| *** reedip_away is now known as reedip | 00:15 | |
| *** salv-orlando has quit IRC | 00:21 | |
| *** dingboopt has joined #openstack-kuryr | 01:03 | |
| *** salv-orlando has joined #openstack-kuryr | 01:29 | |
| *** baohua has joined #openstack-kuryr | 01:30 | |
| *** salv-orlando has quit IRC | 01:31 | |
| baohua | dear kuryr experts, please help review a patch and push it forward, as it's been a long while, and will block further work. thanks a lot :) https://review.openstack.org/#/c/265105/ | 01:38 |
|---|---|---|
| *** kenji has joined #openstack-kuryr | 03:25 | |
| *** kenji is now known as Guest93750 | 03:26 | |
| *** Guest93750 has quit IRC | 03:33 | |
| *** salv-orlando has joined #openstack-kuryr | 04:44 | |
| *** salv-orlando has quit IRC | 04:49 | |
| *** irenab has quit IRC | 05:51 | |
| *** salv-orlando has joined #openstack-kuryr | 06:30 | |
| *** irenab has joined #openstack-kuryr | 06:39 | |
| *** ajo has quit IRC | 07:03 | |
| *** ajo has joined #openstack-kuryr | 07:03 | |
| *** shashank_hegde has quit IRC | 07:23 | |
| *** shashank_hegde has joined #openstack-kuryr | 07:24 | |
| openstackgerrit | Liping Mao proposed openstack/kuryr: Mofify external_ids in /usr/libexec/kuryr/ovs https://review.openstack.org/278224 | 07:44 |
| openstackgerrit | Liping Mao proposed openstack/kuryr: Modify external_ids in /usr/libexec/kuryr/ovs https://review.openstack.org/278224 | 07:47 |
| *** salv-orlando has quit IRC | 07:47 | |
| *** reedip is now known as reedip_away | 08:13 | |
| *** shashank_hegde has quit IRC | 08:20 | |
| *** salv-orlando has joined #openstack-kuryr | 08:59 | |
| *** salv-orlando has quit IRC | 08:59 | |
| *** salv-orlando has joined #openstack-kuryr | 08:59 | |
| *** ajo has quit IRC | 09:16 | |
| *** ajo has joined #openstack-kuryr | 09:16 | |
| *** baohua has quit IRC | 09:30 | |
| *** salv-orl_ has joined #openstack-kuryr | 10:03 | |
| *** salv-orlando has quit IRC | 10:06 | |
| *** apuimedo has joined #openstack-kuryr | 10:36 | |
| *** openstackgerrit_ has joined #openstack-kuryr | 12:27 | |
| *** salv-orl_ has quit IRC | 13:04 | |
| *** salv-orlando has joined #openstack-kuryr | 13:40 | |
| *** fawadkhaliq has joined #openstack-kuryr | 14:10 | |
| *** fawadk has joined #openstack-kuryr | 14:21 | |
| *** fawadk has quit IRC | 14:24 | |
| *** fawadkhaliq has quit IRC | 14:24 | |
| *** fawadkhaliq has joined #openstack-kuryr | 14:24 | |
| *** msagh__ has joined #openstack-kuryr | 14:26 | |
| irenab | apuimedo: gsagie : Cannot make it to kury meeting today, something urgent came up | 14:26 |
| *** fawadkhaliq has quit IRC | 14:26 | |
| apuimedo | Hope nothing bad irenab ;-) | 14:26 |
| *** fawadkhaliq has joined #openstack-kuryr | 14:26 | |
| *** msagh__ has quit IRC | 14:27 | |
| *** salv-orlando has quit IRC | 14:27 | |
| *** gsagie_ has joined #openstack-kuryr | 14:38 | |
| *** banix has joined #openstack-kuryr | 14:41 | |
| gsagie_ | mestery: ping | 14:45 |
| *** irenab has quit IRC | 14:46 | |
| *** salvorlando has joined #openstack-kuryr | 14:56 | |
| *** fawadkhaliq has quit IRC | 15:10 | |
| *** fawadkhaliq has joined #openstack-kuryr | 15:11 | |
| mestery | gsagie_: pong | 15:28 |
| *** devvesa has joined #openstack-kuryr | 15:43 | |
| *** irenab has joined #openstack-kuryr | 15:55 | |
| *** apuimedo has quit IRC | 15:58 | |
| *** apuimedo has joined #openstack-kuryr | 16:04 | |
| *** gsagie_ has quit IRC | 16:09 | |
| *** devvesa has quit IRC | 16:12 | |
