooney | jschwarz: idealy waht we need is a api in nova to notify it that the network has changed or generically that some property of the vm has changed | 15:41 |
| sean-k-mooney | that is alot more work though for somethign that will only be used on upgrade | 15:42 |
| *** st8less has joined #openstack-neutron | 15:42 | |
| *** catintheroof has joined #openstack-neutron | 15:43 | |
| sean-k-mooney | jschwarz|brb: kevinbenton one topic i put in the nova track for the ptg is os-vif intergration with neutron which clearly should be a cross project thing | 15:43 |
| sean-k-mooney | is there time on the neutron track to touch on this | 15:44 |
| *** jose-phillips has joined #openstack-neutron | 15:44 | |
| dasm | ihrachys: i believe newton release requires minor bump https://review.openstack.org/#/c/424224/ because we have at least this feature: https://review.openstack.org/#/c/412462/ | 15:46 |
| *** catinthe_ has quit IRC | 15:46 | |
| *** gcheresh_ has quit IRC | 15:46 | |
| dasm | ihrachys: i didn't mention it in highlights, so i'll add it. | 15:46 |
| *** Swami_ has joined #openstack-neutron | 15:46 | |
| dasm | ihrachys: what do you think about that? worth bumping minor ver or not? | 15:47 |
| kevinbenton | sean-k-mooney: yeah. that would be a good PTG topic | 15:47 |
| *** jschwarz|brb is now known as jschwarz | 15:48 | |
| kevinbenton | jschwarz|brb: find some people to +2 it and i'll just look the other way holding my nose :) | 15:48 |
| jschwarz | kevinbenton, some good :) | 15:49 |
| jschwarz | jlibosva, rossella_, ;-) | 15:49 |
| *** tbachman has quit IRC | 15:49 | |
| *** Swami has quit IRC | 15:49 | |
| rossella_ | jschwarz, lol | 15:49 |
| jschwarz | rossella_, ^_^ | 15:50 |
| jschwarz | rossella_, the latest revision has tests for most of the code (the migration/main.py is mostly "call this function and then that function" so it wasn't worth putting it through the ringer, but the rest has tests) | 15:50 |
| jschwarz | rossella_, if there are any issues I can upload new patchsets | 15:51 |
| *** Sukhdev has joined #openstack-neutron | 15:51 | |
| rossella_ | jschwarz, I will look at it tomorrow, today I have my plate full with other stuff, sorry | 15:51 |
| jschwarz | rossella_, ack, thanks :) | 15:51 |
| openstackgerrit | Maxime Guyot proposed openstack/neutron: Apply QoS policy on network:router_gateway https://review.openstack.org/425218 | 15:52 |
| *** mhickey has joined #openstack-neutron | 15:53 | |
| *** ThiagoCMC has joined #openstack-neutron | 15:53 | |
| *** jchhatbar has quit IRC | 15:55 | |
| *** caboucha has joined #openstack-neutron | 15:55 | |
| *** belharar_ has quit IRC | 15:56 | |
| openstackgerrit | Henry Gessau proposed openstack/neutron: [WIP] Operation Pig Bristle, part 7 https://review.openstack.org/409557 | 15:58 |
| openstackgerrit | Henry Gessau proposed openstack/neutron: [WIP] Operation Pig Bristle, part 5 https://review.openstack.org/409556 | 15:58 |
| openstackgerrit | Henry Gessau proposed openstack/neutron: [WIP] Operation Pig Bristle, part 4 https://review.openstack.org/409555 | 15:58 |
| openstackgerrit | Henry Gessau proposed openstack/neutron: [WIP] Operation Pig Bristle, part 3 https://review.openstack.org/409554 | 15:58 |
| openstackgerrit | Henry Gessau proposed openstack/neutron: Decouple hook and func registration from CommonDbMixin https://review.openstack.org/409553 | 15:58 |
| openstackgerrit | Henry Gessau proposed openstack/neutron: [WIP] Operation Pig Bristle, part 6 https://review.openstack.org/407837 | 15:58 |
| *** bobmel_ has quit IRC | 15:58 | |
| *** bobmel has joined #openstack-neutron | 15:59 | |
| *** abregman has quit IRC | 15:59 | |
| *** abregman has joined #openstack-neutron | 16:01 | |
| *** jhershbe has quit IRC | 16:02 | |
| *** dsneddon_afk is now known as dsneddon | 16:02 | |
| *** sdague has joined #openstack-neutron | 16:03 | |
| *** gongysh has quit IRC | 16:07 | |
| *** anilvenkata has quit IRC | 16:07 | |
| *** eezhova has joined #openstack-neutron | 16:07 | |
| *** itzikb_ has quit IRC | 16:12 | |
| *** mdnadeem has quit IRC | 16:13 | |
| *** gcheresh_ has joined #openstack-neutron | 16:13 | |
| *** ventifus has joined #openstack-neutron | 16:15 | |
