| *** gangil has joined #openstack-neutron-ovn | 00:14 | |
| *** gangil has joined #openstack-neutron-ovn | 00:14 | |
| *** azbiswas_ has quit IRC | 00:15 | |
| *** thumpba has quit IRC | 00:20 | |
| *** gangil1 has joined #openstack-neutron-ovn | 00:21 | |
| *** gangil has quit IRC | 00:21 | |
| *** yamamoto_ has joined #openstack-neutron-ovn | 00:24 | |
| *** yamamoto_ has quit IRC | 00:30 | |
| *** yamamoto_ has joined #openstack-neutron-ovn | 00:30 | |
| *** yamamot__ has joined #openstack-neutron-ovn | 00:34 | |
| *** yamamoto_ has quit IRC | 00:34 | |
| *** yamamot__ has quit IRC | 00:46 | |
| *** gongysh has joined #openstack-neutron-ovn | 01:22 | |
| *** chenli has joined #openstack-neutron-ovn | 01:43 | |
| *** chenli_ has joined #openstack-neutron-ovn | 01:58 | |
| *** jckasper has quit IRC | 01:59 | |
| *** jckasper has joined #openstack-neutron-ovn | 02:00 | |
| *** chenli has quit IRC | 02:00 | |
| *** chenli_ is now known as chenli | 02:00 | |
| *** jckasper has quit IRC | 02:05 | |
| *** jckasper has joined #openstack-neutron-ovn | 02:06 | |
| *** salv-orl_ has joined #openstack-neutron-ovn | 02:09 | |
| *** jckasper has quit IRC | 02:11 | |
| *** salv-orlando has quit IRC | 02:12 | |
| *** chenli has quit IRC | 02:35 | |
| *** yamamoto has joined #openstack-neutron-ovn | 02:46 | |
| *** azbiswas has joined #openstack-neutron-ovn | 02:48 | |
| *** yamamoto has quit IRC | 02:49 | |
| *** gangil1 has quit IRC | 02:52 | |
| *** chenli has joined #openstack-neutron-ovn | 02:53 | |
| *** s3wong has quit IRC | 03:04 | |
| *** azbiswas has quit IRC | 03:24 | |
| *** jamespd_ is now known as jamespd | 03:34 | |
| *** yamamoto has joined #openstack-neutron-ovn | 03:34 | |
| *** gangil has joined #openstack-neutron-ovn | 03:36 | |
| *** gangil has joined #openstack-neutron-ovn | 03:36 | |
| *** allan_h has quit IRC | 03:47 | |
| *** allan_h has joined #openstack-neutron-ovn | 03:47 | |
| *** jckasper has joined #openstack-neutron-ovn | 03:56 | |
| *** gongysh has quit IRC | 04:00 | |
| *** allan_h has quit IRC | 04:09 | |
| *** jckasper has quit IRC | 04:11 | |
| *** jckasper_ has joined #openstack-neutron-ovn | 04:13 | |
| *** jckasper_ has quit IRC | 04:15 | |
| *** jckasper_ has joined #openstack-neutron-ovn | 04:16 | |
| *** jckasper_ has quit IRC | 04:20 | |
| *** azbiswas has joined #openstack-neutron-ovn | 04:33 | |
| *** azbiswas has quit IRC | 04:35 | |
| *** jckasper has joined #openstack-neutron-ovn | 04:46 | |
| *** jckasper has quit IRC | 05:04 | |
| *** jckasper has joined #openstack-neutron-ovn | 05:05 | |
| *** jckasper_ has joined #openstack-neutron-ovn | 05:08 | |
| *** jckasper has quit IRC | 05:09 | |
| openstackgerrit | Merged openstack/networking-ovn: Update the HA section of the FAQ. https://review.openstack.org/274844 | 05:58 |
|---|---|---|
| *** armax_ has joined #openstack-neutron-ovn | 06:11 | |
| *** armax has quit IRC | 06:12 | |
| *** armax_ is now known as armax | 06:12 | |
| *** allan_h has joined #openstack-neutron-ovn | 06:17 | |
| *** allan_h has quit IRC | 06:20 | |
| *** numan_ has joined #openstack-neutron-ovn | 06:27 | |
| *** numan_ has quit IRC | 06:45 | |
| *** numans has joined #openstack-neutron-ovn | 06:46 | |
| *** gangil has quit IRC | 06:50 | |
| *** armax has quit IRC | 06:52 | |
| openstackgerrit | Babu Shanmugam proposed openstack/networking-ovn: DPDK support for OVN https://review.openstack.org/275103 | 07:32 |
| *** arosen has quit IRC | 08:04 | |
| *** arosen has joined #openstack-neutron-ovn | 08:04 | |
| *** arosen is now known as Guest55835 | 08:05 | |
| *** fzdarsky|afk has joined #openstack-neutron-ovn | 08:07 | |