| *** gsagie_ has joined #openstack-kuryr | 16:20 | |
| *** salvorlando has quit IRC | 16:40 | |
| *** gsagie_ has quit IRC | 16:56 | |
| *** salv-orlando has joined #openstack-kuryr | 17:06 | |
| *** salv-orlando has quit IRC | 17:11 | |
| openstackgerrit | Mohammad Banikazemi proposed openstack/kuryr: Make use of one instance of IPDB https://review.openstack.org/280334 | 17:27 |
| *** irenab_ has joined #openstack-kuryr | 17:34 | |
| *** irenab has quit IRC | 17:35 | |
| *** irenab_ is now known as irenab | 17:35 | |
| *** salv-orlando has joined #openstack-kuryr | 17:36 | |
| *** openstackgerrit has quit IRC | 17:45 | |
| *** openstackgerrit has joined #openstack-kuryr | 17:46 | |
| *** mspreitz_ has joined #openstack-kuryr | 18:08 | |
| mspreitz_ | apuimedo, are you there? | 18:08 |
| apuimedo | mspreitz_: affirmative | 18:08 |
| mspreitz_ | I tried that renaming hack, no joy --- due to IPAM difference | 18:08 |
| mspreitz_ | root@dev-mon01-kraken15-host-01:~# docker run --net=mynet_name -d 716669d6a328 a390cdb7b4f865254f5a18305e37f9228ca8be5a5390425a67f3dae032b85e0b Error response from daemon: Cannot start container a390cdb7b4f865254f5a18305e37f9228ca8be5a5390425a67f3dae032b85e0b: failed to create endpoint thirsty_ramanujan on network mynet_name: Plugin Error: NetworkDriver.CreateEndpoint, { "Err": "No subnet exist for the cidrs 10.10.1.2/24 | 18:09 |
| mspreitz_ | here's the story:... | 18:09 |
| mspreitz_ | I installed openstack, including a provider network | 18:09 |
| mspreitz_ | then I used `docker network create` to make Docker network mynet_name, with IP addrs in 10.10... | 18:10 |
| mspreitz_ | then I renamed the new Neutron network to fubar, and renamed the provider net to have the Docker network's Docker UUID | 18:10 |
| mspreitz_ | but it still has its original subnets, not 10.10... | 18:10 |
| mspreitz_ | then I tried `docker run`, and it failed as you see. | 18:11 |
| apuimedo | mspreitz_: if you create the network with the same --subnet --ip-range --gateway as the network you'll replace? | 18:11 |
| mspreitz_ | Um, won't Neutron object? | 18:11 |
| apuimedo | I don't think so, it's fine to have overlapping networks in neutron | 18:12 |
| mspreitz_ | I'm game to try. What would the command line be to `docker network create` with IPAM matching my existing provider network? | 18:12 |
| apuimedo | let me get it for you | 18:12 |
| *** shashank_hegde has joined #openstack-kuryr | 18:13 | |
| apuimedo | docker network create --driver=kuryr --ipam-driver=kuryr --subnet 10.12.0.0/16 --ip-range 10.12.0.0/16 --gateway 10.12.0.1 bcn | 18:13 |
| apuimedo | mspreitz_: ^^ | 18:13 |
| mspreitz_ | let me try... | 18:13 |
| mspreitz_ | big messy error spew. Let me make an etherpad... | 18:16 |
| apuimedo | thanks | 18:18 |
| mspreitz_ | apuimedo, banix: see https://etherpad.openstack.org/p/0fFdv4PPXI | 18:18 |
| mspreitz_ | It ends with a little status, in case that helps | 18:18 |
| apuimedo | wow, I never saw the web error trace | 18:20 |
| apuimedo | only the console one | 18:20 |
| apuimedo | mspreitz_: is the normal binding working for you? | 18:22 |
| apuimedo | without this hacky changing of the network under the wraps? | 18:22 |
| apuimedo | cause it looks to me like you don't have anything in /usr/libexec/kuryr | 18:23 |
| openstackgerrit | Merged openstack/kuryr: Make use of one instance of IPDB https://review.openstack.org/280334 | 18:23 |
| mspreitz_ | oh, let me check | 18:24 |
