*** azbiswas has quit IRC | 00:07 | |
*** azbiswas has joined #openstack-neutron-ovn | 00:08 | |
*** azbiswas has quit IRC | 00:13 | |
*** chandrav has quit IRC | 00:16 | |
*** salv-orlando has joined #openstack-neutron-ovn | 01:00 | |
*** armax has quit IRC | 01:14 | |
*** salv-orlando has quit IRC | 01:17 | |
*** salv-orlando has joined #openstack-neutron-ovn | 02:00 | |
*** salv-orlando has quit IRC | 02:05 | |
*** yamamoto has joined #openstack-neutron-ovn | 02:22 | |
openstackgerrit | Matthew Kassawara proposed openstack/networking-ovn: [docs] Improve deployment tool integration https://review.openstack.org/258224 | 02:24 |
---|---|---|
*** yamamoto has quit IRC | 02:25 | |
*** azbiswas has joined #openstack-neutron-ovn | 02:51 | |
*** fzdarsky_ has joined #openstack-neutron-ovn | 02:52 | |
*** fzdarsky has quit IRC | 02:55 | |
*** yamamoto has joined #openstack-neutron-ovn | 03:12 | |
*** armax has joined #openstack-neutron-ovn | 04:00 | |
*** armax has quit IRC | 04:07 | |
*** armax has joined #openstack-neutron-ovn | 04:20 | |
*** azbiswas_ has joined #openstack-neutron-ovn | 04:32 | |
*** azbiswas has quit IRC | 04:34 | |
*** azbiswas_ has quit IRC | 04:54 | |
*** azbiswas has joined #openstack-neutron-ovn | 05:14 | |
*** chandrav has joined #openstack-neutron-ovn | 05:15 | |
*** salv-orlando has joined #openstack-neutron-ovn | 05:31 | |
*** salv-orlando has quit IRC | 05:48 | |
*** salv-orlando has joined #openstack-neutron-ovn | 05:50 | |
*** salv-orl_ has joined #openstack-neutron-ovn | 05:54 | |
*** salv-orlando has quit IRC | 05:56 | |
*** salv-orlando has joined #openstack-neutron-ovn | 05:58 | |
*** salv-orl_ has quit IRC | 05:58 | |
*** salv-orlando has quit IRC | 06:07 | |
*** salv-orlando has joined #openstack-neutron-ovn | 06:08 | |
*** salv-orl_ has joined #openstack-neutron-ovn | 06:14 | |
*** salv-orlando has quit IRC | 06:14 | |
*** salv-orlando has joined #openstack-neutron-ovn | 06:41 | |
*** salv-orl_ has quit IRC | 06:43 | |
*** armax has quit IRC | 06:57 | |
*** salv-orlando has quit IRC | 07:04 | |
*** azbiswas has quit IRC | 07:20 | |
*** fzdarsky_ has quit IRC | 07:34 | |
*** salv-orlando has joined #openstack-neutron-ovn | 07:36 | |
*** salv-orl_ has joined #openstack-neutron-ovn | 07:47 | |
*** salv-orlando has quit IRC | 07:47 | |
*** salv-orlando has joined #openstack-neutron-ovn | 07:54 | |
*** salv-orl_ has quit IRC | 07:55 | |
*** salv-orl_ has joined #openstack-neutron-ovn | 07:57 | |
*** salv-orlando has quit IRC | 07:57 | |
*** yamamoto has quit IRC | 08:00 | |
*** yamamoto has joined #openstack-neutron-ovn | 08:03 | |
*** salv-orl_ has quit IRC | 08:05 | |
*** yamamoto has quit IRC | 08:06 | |
*** salv-orl_ has joined #openstack-neutron-ovn | 08:08 | |
gsagie | arosen: there is some rally problem | 08:13 |
gsagie | thats why the rally patches are failing | 08:13 |
gsagie | <kun_huang> https://www.irccloud.com/pastebin/NmwmxNMW/ | 08:14 |
gsagie | <kun_huang> boris said https://review.openstack.org/#/c/258232/ this one | 08:14 |
*** salv-orl_ has quit IRC | 08:14 | |
*** yamamoto has joined #openstack-neutron-ovn | 08:18 | |
*** yamamoto has quit IRC | 08:18 | |
*** azbiswas has joined #openstack-neutron-ovn | 08:21 | |
*** azbiswas has quit IRC | 08:25 | |
*** salv-orlando has joined #openstack-neutron-ovn | 08:27 | |
*** salv-orlando has quit IRC | 08:38 | |
openstackgerrit | Gal Sagie proposed openstack/networking-ovn: Change Rally tasks file name to fit the project name https://review.openstack.org/258348 | 08:38 |