| *** fzdarsky|afk is now known as fzdarsky | 08:09 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 08:10 | |
| *** salv-orl_ has quit IRC | 08:12 | |
| *** salv-orlando has quit IRC | 08:14 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 08:14 | |
| *** salv-orlando has quit IRC | 08:52 | |
| *** roeyc has joined #openstack-neutron-ovn | 09:28 | |
| *** Mic22 is now known as _Mic22 | 09:57 | |
| *** _Mic22 is now known as Mic22 | 09:57 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 10:16 | |
| *** chenli has quit IRC | 10:16 | |
| *** salv-orlando has quit IRC | 10:21 | |
| *** yamamoto has quit IRC | 10:27 | |
| *** roeyc has quit IRC | 10:57 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 11:07 | |
| *** fzdarsky is now known as fzdarsky|lunch | 11:19 | |
| *** yamamoto has joined #openstack-neutron-ovn | 11:32 | |
| *** yamamoto has quit IRC | 11:33 | |
| *** roeyc has joined #openstack-neutron-ovn | 11:48 | |
| *** rtheis has joined #openstack-neutron-ovn | 12:21 | |
| *** roeyc has quit IRC | 12:24 | |
| *** roeyc has joined #openstack-neutron-ovn | 12:35 | |
| -openstackstatus- NOTICE: Infra running with lower capacity now, due to a temporary problem affecting one of our nodepool providers. Please expect some delays in your jobs. Apologies for any inconvenience caused. | 12:40 | |
| *** fzdarsky|lunch is now known as fzdarsky | 12:59 | |
| *** openstackgerrit has quit IRC | 13:02 | |
| *** openstackgerrit has joined #openstack-neutron-ovn | 13:03 | |
| rtheis | When using neutron L3 agent with OVN, are routers suppose to be hosted by an L3 agent? | 14:07 |
| Sam-I-Am | y... yes? | 14:07 |
| rtheis | Recent deployment gives me 404 when I try to see the L3 agent hosting the router | 14:08 |
| rtheis | and no qrouter namespace exists | 14:09 |
| Sam-I-Am | interesting. i havent deployed with the conventional l3 agent recently. | 14:10 |
| *** yamamoto has joined #openstack-neutron-ovn | 14:11 | |
| *** thumpba has joined #openstack-neutron-ovn | 14:16 | |
| *** thumpba has quit IRC | 14:17 | |
| *** thumpba has joined #openstack-neutron-ovn | 14:17 | |
| russellb | the gate uses that mode | 14:18 |
| russellb | but i haven't used it myself recently either | 14:18 |
| Sam-I-Am | it does... | 14:19 |
| Sam-I-Am | russellb: did you see the mtu thread? | 14:19 |
| Sam-I-Am | the ovn one | 14:19 |
| Sam-I-Am | also, mornings! | 14:19 |
| russellb | i saw it, read jesse's first response last night | 14:19 |
| russellb | haven't gone back yet this morning | 14:19 |
| Sam-I-Am | its not much more | 14:20 |
| russellb | "yep MTU is hard" | 14:21 |
| Sam-I-Am | it should not be... hard. | 14:23 |
| numans | russellb, Hi, "pip install ovs" is installing 2.4.0, is it because 2.6.0.dev0 is a developer version ? https://pypi.python.org/pypi/ovs | 14:23 |
| Sam-I-Am | we cant break mtu in yet another neutron thing :/ | 14:24 |
| russellb | numans: i'm not sure... is that when you run "pip install" manually? | 14:24 |
| Sam-I-Am | this one with a seemingly more difficult way to fix it than conventional ovs/linuxbridge agents | 14:24 |
| Sam-I-Am | only thing i can hope from my experiments with conventional OVS is that it Just Works once packets get into OVS... and we just need to concern ourselves with providing the right mtu to instances | 14:25 |
| Sam-I-Am | but i cant test that until L3 works | 14:25 |
| numans | russellb, yes. | 14:26 |
| numans | russellb, when i run "pip install ovs" i see 2.4.0 version | 14:26 |
| russellb | ok ... i don't know the answer | 14:26 |