| *** gvrangan has joined #openstack-neutron | 16:15 | |
| *** tbachman has joined #openstack-neutron | 16:16 | |
| ThiagoCMC | Hey guys, is it possible to create an Instance on OpenStack, with 2 Networks, first (eth0) will be on top of Linux Bridges Agent, the second (eth1), will be on top of OpenvSwitch Agent. | 16:17 |
| *** markvoelker has joined #openstack-neutron | 16:17 | |
| *** armax has joined #openstack-neutron | 16:18 | |
| *** markvoelker_ has joined #openstack-neutron | 16:18 | |
| *** markvoelker has quit IRC | 16:22 | |
| *** manheim has quit IRC | 16:24 | |
| *** mvk has joined #openstack-neutron | 16:25 | |
| *** efried has quit IRC | 16:25 | |
| *** manheim has joined #openstack-neutron | 16:25 | |
| *** amarao has quit IRC | 16:25 | |
| *** TMM has quit IRC | 16:25 | |
| *** Leo_ has joined #openstack-neutron | 16:26 | |
| *** udesale has quit IRC | 16:26 | |
| *** Leo_ has quit IRC | 16:27 | |
| *** annegentle has joined #openstack-neutron | 16:29 | |
| mlavalle | dasm: ping | 16:29 |
| *** manheim has quit IRC | 16:29 | |
| *** shaner has joined #openstack-neutron | 16:31 | |
| *** lihi has quit IRC | 16:31 | |
| amuller | jlibosva: John and I have a question for the testing lieutenant | 16:32 |
| *** ankur-gupta-f2 has joined #openstack-neutron | 16:32 | |
| openstackgerrit | Thomas Morin proposed openstack/neutron: Add OVS agent bridge extensions https://review.openstack.org/425074 | 16:33 |
| amuller | jlibosva: For the OVS FW migration script, is there any infrastructure that will allow us to test it for real? | 16:33 |
| jlibosva | amuller: jschwarz tell me | 16:33 |
| jlibosva | amuller: no :) | 16:33 |
| jschwarz | XD | 16:33 |
| amuller | seems like tempest is out of the question... we need to do things like shut down nova compute | 16:33 |
| amuller | fullstack is only so so here | 16:33 |
| *** Leo_ has joined #openstack-neutron | 16:33 | |
| jlibosva | amuller: yeah, we talked a bit about that with jschwarz and fullstack is not using libvirt | 16:34 |
| jschwarz | jlibosva, we could simulate a neutron-virsh for fullstack to simulate a FakeMachine | 16:34 |
| amuller | jlibosva: maybe we could test it in upstream tripleo though | 16:34 |
| jlibosva | amuller: maybe a grenade could be an option. just instead of upgrade changing config file and running script | 16:34 |
| amuller | jlibosva: we'll need to add support for this in tripleo anyway | 16:35 |
| *** AlexeyAbashkin has quit IRC | 16:35 | |
| amuller | we could run that migration in the upstream tripleo gate, maybe | 16:35 |
| jlibosva | amuller: I'm not familiar with their testing env so I can't say :-/ | 16:36 |
| *** ushkalim has quit IRC | 16:36 | |
| amuller | jlibosva: I think we'll have to become familiar with it =[p | 16:36 |
| *** morgabra has quit IRC | 16:36 | |
| amuller | =p | 16:36 |
| jlibosva | amuller: and by "we" you mean "me"? :) | 16:37 |
| amuller | jlibosva: maybe, dunno yet =p | 16:38 |
| *** princenana has joined #openstack-neutron | 16:40 | |
| *** abregman has quit IRC | 16:40 | |
| openstackgerrit | Merged openstack/neutron-fwaas: Netlink solution to improve FWaaS performance https://review.openstack.org/389654 | 16:40 |
| *** moshele has quit IRC | 16:41 | |
| *** oreillyd_ has quit IRC | 16:42 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: ovs fw migration from iptables https://review.openstack.org/413082 | 16:44 |
| *** radhikam has joined #openstack-neutron | 16:47 | |
| electrocucaracha | kevinbenton: ping | 16:47 |
| dasm | mlavalle: pong | 16:48 |
| mlavalle | dasm: I found what I was looking for (the patchset for the workers issue in the functional tests). Thanks for fixing it! | 16:49 |
| dasm | mlavalle: good you have it :) | 16:50 |
| *** bobmel_ has joined #openstack-neutron | 16:50 | |
| dasm | but it was dirk's and dims's job :) | 16:50 |
| openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/neutron: Enforce port QoS policies for all ports. https://review.openstack.org/425280 | 16:51 |
| *** gvrangan has quit IRC | 16:51 | |
| *** bobmel has quit IRC | 16:51 | |
| *** ThiagoCMC has quit IRC | 16:52 | |
| *** bfernando has joined #openstack-neutron | 16:52 | |