| mspreitz_ | I've got no /usr/libexec at all | 18:25 |
| mspreitz_ | banix: should I expect your hacky install script to make and populate /usr/libexec? As far as I can tell, it does not. | 18:26 |
| mspreitz_ | apuimedo: do we need to talk about versions now? | 18:26 |
| apuimedo | versions? | 18:27 |
| mspreitz_ | of kuryr | 18:27 |
| mspreitz_ | I got a kuryr distro from banix a couple of weeks ago | 18:27 |
| apuimedo | cp -r usr/libexec/kuryr /usr/libexec/kuryr | 18:28 |
| mspreitz_ | let me see if it is functioning without worrying about /usr/libexec ... | 18:28 |
| apuimedo | from the kuryr repo | 18:28 |
| mspreitz_ | I get a similar problem when I attempt no renaming. See https://etherpad.openstack.org/p/ccd8Q1twx6 | 18:30 |
| apuimedo | mspreitz_: did you place the /usr/libexec or not? | 18:31 |
| apuimedo | because it complains that it does not find the binding script | 18:31 |
| mspreitz_ | not yet. I found two usr/libexec in what I got from banix, will make a guess and use it... | 18:32 |
| apuimedo | cool | 18:32 |
| apuimedo | gotta go take care of my son, but I should be back in 1.5-2h | 18:32 |
| mspreitz_ | cp -rp, right? | 18:32 |
| mspreitz_ | I copied and fixed up permissions, on both worker nodes. Still getting big error spew. See https://etherpad.openstack.org/p/9Is0Kqpm4v | 18:40 |
| mspreitz_ | banix: are you around now? | 18:45 |
| *** irenab has quit IRC | 18:53 | |
| apuimedo | mspreitz_: which neutron plugin do you have configured? Ovn? | 18:57 |
| *** banix has quit IRC | 18:59 | |
| mspreitz_ | ML2 with OVS and VXLAN | 19:01 |
| *** fawadkhaliq has quit IRC | 19:02 | |
| mspreitz_ | apuimedo: ML2 with OVS and VXLAN | 19:03 |
| apuimedo | can you put a print in kuryr/binding.py just after it does the vif_type = neutron_port.get(VIF_TYPE_KEY, FALLBACK_VIF_TYPE) | 19:04 |
| apuimedo | of the vif_type value | 19:04 |
| apuimedo | and of binding_exec_path | 19:04 |
| mspreitz_ | I think so, just a few mins.. | 19:09 |
| mspreitz_ | now, looking for where stdout went... | 19:14 |
| * mspreitz_ slaps head, restarts kuryr | 19:17 | |
| apuimedo | :-) | 19:18 |
| mspreitz_ | Here is the output: | 19:19 |
| mspreitz_ | vif_type: ovs binding_exec_path: /home/openstack/latest/kuryr/usr/libexec/kuryr/ovs | 19:19 |
| *** irenab has joined #openstack-kuryr | 19:19 | |
| mspreitz_ | BTW, /home/openstack/latest/kuryr/usr/libexec is what I copied to /usr/ | 19:19 |
| mspreitz_ | well, not actually! | 19:20 |
| mspreitz_ | There is no /home/openstack/latest/kuryr/usr/libexec ! | 19:21 |
| mspreitz_ | There is no /home/openstack/latest | 19:21 |
| mspreitz_ | apuimedo: ^^ | 19:22 |
| mspreitz_ | apuimedo: /home/mspreitzer/kuryr-installer/latest/kuryr/usr/libexec is what I copied to /usr | 19:22 |
| apuimedo | oh, that is a funny path :P | 19:22 |
| apuimedo | banix installer must be quite unusual :P | 19:22 |
| mspreitz_ | Worse yet, I adapted it so that part is now an Ansible playbook | 19:23 |
| apuimedo | anyway, if you put it in the path it expects it should work | 19:23 |
| apuimedo | mspreitz_: for Ursula? | 19:23 |
| mspreitz_ | not yet, Ursula and I are not getting along | 19:23 |
| apuimedo | :P | 19:23 |
| apuimedo | deployment tools are tough lovers | 19:25 |
| mspreitz_ | Well, the copy eliminated the complaint... now to see if it is actually wired up correctly... | 19:25 |
| fkautz | apuimedo: do you have an easy way of tracing calls made from the libnetwork plugin to neutron? | 19:26 |