| numans | russellb, do we need to add 2.6.0 in requirements.txt ? | 14:26 |
| russellb | try "pip install --upgrade ovs" | 14:27 |
| numans | russellb, i tried its the same. when i run "pip install ovs==2.6.0.dev" its install 2.6.0 version | 14:27 |
| russellb | ok, well i can't really change the "dev" part, because that's what it is | 14:29 |
| russellb | i could backport this to the 2.4 branch and release a 2.4.1 perhaps ... | 14:30 |
| *** jckasper_ has quit IRC | 14:30 | |
| russellb | "this", i mean the lock fix | 14:30 |
| russellb | i assume that's what you need? | 14:30 |
| *** jckasper has joined #openstack-neutron-ovn | 14:30 | |
| numans | russellb, yes | 14:32 |
| numans | russellb, thanks | 14:32 |
| *** jckasper has quit IRC | 14:33 | |
| russellb | the pain here is actually that the ovs python library is bundled with ovs, so our version numbers of python releases have to match | 14:33 |
| *** jckasper has joined #openstack-neutron-ovn | 14:33 | |
| russellb | i can't really call it 2.4.1 unless ovs itself releases a 2.4.1 | 14:34 |
| numans | ok | 14:34 |
| russellb | i could do something like ... 2.4.1rc1 | 14:34 |
| russellb | but that probably doesn't help | 14:34 |
| numans | russellb, you think its a good idea to use master python ovs in networking-ovn like the way it uses ovs | 14:37 |
| russellb | yes, might as well | 14:38 |
| russellb | we could just pip install from the ovs git tree then.. | 14:38 |
| numans | yes.. i tried sometime to test that fix. i can submit a patch for that | 14:39 |
| russellb | ok, that works | 14:39 |
| numans | great. i will submit a patch then | 14:39 |
| russellb | setup_package $DEST/networking-ovn | 14:40 |
| russellb | i don't remember what all setup_package does ... | 14:40 |
| russellb | but i'd look at that | 14:40 |
| russellb | setup_package $DEST/$OVN_REPO_NAME/python | 14:40 |
| russellb | or something | 14:40 |
| *** regXboi has joined #openstack-neutron-ovn | 14:41 | |
| *** yamamoto has quit IRC | 14:43 | |
| mestery | rtheis: Hostname mismatch? I've seen that before | 14:44 |
| mestery | russellb: When you have a moment, would appreciate a gentle nudge on this one: https://review.openstack.org/#/c/275397/ | 14:44 |
| russellb | mestery: done | 14:45 |
| rtheis | mestery: I deployed all-in-one with devstack. Where should I check for hostname mismatch? | 14:46 |
| *** yamamoto has joined #openstack-neutron-ovn | 14:47 | |
| *** tongli has joined #openstack-neutron-ovn | 14:47 | |
| tongli | @Sam-I-Am, hi matt, are you there? | 14:48 |
| Sam-I-Am | tongli: hi | 14:48 |
| tongli | when I run vagrant up today, it keeps complaining the devstack and networking-ovn directory do not exist. | 14:49 |
| Sam-I-Am | tongli: couple things... | 14:50 |
| Sam-I-Am | the docs say you need to clone those repos prior to running vagrant | 14:50 |
| openstackgerrit | Merged openstack/networking-ovn: Vagrant: Stop sharing the ~/devstack directory https://review.openstack.org/275397 | 14:50 |
| Sam-I-Am | then kyle put in a patch to remove devstack from the shared directories, so when that patch merges, you only need to clone networking-ovn prior to running vagrant | 14:50 |
| tongli | oh, that is a recent change, right? | 14:50 |
| Sam-I-Am | for whatever reason, vagrant does not handle missing shared directories well | 14:50 |
| Sam-I-Am | it should just... not share them. but it fails instead. | 14:50 |
| Sam-I-Am | the vagrant stuff is getting lots of changes lately | 14:51 |
| mestery | russellb: Thanks my friend :) | 14:51 |
| russellb | np | 14:51 |
| *** yamamoto_ has joined #openstack-neutron-ovn | 14:52 | |