| *** andreas_s has quit IRC | 16:53 | |
| *** leitan has joined #openstack-neutron | 16:54 | |
| openstackgerrit | Shashank Kumar Shankar proposed openstack/neutron: Integration of Port Binding Level OVO https://review.openstack.org/382037 | 16:56 |
| openstackgerrit | tonytan4ever proposed openstack/neutron: Support for Agent Object for filtering on "LIKE" statement https://review.openstack.org/419152 | 16:56 |
| *** tonytan4ever has joined #openstack-neutron | 16:56 | |
| *** nagarjung has joined #openstack-neutron | 16:59 | |
| ltomasbo | ping armax: can you give me more details about your comments in https://review.openstack.org/#/c/421880 | 16:59 |
| *** markvoelker_ has quit IRC | 16:59 | |
| *** mickeys has joined #openstack-neutron | 16:59 | |
| *** bobmel_ has quit IRC | 16:59 | |
| *** markvoelker has joined #openstack-neutron | 16:59 | |
| *** Miouge has quit IRC | 16:59 | |
| bfernando | hello someone knows the project where keywords for the policy.json are defined | 17:01 |
| bfernando | ? | 17:03 |
| *** markvoelker has quit IRC | 17:03 | |
| *** bobmel has joined #openstack-neutron | 17:03 | |
| *** panda is now known as panda|bbl | 17:04 | |
| *** mickeys has quit IRC | 17:05 | |
| *** Jeffrey4l_ has quit IRC | 17:06 | |
| *** liangy has quit IRC | 17:07 | |
| *** jemcevoy has quit IRC | 17:11 | |
| njohnston | bfernando: do you mean https://github.com/openstack/oslo.policy ? | 17:13 |
| bfernando | njohnston, perfect thx ;D | 17:15 |
| *** gcheresh_ has quit IRC | 17:15 | |
| *** hishh has quit IRC | 17:18 | |
| *** manheim has joined #openstack-neutron | 17:20 | |
| *** jaosorior has quit IRC | 17:22 | |
| *** ataraday_ has quit IRC | 17:22 | |
| *** reedip_ has quit IRC | 17:23 | |
| *** ramishra_ has joined #openstack-neutron | 17:23 | |
| *** jheroux has joined #openstack-neutron | 17:24 | |
| *** ramishra has quit IRC | 17:25 | |
| *** Jeffrey4l has joined #openstack-neutron | 17:25 | |
| *** manheim has quit IRC | 17:25 | |
| openstackgerrit | Kevin Benton proposed openstack/neutron: Fix a bad docstring in provisioning blocks module https://review.openstack.org/425308 | 17:26 |
| *** manheim has joined #openstack-neutron | 17:26 | |
| *** fzdarsky is now known as fzdarsky|afk | 17:27 | |
| *** Swami_ has quit IRC | 17:27 | |
| *** iyamahat has joined #openstack-neutron | 17:29 | |
| *** Jeffrey4l has quit IRC | 17:30 | |
| *** ygbo has quit IRC | 17:30 | |
| *** manheim has quit IRC | 17:30 | |
| *** rossella_ has quit IRC | 17:31 | |
| *** jaosorior has joined #openstack-neutron | 17:32 | |
| *** anilvenkata has joined #openstack-neutron | 17:33 | |
| *** tesseract has quit IRC | 17:33 | |
| *** karts has quit IRC | 17:34 | |
| *** baoli has joined #openstack-neutron | 17:35 | |
| *** princenana has quit IRC | 17:35 | |
| *** kevo has joined #openstack-neutron | 17:36 | |
| *** aarefiev is now known as aarefiev_afk | 17:37 | |
| *** tonytan4ever has quit IRC | 17:39 | |
| *** tonytan4ever has joined #openstack-neutron | 17:40 | |
| openstackgerrit | Jakub Libosvar proposed openstack/neutron: DNM: Try to increase ssh timeout for ubuntu images https://review.openstack.org/425165 | 17:40 |
| *** bobmel has quit IRC | 17:42 | |
| *** mhickey has quit IRC | 17:44 | |
| *** tbarron has joined #openstack-neutron | 17:44 | |
| *** mmalik4 has joined #openstack-neutron | 17:45 | |
| *** Swami has joined #openstack-neutron | 17:46 | |
| *** tbachman has quit IRC | 17:46 | |
| *** nplanel has quit IRC | 17:47 | |
| *** davidsha has quit IRC | 17:47 | |
| *** iyamahat has quit IRC | 17:48 | |
| dims | dasm : thanks for sharing the credit LOL | 17:49 |
| *** lucasagomes is now known as lucas-afk | 17:54 | |
| *** caboucha has quit IRC | 17:54 | |
| *** jlibosva has quit IRC | 17:54 | |
| manjeets | ihrachys, ping ( reg https://review.openstack.org/#/c/353088/35/neutron/services/auto_allocate/db.py@226) | 17:55 |
| manjeets | to ensure order can we have an optional argument if set we can order by network_id to ensure the order of networks returned | 17:56 |
| *** ralonsoh has quit IRC | 17:58 | |
| *** beagles is now known as beagles-biab | 17:59 | |
| *** sam_s has joined #openstack-neutron | 18:02 | |