*** salv-orlando has joined #openstack-neutron-ovn | 08:45 | |
*** salv-orlando has quit IRC | 08:46 | |
*** salv-orlando has joined #openstack-neutron-ovn | 08:47 | |
*** chandrav has quit IRC | 08:47 | |
openstackgerrit | garyk proposed openstack/networking-ovn: Undo neutron network if lswitch was not created https://review.openstack.org/256021 | 08:51 |
*** salv-orlando has quit IRC | 08:57 | |
*** salv-orlando has joined #openstack-neutron-ovn | 08:59 | |
*** salv-orlando has quit IRC | 09:08 | |
*** salv-orlando has joined #openstack-neutron-ovn | 09:26 | |
arosen | gsagie: gotcha, my first patch that merged into rally to the tests a little further. | 09:40 |
arosen | i think the problem is actually in project-config | 09:40 |
arosen | it looks like it isn't passing the right path to the yaml file (you can see in the logs it's missing the file name) | 09:41 |
arosen | http://logs.openstack.org/67/257067/4/check/gate-networking-ovn-dsvm-rally-nv/632e37f/console.html#_2015-12-16_07_50_52_242 | 09:41 |
arosen | this causes the error here: | 09:41 |
arosen | http://logs.openstack.org/67/257067/4/check/gate-networking-ovn-dsvm-rally-nv/632e37f/console.html#_2015-12-16_07_50_53_805 | 09:42 |
gsagie | arosen: https://review.openstack.org/#/c/258347/ | 09:42 |
arosen | cool | 09:42 |
arosen | :) | 09:42 |
arosen | gsagie: this was biting us before that in the ovn gate : https://review.openstack.org/#/c/258232/ | 09:43 |
arosen | alright i'm calling it though. later. | 09:43 |
gsagie | good night! | 09:44 |
gsagie | hopefully they approve it soon and i will be able to test it | 09:45 |
*** yamamoto has joined #openstack-neutron-ovn | 09:47 | |
*** salv-orlando has quit IRC | 09:54 | |
*** flaviof_ has joined #openstack-neutron-ovn | 09:58 | |
*** flaviof has quit IRC | 10:00 | |
*** azbiswas has joined #openstack-neutron-ovn | 10:01 | |
*** azbiswas has quit IRC | 10:07 | |
*** fzdarsky_ has joined #openstack-neutron-ovn | 10:12 | |
*** salv-orlando has joined #openstack-neutron-ovn | 10:41 | |
*** yamamoto has quit IRC | 11:23 | |
*** gongysh_ has joined #openstack-neutron-ovn | 11:29 | |
*** gongysh_ has quit IRC | 11:32 | |
*** gongysh_ has joined #openstack-neutron-ovn | 11:37 | |
*** roeyc has joined #openstack-neutron-ovn | 11:43 | |
*** salv-orlando has quit IRC | 11:47 | |
*** salv-orlando has joined #openstack-neutron-ovn | 11:47 | |
*** yamamoto has joined #openstack-neutron-ovn | 12:00 | |
*** yamamoto_ has joined #openstack-neutron-ovn | 12:02 | |
*** salv-orlando has quit IRC | 12:03 | |
*** yamamoto has quit IRC | 12:05 | |
*** salv-orlando has joined #openstack-neutron-ovn | 12:09 | |
*** salv-orlando has quit IRC | 12:11 | |
openstackgerrit | Merged openstack/networking-ovn: Undo neutron network if lswitch was not created https://review.openstack.org/256021 | 12:27 |
*** rtheis has joined #openstack-neutron-ovn | 12:29 | |
*** yamamoto_ has quit IRC | 12:36 | |
*** gongysh_ has quit IRC | 12:51 | |
*** yamamoto has joined #openstack-neutron-ovn | 12:53 | |
*** yamamoto has quit IRC | 12:54 | |
*** flaviof_ is now known as flaviof | 13:35 | |
*** azbiswas has joined #openstack-neutron-ovn | 13:37 | |
russellb | Sam-I-Am: yes, openvswitch-ovn includes both ovn-controller and ovn-northd | 13:38 |