| mestery | rtheis: So you have no additional compute nodes? If that's the case, then this isn't hte problem, as it would only affect you if you had additional compute nodes | 14:52 |
| *** yamamoto has quit IRC | 14:52 | |
| rtheis | mestery: sure, I don't have any additional compute nodes at this time. I did add a network node running only the dhcp agent. | 14:53 |
| Mic22 | Hi guys, i get a strange error. I have 2 computes with two ports hooked in the system to get connectivity. From every box ping works but i can ssh only on vms hosted on the same box | 14:53 |
| Mic22 | security roules allow all | 14:54 |
| Mic22 | any ideas? | 14:54 |
| Mic22 | from tshark this is the output: | 14:55 |
| Mic22 | 14 0.212833923 10.0.0.201 -> 10.0.0.12 TCP 1516 [TCP Retransmission] 49040 > ssh [ACK] Seq=24 Ack=902 Win=30976 Len=1448 TSval=85298655 TSecr=513833 | 14:55 |
| Mic22 | 15 0.617814998 10.0.0.201 -> 10.0.0.12 TCP 1516 [TCP Retransmission] 49040 > ssh [ACK] Seq=24 Ack=902 Win=30976 Len=1448 TSval=85299060 TSecr=513833 | 14:55 |
| Mic22 | 16 1.427805213 10.0.0.201 -> 10.0.0.12 TCP 1516 [TCP Retransmission] 49040 > ssh [ACK] Seq=24 Ack=902 Win=30976 Len=1448 TSval=85299870 TSecr=513833 | 14:55 |
| Mic22 | 17 3.045842277 10.0.0.201 -> 10.0.0.12 TCP 1516 [TCP Retransmission] 49040 > ssh [ACK] Seq=24 Ack=902 Win=30976 Len=1448 TSval=85301488 TSecr=513833 | 14:55 |
| russellb | i think i missed some scrollback (laptop locked up on me) | 14:56 |
| russellb | but, if ping is working but ssh fails during connection setup, that's usually an MTU problem | 14:57 |
| Sam-I-Am | does it establish a tcp session? | 14:57 |
| openstackgerrit | Richard Theis proposed openstack/networking-ovn: WIP: Add network availability zone support https://review.openstack.org/275415 | 14:57 |
| *** yamamoto_ has quit IRC | 14:57 | |
| russellb | you can usually run ssh with --debug and see that it fails during key exchange or something, i can't remember exactly | 14:58 |
| russellb | not fail, but hangs | 14:58 |
| Sam-I-Am | use -v | 14:58 |
| Mic22 | russellb you got the problem at the first attempt! | 14:58 |
| Mic22 | it way mtu | 14:58 |
| russellb | yay | 14:59 |
| Sam-I-Am | well, not yay | 14:59 |
| russellb | right | 14:59 |
| Mic22 | i disabled dhcp booted with static addressed and | 14:59 |
| Sam-I-Am | last thing we need are mtu problems :/ | 14:59 |
| russellb | Sam-I-Am: fwiw i've mentally assigned "MTU strategy for OVN" to you | 14:59 |
| Mic22 | mtu was 1500 | 14:59 |
| Sam-I-Am | hopefully this is just the vm mtu problem | 14:59 |
| tongli | @Sam-I-Am, so is there someone fixing that problem? so that vagrant can run, now even if I clone devstack, the sync still breaks. | 14:59 |
| Sam-I-Am | Mic22: try subtracting 42 for that for geneve... i think thats the oberhead number | 14:59 |
| regXboi | russellb: so... I think I may have a poor-man's incremental approach to lflow_run - I'm hoping to run a test later today and if it works, I'll send a patch set to dev | 14:59 |
| Sam-I-Am | tongli: did you clone both networking-ovn and devstack to your home dir? | 15:00 |
| russellb | regXboi: cool :) | 15:00 |
| Mic22 | problem is I need another overlay in the vm for kubernates | 15:00 |
| Mic22 | so i have double overhead | 15:00 |
| Sam-I-Am | yuck | 15:00 |
| regXboi | russellb: yeah, right now I'm running a test of a poor-man's poor-man's change for other issues and have to wait for it to finish first | 15:01 |
| *** yamamoto has joined #openstack-neutron-ovn | 15:01 | |
| tongli | @Sam-I-Am, oh, i c what the problem is. | 15:01 |