| *** sdague has quit IRC | 18:03 | |
| *** bobmel has joined #openstack-neutron | 18:06 | |
| *** tonytan_brb has joined #openstack-neutron | 18:06 | |
| *** tonytan4ever has quit IRC | 18:07 | |
| *** eezhova has quit IRC | 18:08 | |
| openstackgerrit | Merged openstack/neutron-lib: Add action map for neutron-fwaas API definition https://review.openstack.org/421534 | 18:10 |
| *** matrohon__ has quit IRC | 18:10 | |
| openstackgerrit | Manjeet Singh Bhatia proposed openstack/neutron: OVO External Networks https://review.openstack.org/353088 | 18:13 |
| ihrachys | manjeets: we may as well always apply the order in get? | 18:13 |
| *** jlinkes has joined #openstack-neutron | 18:18 | |
| *** manheim has joined #openstack-neutron | 18:19 | |
| *** mickeys has joined #openstack-neutron | 18:21 | |
| *** st8less has quit IRC | 18:22 | |
| manjeets | ihrachys, we can do that was thinking of using sql_alchemy's order_by but after looking at _get_collection i think its better to do in base | 18:24 |
| *** efoley_ has quit IRC | 18:26 | |
| *** jlibosva has joined #openstack-neutron | 18:27 | |
| *** abhiraut has joined #openstack-neutron | 18:30 | |
| openstackgerrit | Victor Morales proposed openstack/neutron: Change _VALID_CLS to store obj names https://review.openstack.org/425332 | 18:32 |
| *** catinthe_ has joined #openstack-neutron | 18:33 | |
| *** korzen has joined #openstack-neutron | 18:34 | |
| *** sambetts is now known as sambetts|afk | 18:34 | |
| *** catintheroof has quit IRC | 18:35 | |
| *** Sukhdev has quit IRC | 18:35 | |
| *** Sukhdev has joined #openstack-neutron | 18:36 | |
| *** trown is now known as trown|lunch | 18:37 | |
| *** wolverineav has quit IRC | 18:39 | |
| *** wolverineav has joined #openstack-neutron | 18:39 | |
| *** gcheresh_ has joined #openstack-neutron | 18:40 | |
| *** Jeffrey4l has joined #openstack-neutron | 18:41 | |
| *** mvk has quit IRC | 18:42 | |
| *** wolverineav has quit IRC | 18:43 | |
| *** yamahata has joined #openstack-neutron | 18:43 | |
| *** iyamahat has joined #openstack-neutron | 18:43 | |
| *** korzen has quit IRC | 18:44 | |
| *** jpena is now known as jpena|off | 18:45 | |
| *** Jeffrey4l has quit IRC | 18:46 | |
| *** Leo_ has quit IRC | 18:47 | |
| *** sean-k-mooney has quit IRC | 18:47 | |
| *** sean-k-mooney has joined #openstack-neutron | 18:48 | |
| *** gvrangan has joined #openstack-neutron | 18:53 | |
| *** pcaruana has quit IRC | 18:53 | |
| *** tbachman has joined #openstack-neutron | 18:54 | |
| *** tonytan4ever has joined #openstack-neutron | 18:56 | |
| *** radhikam has quit IRC | 18:57 | |
| *** nagarjung has quit IRC | 18:57 | |
| *** tonytan_brb has quit IRC | 18:58 | |
| *** radhikam has joined #openstack-neutron | 18:58 | |
| *** tbachman has quit IRC | 18:59 | |
| *** tbachman has joined #openstack-neutron | 18:59 | |
| *** sdague has joined #openstack-neutron | 19:00 | |
| *** sshank_ has quit IRC | 19:00 | |
| *** vjmorale has quit IRC | 19:01 | |
| *** jlinkes has quit IRC | 19:01 | |
| *** sshank has joined #openstack-neutron | 19:02 | |
| *** nicolasbock has quit IRC | 19:03 | |
| *** vjmorale has joined #openstack-neutron | 19:04 | |
| *** esmiurium has quit IRC | 19:05 | |
| *** ociuhandu has quit IRC | 19:07 | |
| *** abhiraut has quit IRC | 19:08 | |
| *** jlinkes has joined #openstack-neutron | 19:08 | |
| *** MasterOfBugs has joined #openstack-neutron | 19:08 | |
| *** ijw has joined #openstack-neutron | 19:08 | |
| *** manheim has quit IRC | 19:08 | |
| prometheanfire | do routers support dnat? | 19:08 |
| *** MasterOfBugs has quit IRC | 19:08 | |
| *** ijw has quit IRC | 19:09 | |
| *** manheim has joined #openstack-neutron | 19:09 | |
| *** MasterOfBugs has joined #openstack-neutron | 19:09 | |
| *** ijw has joined #openstack-neutron | 19:09 | |
| *** pramodrj07 has joined #openstack-neutron | 19:09 | |
| *** abhiraut has joined #openstack-neutron | 19:09 | |
| ijw | dhellmann: you about? | 19:09 |
| *** Jeffrey4l has joined #openstack-neutron | 19:10 | |
| *** mlavalle has quit IRC | 19:10 | |
| *** moshele has joined #openstack-neutron | 19:11 | |