*** dslev_ has quit IRC | 13:38 | |
russellb | Sam-I-Am: that could change ... someone just posted patches for the debian packages, and he did it differently. I'd like to match, so I'll probably change the rpms to match wherever that ends up | 13:38 |
Sam-I-Am | russellb: mornings | 13:38 |
russellb | morning :) | 13:39 |
Sam-I-Am | where are these deb packages? | 13:39 |
russellb | proposed patches | 13:39 |
russellb | this is the start of the patch series -- http://openvswitch.org/pipermail/dev/2015-December/063432.html | 13:40 |
Sam-I-Am | ah | 13:41 |
*** azbiswas has quit IRC | 13:41 | |
russellb | source of the packaing for rpm and debs is in the ovs git tree | 13:41 |
Sam-I-Am | usually rh and debian do things differently | 13:41 |
Sam-I-Am | thats what makes maintaining the install guide fun | 13:42 |
Sam-I-Am | the deployment tool integration docs may also need to include source | 13:43 |
Sam-I-Am | or if ovs decided to roll their own packages | 13:43 |
russellb | ok | 13:43 |
Sam-I-Am | the networking guide, on the other hand, assumes people know how to install the software bits... it just covers configuration | 13:47 |
Sam-I-Am | makes it a lot easier when its not distro-specific | 13:47 |
*** dslev has joined #openstack-neutron-ovn | 13:49 | |
*** yamamoto has joined #openstack-neutron-ovn | 13:55 | |
russellb | Sam-I-Am: yeah, that makes sense. This particular doc is just part of bootstrapping OVN. | 13:55 |
russellb | I *know* nobody knows how to install the bits yet | 13:55 |
russellb | teaching the devs of deployment tools how to integrate OVN seemed like a good step to get there | 13:55 |
russellb | because ideally OVN gets integrated into TripleO, Fuel, etc in the coming year | 13:56 |
russellb | i'm not sure how many people actually install from source in practice ... | 13:56 |
russellb | i expect very few ... | 13:56 |
Sam-I-Am | openstack-ansible installs openstack from source, but currently uses distro packages for system utils | 13:58 |
Sam-I-Am | however, if distros have trouble packaging ovs-ovn or dont update it enough, we'll probably see some try to build it from source | 13:59 |
*** yamamoto has quit IRC | 14:02 | |
openstackgerrit | Merged openstack/networking-ovn: Fix remote_group handling when remote_group is not self https://review.openstack.org/257067 | 14:05 |
russellb | Sam-I-Am: do you think a "how to install from source" guide would be better than this deployment integration guide? | 14:12 |
russellb | err, an install guide that includes both packages and source installs | 14:12 |
russellb | i suppose that serves the same purpose as what i started | 14:12 |
russellb | maybe we can morph it that direction | 14:12 |
Sam-I-Am | russellb: yeah. rdo, deb, obs packaging + source option (for ovs + ovn at least, not entire openstack) | 14:13 |
* russellb nods | 14:13 | |
russellb | so bad news | 14:13 |
russellb | i'm out of coffee | 14:13 |
russellb | good news, i had some backup tea | 14:13 |
Sam-I-Am | sounds like my problem yesterday | 14:13 |
russellb | i just have a keurig ... out of k-cups | 14:14 |
russellb | but box ordered! | 14:14 |
Sam-I-Am | my guess is the deployment tool integration guide will morph into the upstream install guide when/if we include ovn... in which case doing more work now makes sense | 14:14 |
Sam-I-Am | this sounds like planning fail | 14:14 |
russellb | truth | 14:14 |
Sam-I-Am | i like writing reusable docs | 14:15 |