| * apuimedo crosses fingers | 19:26 | |
| apuimedo | fkautz: well, if you run kuryr in foreground it prints the requests I think | 19:27 |
| mspreitz_ | hmm, the flailures have left some junk that may be harmfull, let me see about clearing it out... | 19:27 |
| fkautz | apuimedo: thanks, i'll give that a try | 19:27 |
| apuimedo | fkautz: well, it prints the requests to kuryr | 19:28 |
| apuimedo | the requests to neutron maybe only if they fail, actually | 19:28 |
| apuimedo | you'd like to see the underlying http request? | 19:28 |
| apuimedo | mspreitz_: clear it out with neutron cli | 19:28 |
| apuimedo | first | 19:28 |
| apuimedo | then delete with docker network | 19:28 |
| mspreitz_ | The junk to which I am referring is ip routes | 19:28 |
| apuimedo | it's a sureway to get it clean ;-) | 19:28 |
| apuimedo | ah, ok | 19:29 |
| apuimedo | mspreitz_: in MidoNet we don't have that :P | 19:29 |
| mspreitz_ | actually, rather than clean out the old, let me just start with a fresh network | 19:29 |
| *** salv-orlando has quit IRC | 19:30 | |
| apuimedo | okey dokey | 19:30 |
| apuimedo | going for dinner, will check later | 19:30 |
| mspreitz_ | apuimedo: thanks | 19:31 |
| *** banix has joined #openstack-kuryr | 19:36 | |
| mspreitz_ | It worked! | 19:40 |
| fkautz | :D | 19:41 |
| mspreitz_ | cumulative hack list to make kuryr work at all: copy libexec to /usr, to ~openstack | 19:41 |
| mspreitz_ | hack to connect to existing provider network: create new network with same IPAM, switch Neutron names around | 19:41 |
| mspreitz_ | and beware broken networks | 19:42 |
| mspreitz_ | of which my provider network is now one | 19:42 |
| mspreitz_ | To be a little more precise: the places to which I had to copy libexec are: /usr and /home/openstack/latest/kuryr/usr | 19:43 |
| banix | mspreitz_: not the place to discuss this, but the tar file you have, should include the usr/libexec directory | 19:44 |
| mspreitz_ | I found libexec in /home/mspreitzer/kuryr-installer/latest/kuryr/usr/libexec | 19:44 |
| mspreitz_ | banix: that was from your distro tar | 19:44 |
| mspreitz_ | your install scripting did not do any of those copies. I am going to update my Ansible playbook, which I use instead of your install_kuryr.sh | 19:45 |
| apuimedo | ;-) | 19:45 |
| mspreitz_ | both invoke scripts/run_kuryr.sh | 19:45 |
| banix | I have missed some of the logs above and will read but you shouldnt need to copy anything anywhere …. | 19:45 |
| mspreitz_ | You will see that I needed to make two copies of libexec | 19:46 |
| apuimedo | fkautz: can you file a bug on launchpad please so that we add debug level logging messages for calls to Neutron APIs? | 19:46 |
| banix | apuimedo: shouldn't we create a prper service for Kuryr to get things more steamlined… | 19:47 |
| banix | apuimedo: our logging in general need work…. have made some changes that i will push up soon | 19:48 |
| apuimedo | banix: when you mean proper service, you mean a systemd or upstart service files? | 19:48 |
| banix | yup | 19:48 |
| banix | :) | 19:48 |
| apuimedo | I have an upstart service one | 19:48 |
| apuimedo | I should push it to init/upstart | 19:49 |
| apuimedo | and we should put one for systemd | 19:49 |
| banix | apuimedo: that would be a great start | 19:49 |
| apuimedo | and then distros can decide if to pick those up | 19:49 |
| apuimedo | banix: you want the upstart one? | 19:49 |