| prometheanfire | says it's obsolete :| https://blueprints.launchpad.net/neutron/+spec/router-port-forwarding | 19:11 |
| *** liangy has joined #openstack-neutron | 19:11 | |
| *** esmiurium has joined #openstack-neutron | 19:12 | |
| *** mlavalle has joined #openstack-neutron | 19:12 | |
| *** Sukhdev has quit IRC | 19:13 | |
| *** Jeffrey4l has quit IRC | 19:15 | |
| *** anilvenkata has quit IRC | 19:15 | |
| *** moshele has quit IRC | 19:18 | |
| *** manheim has quit IRC | 19:18 | |
| *** abhiraut has quit IRC | 19:18 | |
| *** manheim has joined #openstack-neutron | 19:19 | |
| *** jaosorior has quit IRC | 19:21 | |
| *** manheim has quit IRC | 19:23 | |
| *** Leo_ has joined #openstack-neutron | 19:25 | |
| *** abhiraut has joined #openstack-neutron | 19:26 | |
| *** sdague has quit IRC | 19:26 | |
| *** wolverineav has joined #openstack-neutron | 19:27 | |
| *** sdague has joined #openstack-neutron | 19:27 | |
| *** abhiraut has quit IRC | 19:27 | |
| *** jlinkes has quit IRC | 19:30 | |
| *** abregman has joined #openstack-neutron | 19:32 | |
| *** trown|lunch is now known as trown | 19:33 | |
| *** abhiraut has joined #openstack-neutron | 19:35 | |
| *** korzen has joined #openstack-neutron | 19:39 | |
| korzen | hi ihrachys | 19:40 |
| ihrachys | korzen: hi | 19:40 |
| korzen | ihrachys, I have addressed Kevin comments in https://review.openstack.org/#/c/407868/ port binding ovo adoption | 19:40 |
| korzen | but the patch is not passing the gate | 19:41 |
| dasm | korzen: functional tests are broken. fix is here: https://review.openstack.org/#/c/424906/ | 19:41 |
| korzen | it seems like gate failures, not patch's fault | 19:41 |
| dasm | and linuxbridge one is affected by oom-killer | 19:42 |
| dasm | soo.. now you're lucky :) | 19:42 |
| *** mitchjameson has joined #openstack-neutron | 19:42 | |
| *** radhikam has quit IRC | 19:43 | |
| korzen | dasm, ok, I thought that it was already fixed | 19:45 |
| dasm | korzen: still in queue | 19:45 |
| ihrachys | korzen: yeah. there is also some fullstack failure that seems legit | 19:45 |
| ihrachys | korzen: but not for the patch | 19:45 |
| *** jprovazn has quit IRC | 19:46 | |
| korzen | ihrachys, fullstack seems like high workload in gate or smth | 19:47 |
| ihrachys | well there is probably some legit issue there still, I see non existent attribute accessed by eventle | 19:47 |
| ihrachys | *eventlet | 19:47 |
| *** manheim has joined #openstack-neutron | 19:49 | |
| *** Sukhdev has joined #openstack-neutron | 19:52 | |
| *** radhikam has joined #openstack-neutron | 19:52 | |
| *** eezhova has joined #openstack-neutron | 19:53 | |
| *** Jeffrey4l has joined #openstack-neutron | 19:58 | |
| *** bobmel has quit IRC | 19:59 | |
| *** esmiurium has quit IRC | 19:59 | |
| *** bobmel has joined #openstack-neutron | 19:59 | |
| haleyb | prometheanfire: none of that code was ever merged | 20:01 |
| *** nplanel has joined #openstack-neutron | 20:02 | |
| prometheanfire | haleyb: oh boo | 20:02 |
| openstackgerrit | Manjeet Singh Bhatia proposed openstack/neutron: OVO External Networks https://review.openstack.org/353088 | 20:03 |
| manjeets | electrocucaracha, ihrachys ^^ tried to add some ordering not sure if that would be enough though but reviews will help | 20:04 |
| *** jlibosva has quit IRC | 20:04 | |
| *** Jeffrey4l has quit IRC | 20:04 | |
| *** Sukhdev has quit IRC | 20:05 | |
| *** sdague has quit IRC | 20:06 | |
| *** esmiurium has joined #openstack-neutron | 20:07 | |
| *** morgabra has joined #openstack-neutron | 20:08 | |
| *** morgabra has quit IRC | 20:08 | |
| *** morgabra has joined #openstack-neutron | 20:08 | |
| *** abhiraut has quit IRC | 20:11 | |
| openstackgerrit | Aradhana Singh proposed openstack/neutron: Refactoring config options for ml2 config opts https://review.openstack.org/340228 | 20:11 |
| *** john-davidge has quit IRC | 20:12 | |
| *** jlibosva has joined #openstack-neutron | 20:12 | |
| *** jaosorior has joined #openstack-neutron | 20:12 | |
| *** mvk has joined #openstack-neutron | 20:13 | |
| *** abregman has quit IRC | 20:13 | |
| *** abregman has joined #openstack-neutron | 20:13 | |
| *** abhiraut has joined #openstack-neutron | 20:14 | |
| *** korzen has quit IRC | 20:15 | |