Sam-I-Am | minus the packaging bits, the deployment guide smells like the start of the networking guide scenario | 14:16 |
gsagie | "reusable docs" .. nice :) | 14:16 |
Sam-I-Am | gsagie: efficiency | 14:16 |
Sam-I-Am | russellb: btw, hope i didnt butcher your deployment guide too much | 14:17 |
russellb | Sam-I-Am: heh, i don't care | 14:17 |
russellb | i just want the info to exist | 14:17 |
russellb | the diff looks like a rewrite :) | 14:17 |
Sam-I-Am | yeah, i ended up making it more step-oriented | 14:18 |
Sam-I-Am | docs diffs always suck | 14:18 |
russellb | cool, i'm not in tune with doc conventions | 14:18 |
russellb | this all seems normal though | 14:18 |
Sam-I-Am | they always look worse than they are | 14:18 |
russellb | i've done professional writing, and then a copy editor just reworded everything to make me not sound stupid | 14:18 |
Sam-I-Am | heh | 14:18 |
russellb | should have put the copy editor's name on the cover of my book | 14:19 |
russellb | seriously, she was amazing | 14:19 |
russellb | anyway. | 14:19 |
Sam-I-Am | i figure this stuff is probably aiming for central docs at some point, so might as well get it out of the way sooner than later | 14:19 |
russellb | you're doing great stuff, i appreciate it | 14:19 |
russellb | yep | 14:19 |
Sam-I-Am | i just want to make sure i didnt make the technical bits inaccurate | 14:19 |
Sam-I-Am | there are a couple of questions for you in the patch | 14:19 |
russellb | i'm in the midst of answering! | 14:20 |
Sam-I-Am | excellent | 14:20 |
Sam-I-Am | without caffeine though? | 14:20 |
russellb | i have tea | 14:20 |
Sam-I-Am | tea = sub-par replacement for coffee | 14:20 |
russellb | caffientated | 14:20 |
russellb | heh | 14:20 |
Sam-I-Am | i went about 3 hours before caffeine yesterday... i had nightmares last night about it | 14:21 |
russellb | http://www.tazo.com/Product/Detail/12 | 14:21 |
russellb | note the caffeine level | 14:21 |
russellb | 5/5 | 14:21 |
Sam-I-Am | they need "nerd-level caffiene" .. it goes to 11 | 14:21 |
russellb | :) | 14:22 |
*** roeyc has quit IRC | 14:22 | |
*** roeyc has joined #openstack-neutron-ovn | 14:23 | |
*** yamamoto has joined #openstack-neutron-ovn | 14:29 | |
*** yamamoto has quit IRC | 14:29 | |
Sam-I-Am | russellb: is 1400 the max mtu for geneve? | 14:29 |
Sam-I-Am | with vxlan we can get away with 1454 iirc, 1450 for gre | 14:30 |
russellb | 1400 is made up | 14:30 |
russellb | devstack == pile of hacks | 14:30 |
russellb | but let me see here ... | 14:30 |
Sam-I-Am | devhack | 14:30 |
*** yamamoto has joined #openstack-neutron-ovn | 14:30 | |
russellb | Sam-I-Am: geneve is a little more complicated, because it's variable | 14:31 |
Sam-I-Am | yeah, which i found strange | 14:31 |
Sam-I-Am | not good for hardware implementations | 14:31 |
gsagie | Sam-I-Am : its called TLV, its pretty common for networking hardware | 14:32 |
russellb | See "Geneve: What is it and why is OVN using it? (Jesse Gross, VMware)" on http://openvswitch.org/support/ovscon2015/ | 14:32 |
Sam-I-Am | yeah, except its not well-defined tlv | 14:32 |
russellb | Jesse did a very nice job explaining "why geneve" and also the status of hardware implementations | 14:32 |
russellb | the TLV format is well defined | 14:32 |
russellb | the meanings are not | 14:33 |
Sam-I-Am | i sort of got the impression that geneve was a bit 'open' | 14:33 |
russellb | check out the talk, worth while i think :) | 14:33 |