| Sam-I-Am | i havent used kubernetes, but with lxc the containers were just bridged rather than tunneled | 15:01 |
| tongli | ok. let me try that. | 15:01 |
| Mic22 | Sam-I-Am kubernates requires some networking specs where every vm has it's own /24 | 15:02 |
| Mic22 | and every /24 must be routable each others | 15:02 |
| Sam-I-Am | thats... interesting | 15:02 |
| Mic22 | in the current situation of OVN the only solution I can see is 254 ports per vm and this is a little bit fucked up | 15:03 |
| Mic22 | don't you think? | 15:03 |
| Sam-I-Am | i guess. like i said... havent messed with kubes | 15:03 |
| Sam-I-Am | just lxc and docker | 15:03 |
| *** shettyg has joined #openstack-neutron-ovn | 15:04 | |
| Mic22 | well actually i think i overkilled... I just need to setup routes in the vms... | 15:08 |
| *** numans has quit IRC | 15:10 | |
| tongli | @Sam-I-Am, can we change the synch directory like this "../../networking-ovn" rather than "~/networking-ovn"? | 15:10 |
| tongli | using ~/networking-ovn, one has to place the directory at the root. | 15:11 |
| tongli | vs ../../networking-ovn, we just need to make sure that the devstack project and networking-ovn at the same level. | 15:11 |
| tongli | normally one's root directory contains so much stuff, kind not nice to have a hard requirement like that. | 15:11 |
| Sam-I-Am | ~ is the home dir shortcut and ../.. would always assume networking-ovn is two dirs up, which it may not necessarily be | 15:12 |
| Sam-I-Am | why are you running this as root? | 15:12 |
| Sam-I-Am | i'm deploying all this as a user, and my stuff is in /home/matt/networking-ovn | 15:12 |
| Sam-I-Am | in which case ~ references /home/matt | 15:12 |
| tongli | @Sam-I-Am, not as root, | 15:12 |
| tongli | root directory, see this line in Vagrantfile. | 15:12 |
| tongli | config.vm.synced_folder "~/networking-ovn", "/home/vagrant/networking-ovn" | 15:13 |
| Sam-I-Am | what about it? | 15:13 |
| tongli | that means the vagrant will always look for networking-ovn directory from the root directory of the current user | 15:13 |
| Sam-I-Am | yeah... like /home/user/networking-ovn | 15:14 |
| tongli | if a user did not clone that project at root directory, it will fail. | 15:14 |
| Sam-I-Am | yes. but theres no way in vagrant to loosely specify directories | 15:15 |
| Sam-I-Am | it would be nice if vagrant would ignore missing stuff, but it does not | 15:15 |
| tongli | vs if we just use "../../networking-ovn", then one can clone the devstack and networking-ovn project any where as long as these two projects are at the same directory. | 15:15 |
| Sam-I-Am | ../../ still implies that networking-ovn is two dirs above the Vagrantfile | 15:15 |
| Sam-I-Am | so that doesnt really fix the problem | 15:15 |
| tongli | yes, that is always the case. | 15:16 |
| tongli | since Vagrantfile is part of networking-ovn project and is always in the vagrant folder. | 15:16 |
| Sam-I-Am | it just means you can have networking-ovn further down the dir tree, but it would always have to be two dirs up from the vagrant file ... sort of just changes the problem | 15:16 |
| tongli | that will not change. | 15:16 |
| Sam-I-Am | tongli: you can also just change the vagrant file locally | 15:16 |
| Sam-I-Am | whats there is just a suggestion | 15:17 |
| tongli | Vagrantfile is always at the vagrant directory, no? | 15:17 |
| Sam-I-Am | no one is going to be happy with every decision in it | 15:17 |
| Sam-I-Am | yes | 15:17 |
| Sam-I-Am | i think most people are going to change at least one thing in the vagrant stuff before deploying | 15:17 |
| tongli | I do not think that you can move the Vagrantfile randomly without changing other scripts locations. | 15:17 |