| *** nicolasbock has joined #openstack-neutron | 20:16 | |
| *** beagles-biab is now known as beagles | 20:16 | |
| *** jaosorior has quit IRC | 20:17 | |
| *** sdague has joined #openstack-neutron | 20:18 | |
| openstackgerrit | John Schwarz proposed openstack/neutron: ovs fw migration from iptables https://review.openstack.org/413082 | 20:20 |
| *** TMM has joined #openstack-neutron | 20:21 | |
| *** moshele has joined #openstack-neutron | 20:23 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/neutron: Updated from global requirements https://review.openstack.org/423645 | 20:24 |
| *** jaosorior has joined #openstack-neutron | 20:24 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/neutron-fwaas: Updated from global requirements https://review.openstack.org/425393 | 20:24 |
| *** abhiraut has quit IRC | 20:24 | |
| *** catinthe_ has quit IRC | 20:27 | |
| *** abregman has quit IRC | 20:28 | |
| *** bobmel has quit IRC | 20:31 | |
| *** bobmel has joined #openstack-neutron | 20:34 | |
| *** slaweq has quit IRC | 20:40 | |
| *** bfernando has quit IRC | 20:41 | |
| *** slaweq has joined #openstack-neutron | 20:41 | |
| *** itzikb_ has joined #openstack-neutron | 20:44 | |
| *** Jeffrey4l has joined #openstack-neutron | 20:45 | |
| *** diltram has quit IRC | 20:46 | |
| *** jamielennox|away has quit IRC | 20:46 | |
| *** diltram has joined #openstack-neutron | 20:46 | |
| *** panda|bbl is now known as panda | 20:50 | |
| *** adriant has joined #openstack-neutron | 20:51 | |
| *** jlibosva has quit IRC | 20:52 | |
| *** shaner has quit IRC | 20:53 | |
| *** sdague has quit IRC | 20:53 | |
| *** sdague has joined #openstack-neutron | 20:54 | |
| *** panda is now known as panda|zZ | 20:54 | |
| openstackgerrit | Ihar Hrachyshka proposed openstack/neutron: Use unique subnetpools in SubnetPoolPrefixDbObjectTestCase https://review.openstack.org/425409 | 20:55 |
| *** tbachman has quit IRC | 20:55 | |
| ihrachys | kevinbenton: ^ please, just saw a patch hitting it in gate | 20:55 |
| *** jamielennox|away has joined #openstack-neutron | 20:56 | |
| *** jamielennox|away is now known as jamielennox | 20:56 | |
| dasm | ihrachys: probably you'll hit this bug: https://review.openstack.org/#/c/424906/ | 20:59 |
| dasm | ihrachys: it's still in queue | 20:59 |
| ihrachys | still there? huh | 20:59 |
| dasm | njohnston raised a question on ML about gate being stuck. it's progressing, but very slowly | 21:00 |
| dasm | this patch didn't even started to be tested against gate | 21:00 |
| njohnston | that was neiljerram :-) | 21:00 |
| dasm | njohnston: mea culpa | 21:00 |
| * haleyb will just take a break until that merges | 21:01 | |
| *** shaner has joined #openstack-neutron | 21:04 | |
| *** abhiraut has joined #openstack-neutron | 21:08 | |
| *** janzian has quit IRC | 21:11 | |
| *** Leo_ has quit IRC | 21:13 | |
| *** john-davidge has joined #openstack-neutron | 21:13 | |
| *** princenana has joined #openstack-neutron | 21:14 | |
| *** tbachman has joined #openstack-neutron | 21:14 | |
| *** john-davidge has quit IRC | 21:18 | |
| *** dsneddon is now known as dsneddon_afk | 21:21 | |
| *** Leo_ has joined #openstack-neutron | 21:22 | |
| *** Leom has joined #openstack-neutron | 21:23 | |
| *** Leo_ has quit IRC | 21:23 | |
| *** dsneddon has joined #openstack-neutron | 21:23 | |
| *** janzian has joined #openstack-neutron | 21:23 | |
| *** yamamoto has joined #openstack-neutron | 21:27 | |
| *** mriedem has joined #openstack-neutron | 21:29 | |
| *** tonytan4ever has quit IRC | 21:29 | |
| *** iranzo has joined #openstack-neutron | 21:29 | |
| *** liangy has quit IRC | 21:29 | |
| *** tonytan4ever has joined #openstack-neutron | 21:29 | |
| *** tonytan4ever has quit IRC | 21:30 | |
| *** tonytan4ever has joined #openstack-neutron | 21:31 | |
| *** jaosorior has quit IRC | 21:31 | |
| *** eezhova has quit IRC | 21:33 | |
| *** tonytan4ever has quit IRC | 21:33 | |
| *** tonytan4ever has joined #openstack-neutron | 21:34 | |
| *** Jeffrey4l_ has joined #openstack-neutron | 21:34 | |
| *** Jeffrey4l has quit IRC | 21:35 | |
| *** yamamoto has quit IRC | 21:36 | |
| *** jamesdenton has quit IRC | 21:38 | |