Sam-I-Am | will do | 14:34 |
Sam-I-Am | gre -> vxlan -> geneve | 14:34 |
gsagie | Sam-I-Am: yes, there is no standard definitions for the types, if thats what you mean (at least thats what i know) | 14:34 |
Sam-I-Am | geneve -> ??? | 14:34 |
Sam-I-Am | gsagie: thats what i meant | 14:34 |
gsagie | mpls | 14:34 |
gsagie | back to basics | 14:35 |
Sam-I-Am | lol yeah. lets use that. | 14:35 |
russellb | looks like header is 54 bytes (outer ethernet, outer ipv4, outer udp, then geneve) | 14:35 |
russellb | *at least* 54 | 14:35 |
russellb | then variable options | 14:35 |
russellb | and OVN uses some | 14:35 |
russellb | well, uses 1 | 14:35 |
russellb | so another 8 bytes i think | 14:36 |
russellb | so 62? | 14:37 |
russellb | i'd probably want to verify this in wireshark or something heh | 14:37 |
Sam-I-Am | looks that way | 14:37 |
*** regXboi has joined #openstack-neutron-ovn | 14:38 | |
openstackgerrit | garyk proposed openstack/networking-ovn: Ensure that portbindings are tested https://review.openstack.org/257266 | 14:38 |
Sam-I-Am | russellb: good idea | 14:39 |
Sam-I-Am | we need a good safe value where people can start from | 14:39 |
Sam-I-Am | also, does ovn end-to-end jumbo frames? | 14:40 |
russellb | how would that work | 14:41 |
russellb | not sure what that means | 14:41 |
Sam-I-Am | geneve looks very similar to vxlan minus the variable-length options | 14:41 |
russellb | yeah | 14:41 |
Sam-I-Am | russellb: vm has vif with mtu of 9000... and thats carried through to the external network | 14:41 |
russellb | that would just get dropped if the attached network doesn't support it | 14:42 |
russellb | i think | 14:42 |
russellb | no different than whatever happens with ml2+ovs today in this case | 14:43 |
russellb | though the answer might be "it depends" | 14:43 |
Sam-I-Am | true. however, mtu with neutron+nova has always been weird. | 14:43 |
russellb | i wonder if it works if it goes through the router namespace on the hypervisor | 14:43 |
russellb | kernel stack might do the fragmentation in that case | 14:44 |
* russellb making stuff up | 14:44 | |
Sam-I-Am | isnt that how this works? | 14:44 |
Sam-I-Am | russellb: the main problem i remember is making sure all interfaces in the path have the desired mtu | 14:45 |
openstackgerrit | garyk proposed openstack/networking-ovn: Remove deprecation warnings from unit tests https://review.openstack.org/258550 | 14:52 |
Sam-I-Am | its interesting how mtu math doesnt exactly work out | 14:54 |
Sam-I-Am | caffeeeeiiinnnnee | 15:07 |
Sam-I-Am | i can start thinking now | 15:07 |
*** roeyc has quit IRC | 15:57 | |
*** roeyc has joined #openstack-neutron-ovn | 15:58 | |
russellb | i proposed making our tempest job voting again while it's passing -- https://review.openstack.org/#/c/258595/ | 15:59 |
*** roeyc has quit IRC | 15:59 | |
*** roeyc has joined #openstack-neutron-ovn | 16:01 | |
*** roeyc has quit IRC | 16:02 | |
Sam-I-Am | russellb: at the end of my patch i added a section for verification... something a deployment tool or human can run to make sure things are working. are there any useful things to do here? | 16:07 |
Sam-I-Am | sort of like neutron agent-list, etc. | 16:07 |
Sam-I-Am | poking at dbs | 16:07 |
russellb | yep there's some stuff | 16:08 |
russellb | # ovn-sbctl show | 16:08 |
russellb | would list all hypervisors (ovn-controller instances) that have registered themselves | 16:08 |