| apuimedo | or do you use el7? | 19:49 |
| mspreitz_ | I use upstart today | 19:50 |
| banix | no but I am sure there are those who so | 19:50 |
| mspreitz_ | everyone uses Ubuntu 14.04 still, right? | 19:50 |
| apuimedo | well, I use arch linux | 19:51 |
| apuimedo | Ubuntu folks internally are already using 16.04 | 19:51 |
| banix | does that mean they are working in future apuimedo ? | 19:52 |
| mspreitz_ | last I heard, OpenStack gets the most/best testing on Ubuntu 14.04 LTS and it is a waste of my time to try anything else | 19:52 |
| apuimedo | banix: yes | 19:52 |
| banix | what time machine do they use? | 19:52 |
| apuimedo | I have a 16.04 deployment too :P | 19:52 |
| banix | you too?!!!! | 19:52 |
| apuimedo | I want to have kuryr ready for 16.04 | 19:52 |
| apuimedo | so this weekend I installed the alpha | 19:53 |
| banix | makes sense | 19:53 |
| mspreitz_ | so, getting back to the point, upstart support please. | 19:53 |
| apuimedo | I'll send it now | 19:53 |
| openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr: Add upstart config file https://review.openstack.org/280378 | 19:55 |
| apuimedo | banix: mspreitz_ ^^ | 19:55 |
| apuimedo | as you will see, it is rather simple | 19:56 |
| apuimedo | there is a patch I should send soon to run it as a regular user, not as root | 19:56 |
| apuimedo | but I never get to cleaning it up | 19:56 |
| banix | apuimedo: thanks and yes to the non root user patch :) | 19:58 |
| apuimedo | banix: it's going to be a contrib thing | 19:58 |
| apuimedo | with instructions for packagers, eh? | 19:58 |
| banix | you may have mentioned this before but what is the general approach? | 19:59 |
| apuimedo | it's a small C program that inherits CAP_NET_ADMIN | 20:04 |
| apuimedo | and launches the interpreter hardcoded to run Kuryr | 20:04 |
| *** openstackgerrit has quit IRC | 20:17 | |
| *** openstackgerrit_ is now known as openstackgerrit | 20:17 | |
| *** openstackgerrit has quit IRC | 20:17 | |
| *** openstackgerrit_ has joined #openstack-kuryr | 20:17 | |
| *** openstackgerrit_ is now known as openstackgerrit | 20:18 | |
| *** openstackgerrit_ has joined #openstack-kuryr | 20:18 | |
| fkautz | apuimedo: I'll file a bug | 20:29 |
| apuimedo | thanks | 20:29 |
| *** salv-orlando has joined #openstack-kuryr | 20:42 | |
| *** irenab_ has joined #openstack-kuryr | 21:00 | |
| *** irenab has quit IRC | 21:01 | |
| *** irenab_ is now known as irenab | 21:01 | |
| openstackgerrit | Mohammad Banikazemi proposed openstack/kuryr: Adds config options for using SSL https://review.openstack.org/280404 | 21:09 |
| openstackgerrit | Mohammad Banikazemi proposed openstack/kuryr: Add binding script for LB https://review.openstack.org/280411 | 21:28 |
| *** salv-orl_ has joined #openstack-kuryr | 22:03 | |
| *** salv-orlando has quit IRC | 22:05 | |
| *** banix has quit IRC | 22:10 | |
| *** itsuugo has quit IRC | 22:24 | |
| *** itsuugo has joined #openstack-kuryr | 22:27 | |
| *** WANG_Feng has quit IRC | 22:28 | |
| *** WANG_Feng has joined #openstack-kuryr | 22:28 | |
| *** shashank_hegde has quit IRC | 23:12 | |
| *** shashank_hegde has joined #openstack-kuryr | 23:26 | |
| *** yuanying_flu is now known as yuanying | 23:29 | |
| *** mspreitz_ has quit IRC | 23:44 | |
| *** banix has joined #openstack-kuryr | 23:45 | |
| *** reedip_away is now known as reedip | 23:50 | |
| *** banix has quit IRC | 23:55 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!