| *** jidar has quit IRC | 21:39 | |
| *** ociuhandu has joined #openstack-neutron | 21:39 | |
| *** jidar has joined #openstack-neutron | 21:40 | |
| *** tbachman has quit IRC | 21:45 | |
| *** belharar_ has joined #openstack-neutron | 21:45 | |
| *** tonytan4ever has quit IRC | 21:46 | |
| *** tonytan4ever has joined #openstack-neutron | 21:46 | |
| *** abhiraut has quit IRC | 21:47 | |
| *** manheim has quit IRC | 21:49 | |
| *** gcheresh_ has quit IRC | 21:49 | |
| *** boden has joined #openstack-neutron | 21:50 | |
| *** abhiraut has joined #openstack-neutron | 21:50 | |
| *** manheim has joined #openstack-neutron | 21:50 | |
| *** tonytan4ever has quit IRC | 21:56 | |
| *** tonytan4ever has joined #openstack-neutron | 21:56 | |
| openstackgerrit | boden proposed openstack/neutron: use neutron_lib's portbindings api-def https://review.openstack.org/422210 | 21:57 |
| *** cleong has quit IRC | 21:57 | |
| *** teclator has quit IRC | 21:59 | |
| *** teclator has joined #openstack-neutron | 21:59 | |
| *** mickeys has quit IRC | 21:59 | |
| electrocucaracha | ihrachys: ping | 22:01 |
| ihrachys | electrocucaracha: howdy | 22:01 |
| electrocucaracha | ihrachys: I was reviewing the patch 425409 | 22:01 |
| *** sdague has quit IRC | 22:02 | |
| electrocucaracha | ihrachys: definitely, make sense the change but do you think that the same approach can be applicable on other _create_test_ * methods https://github.com/openstack/neutron/blob/master/neutron/tests/unit/objects/test_base.py#L1265-L1376 | 22:02 |
| electrocucaracha | ihrachys: most of them hold the object in a property which doesn't make sense | 22:03 |
| ihrachys | electrocucaracha: I think it depends on how those other methods are used. it may sometimes make sense to use the same ID | 22:03 |
| electrocucaracha | ihrachys: yeah, it should be review case by case | 22:03 |
| *** leitan has quit IRC | 22:03 | |
| ihrachys | electrocucaracha: I would review patches that get rid of those properties | 22:04 |
| ihrachys | I just tackled that one because I saw the failure in real life | 22:04 |
| *** crose has quit IRC | 22:04 | |
| *** leitan has joined #openstack-neutron | 22:04 | |
| electrocucaracha | ihrachys: gotcha, I'm going to do it in a separate patch, it's more like refactoring stuff but I'd prefer to have it as a standard | 22:05 |
| ihrachys | electrocucaracha: makes sense, add me to reviewers :) | 22:06 |
| electrocucaracha | ihrachys: sure | 22:06 |
| *** abhiraut has quit IRC | 22:08 | |
| *** leitan has quit IRC | 22:08 | |
| *** tonytan4ever has quit IRC | 22:11 | |
| *** gvrangan has quit IRC | 22:12 | |
| *** moshele has quit IRC | 22:16 | |
| *** ihrachys has quit IRC | 22:20 | |
| boden | anybody know if there’s a “destroy” hook for plugins (e.g. do something on plugin exit)? | 22:21 |
| *** yamamoto has joined #openstack-neutron | 22:22 | |
| openstackgerrit | Merged openstack/neutron-fwaas: Updated from global requirements https://review.openstack.org/425393 | 22:22 |
| *** gouthamr has quit IRC | 22:24 | |
| *** gvrangan has joined #openstack-neutron | 22:24 | |
| *** thorst_ has quit IRC | 22:28 | |
| *** baoli has quit IRC | 22:31 | |
| *** boden has quit IRC | 22:31 | |
| *** mickeys has joined #openstack-neutron | 22:31 | |
| *** mmalik4 has quit IRC | 22:32 | |
| *** jheroux has quit IRC | 22:33 | |
| *** annegentle has quit IRC | 22:34 | |
| *** nplanel has quit IRC | 22:39 | |
| *** Sukhdev has joined #openstack-neutron | 22:41 | |
| *** tbachman has joined #openstack-neutron | 22:43 | |
| raj_ | Toss-up question - the bw limiting in Newton which one can apply to a port - is that expected to apply in both directions, or just outbound? | 22:43 |
| *** sterdnotshaken has joined #openstack-neutron | 22:44 | |
| *** abhiraut has joined #openstack-neutron | 22:44 | |
| *** boden has joined #openstack-neutron | 22:47 | |
| sterdnotshaken | Can someone clarify neutron availability zones for me please? If I assign a network and router to a new availability zone (using routers in ha mode), shouldn't it create said routers ONLY on network nodes that are associated with the availability zone? | 22:47 |
| amuller | sterdnotshaken: I know this one =D Looking it up... | 22:49 |