*** armax has joined #openstack-neutron-ovn | 16:09 | |
Sam-I-Am | russellb: ok, i'll throw that in for now and add sample output when i get a deployment going | 16:12 |
russellb | cool | 16:13 |
Sam-I-Am | i'll get this patch up... then start working on diagrams | 16:13 |
-openstackstatus- NOTICE: Gerrit will be offline for a software upgrade from 17:00 to 21:00 UTC. See: http://lists.openstack.org/pipermail/openstack-dev/2015-December/081037.html | 16:20 | |
*** azbiswas has joined #openstack-neutron-ovn | 16:29 | |
openstackgerrit | Matthew Kassawara proposed openstack/networking-ovn: [docs] Improve deployment tool integration https://review.openstack.org/258224 | 16:29 |
*** gsagie_ has joined #openstack-neutron-ovn | 16:36 | |
russellb | Sam-I-Am: ascii diagrams only | 16:40 |
Sam-I-Am | welp, i'm out | 16:41 |
* Sam-I-Am drops mic | 16:41 | |
*** azbiswas_ has joined #openstack-neutron-ovn | 16:46 | |
*** azbiswas has quit IRC | 16:48 | |
*** azbiswas has joined #openstack-neutron-ovn | 16:48 | |
*** azbiswas_ has quit IRC | 16:52 | |
*** arosen has quit IRC | 17:01 | |
*** shettyg has joined #openstack-neutron-ovn | 17:02 | |
-openstackstatus- NOTICE: Gerrit is offline for a software upgrade from 17:00 to 21:00 UTC. See: http://lists.openstack.org/pipermail/openstack-dev/2015-December/081037.html | 17:05 | |
*** ChanServ changes topic to "Gerrit is offline for a software upgrade from 17:00 to 21:00 UTC. See: http://lists.openstack.org/pipermail/openstack-dev/2015-December/081037.html" | 17:05 | |
*** gsagie_ has quit IRC | 17:08 | |
*** fzdarsky_ is now known as fzdarsky|afk | 17:14 | |
*** salv-orlando has joined #openstack-neutron-ovn | 17:21 | |
*** arosen has joined #openstack-neutron-ovn | 17:25 | |
*** salv-orlando has quit IRC | 17:34 | |
*** salv-orlando has joined #openstack-neutron-ovn | 17:34 | |
*** azbiswas_ has joined #openstack-neutron-ovn | 17:37 | |
*** azbiswas has quit IRC | 17:40 | |
*** azbiswas has joined #openstack-neutron-ovn | 17:47 | |
*** azbiswas_ has quit IRC | 17:51 | |
*** fzdarsky|afk is now known as fzdarsky | 17:56 | |
*** azbiswas_ has joined #openstack-neutron-ovn | 18:01 | |
*** azbiswas has quit IRC | 18:04 | |
*** salv-orlando has quit IRC | 18:21 | |
*** salv-orlando has joined #openstack-neutron-ovn | 18:23 | |
*** yamamoto has quit IRC | 18:38 | |
*** yamamoto has joined #openstack-neutron-ovn | 18:38 | |
*** chandrav has joined #openstack-neutron-ovn | 18:41 | |
*** chandrav has quit IRC | 18:46 | |
*** yamamoto has quit IRC | 18:48 | |
*** chandrav has joined #openstack-neutron-ovn | 19:00 | |
*** yamamoto has joined #openstack-neutron-ovn | 19:18 | |
*** yamamoto has quit IRC | 19:19 | |
*** salv-orlando has quit IRC | 19:27 | |
*** dslev has quit IRC | 19:58 | |
*** yamamoto has joined #openstack-neutron-ovn | 20:19 | |
*** yamamoto has quit IRC | 20:25 | |
*** ChanServ changes topic to "http://docs.openstack.org/developer/networking-ovn/ -=- OVN meeting Thursdays 10:15am Pacific / 1:15pm Eastern #openvswitch" | 21:01 | |
-openstackstatus- NOTICE: Gerrit has been upgraded to 2.11. Please report any issues in #openstack-infra as soon as possible. | 21:01 | |
*** regXboi has quit IRC | 21:09 | |
*** rtheis has quit IRC | 22:18 | |
*** shettyg has quit IRC | 22:59 | |
*** shettyg has joined #openstack-neutron-ovn | 22:59 | |
*** shettyg has quit IRC | 23:03 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!