| amuller | sterdnotshaken: here is the code https://github.com/openstack/neutron/blob/master/neutron/scheduler/l3_agent_scheduler.py#L421 | 22:49 |
| amuller | sterdnotshaken: you need to configure the appropriate L3 scheduler that supports L3 HA and AZs | 22:50 |
| amuller | the default l3 scheduler (least routers) is oblivious to AZs | 22:50 |
| *** TMM has quit IRC | 22:50 | |
| *** dave-mcc_ has joined #openstack-neutron | 22:52 | |
| *** dave-mccowan has quit IRC | 22:54 | |
| sterdnotshaken | amuller: Thanks for your answer! Do you know if availibility zones and this L3 scheduler work when using DVR mode? | 22:55 |
| *** belharar_ has quit IRC | 22:55 | |
| sterdnotshaken | It would seem they would be incompatible? | 22:55 |
| amuller | sterdnotshaken: I think it would, for the centralized / SNAT portion of the router | 22:55 |
| amuller | of course the distributed portion is on every compute so it's not relevant for scheduling | 22:56 |
| *** nplanel has joined #openstack-neutron | 22:56 | |
| sterdnotshaken | I see… Is there a way, when I have configured our cluster as DVR to create a router that isn't DVR? | 22:56 |
| *** vhosakot has joined #openstack-neutron | 22:57 | |
| *** abhiraut has quit IRC | 22:57 | |
| sterdnotshaken | neutron router-create cust7 --distributed False | 22:57 |
| sterdnotshaken | I forgot about that... | 22:57 |
| sterdnotshaken | amuller: Really appreciate your help! | 22:57 |
| amuller | :) | 22:57 |
| *** amuller has quit IRC | 22:59 | |
| *** catintheroof has joined #openstack-neutron | 23:00 | |
| *** manheim has quit IRC | 23:03 | |
| *** Swami has quit IRC | 23:04 | |
| *** abhiraut has joined #openstack-neutron | 23:05 | |
| *** manheim has joined #openstack-neutron | 23:05 | |
| *** dave-mccowan has joined #openstack-neutron | 23:05 | |
| *** dave-mcc_ has quit IRC | 23:05 | |
| *** claudiub has quit IRC | 23:05 | |
| *** tbachman has quit IRC | 23:05 | |
| *** abhiraut has quit IRC | 23:06 | |
| *** jperry has quit IRC | 23:08 | |
| *** esmiurium has quit IRC | 23:09 | |
| *** princenana has quit IRC | 23:11 | |
| *** marst has quit IRC | 23:13 | |
| *** john-davidge has joined #openstack-neutron | 23:14 | |
| *** wolverineav has quit IRC | 23:15 | |
| *** abhiraut has joined #openstack-neutron | 23:16 | |
| *** esmiurium has joined #openstack-neutron | 23:18 | |
| *** john-davidge has quit IRC | 23:18 | |
| *** jperry has joined #openstack-neutron | 23:18 | |
| *** tbachman has joined #openstack-neutron | 23:20 | |
| *** abhiraut has quit IRC | 23:26 | |
| *** wolverineav has joined #openstack-neutron | 23:27 | |
| *** abhiraut has joined #openstack-neutron | 23:28 | |
| *** vhosakot has quit IRC | 23:29 | |
| *** kbringard has quit IRC | 23:30 | |
| *** tbachman has quit IRC | 23:31 | |
| *** zkassab has quit IRC | 23:32 | |
| *** wolverineav has quit IRC | 23:32 | |
| *** iranzo has quit IRC | 23:33 | |
| *** abhiraut has quit IRC | 23:34 | |
| *** mlavalle has quit IRC | 23:35 | |
| *** hichihara has joined #openstack-neutron | 23:35 | |
| *** sterdnotshaken has quit IRC | 23:36 | |
| *** nagarjung has joined #openstack-neutron | 23:36 | |
| *** wolverineav has joined #openstack-neutron | 23:38 | |
| *** nplanel has quit IRC | 23:41 | |
| *** wolverineav has quit IRC | 23:42 | |
| *** Leom has quit IRC | 23:42 | |
| *** Swami has joined #openstack-neutron | 23:46 | |
| *** baoli has joined #openstack-neutron | 23:47 | |
| *** tbachman has joined #openstack-neutron | 23:47 | |
| *** bobmel has quit IRC | 23:48 | |
| *** nagarjung has quit IRC | 23:48 | |
| *** Swami_ has joined #openstack-neutron | 23:48 | |
| *** Swami has quit IRC | 23:51 | |
| *** Swami has joined #openstack-neutron | 23:52 | |
| *** itzikb_ has quit IRC | 23:52 | |
| *** janzian has quit IRC | 23:52 | |
| *** Swami_ has quit IRC | 23:55 | |
| *** wolverineav has joined #openstack-neutron | 23:58 | |
| *** catintheroof has quit IRC | 23:58 | |
| *** manheim has quit IRC | 23:59 | |
| *** catintheroof has joined #openstack-neutron | 23:59 | |
| *** manheim has joined #openstack-neutron | 23:59 | |
| *** Swami has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!