| mestery | Sam-I-Am: https://review.openstack.org/#/c/275748/ | 15:18 |
| tongli | I assume that the relationship between Vagrantfile and other scripts are relatively stable. | 15:18 |
| Sam-I-Am | mestery: one sec | 15:19 |
| Sam-I-Am | mestery: found a thing | 15:23 |
| mestery | Sam-I-Am: nice | 15:25 |
| russellb | in case you didn't hate software enough, check out this pip upgrade attempt: http://paste.openstack.org/show/485859/ | 15:25 |
| russellb | and read the version numbers mentioned throughout | 15:25 |
| openstackgerrit | Kyle Mestery proposed openstack/networking-ovn: Vagrant: Increase the size of the provider network subnet https://review.openstack.org/275748 | 15:28 |
| mestery | Woot! russellb, all your patches from your localnet branch are now upstream! | 15:41 |
| mestery | Awesome! | 15:41 |
| Sam-I-Am | wooo | 15:41 |
| russellb | :) | 15:43 |
| russellb | yeah i just pushed a bunch of stuff | 15:43 |
| Sam-I-Am | mestery: https://review.openstack.org/#/c/275418/ | 15:44 |
| Sam-I-Am | it aint pretty, but no other choice right now | 15:44 |
| mestery | Sam-I-Am: nice | 15:45 |
| * mestery hands russellb cake | 15:45 | |
| openstackgerrit | Matthew Kassawara proposed openstack/networking-ovn: Devstack: Add native MTU option https://review.openstack.org/275418 | 15:47 |
| Sam-I-Am | i like how my linter job failed due to an in-between merge of a doc with a long line | 15:48 |
| Sam-I-Am | er, patch enabling the linter job | 15:48 |
| Sam-I-Am | thats hilarious | 15:48 |
| regXboi | Sam-I-Am: *of course* it would | 15:54 |
| *** salv-orl_ has joined #openstack-neutron-ovn | 15:54 | |
| *** thumpba has quit IRC | 15:55 | |
| *** salv-orlando has quit IRC | 15:57 | |
| openstackgerrit | Matthew Kassawara proposed openstack/networking-ovn: Docs: Implement doc8 linter https://review.openstack.org/274948 | 16:17 |
| Sam-I-Am | russellb: i dont think i can +2 this... https://review.openstack.org/#/c/274236/ | 16:25 |
| Sam-I-Am | but it should get in | 16:25 |
| russellb | there are 2 patches ahead of it that have to merge first | 16:26 |
| russellb | but you're right, it's not really appropriate for either of us to +2 a patch we helped write | 16:27 |
| Sam-I-Am | if its dependent, another +2A would let it launch when the other patches are approved | 16:27 |
| russellb | Sam-I-Am: you and mestery could both merge https://review.openstack.org/#/c/274234/2 | 16:28 |
| russellb | the 2nd patch is already approved (blocked on the first) | 16:28 |
| Sam-I-Am | ok | 16:29 |
| mestery | Looking | 16:29 |
| Sam-I-Am | might need to check 274234 for long lines and stuff | 16:29 |
| Sam-I-Am | let me lint it | 16:29 |
| Sam-I-Am | so my linter patch doesnt blow up again :) | 16:30 |
| * mestery waits for Sam-I-Am | 16:30 | |
| Sam-I-Am | looks good for now, lets push it along. | 16:31 |
| Sam-I-Am | i can edit later | 16:31 |
| * Sam-I-Am looks at large pile of things to clarify | 16:31 | |
| Sam-I-Am | my morning isnt going so well... coffee pot malfunction caused a lot of coffee to go everywhere else :/ | 16:32 |
| Sam-I-Am | whats worse is not having coffee while cleaning up the mess | 16:32 |
| Sam-I-Am | licking it up off the counter was not successful | 16:32 |
| openstackgerrit | Merged openstack/networking-ovn: doc: Further reflect that OVN L3 is the default. https://review.openstack.org/274234 | 16:35 |
| *** thumpba has joined #openstack-neutron-ovn | 16:35 | |
| russellb | my morning is slow, mostly because of a cold i'm fighting | 16:36 |
| Sam-I-Am | yuck | 16:36 |
| mestery | Colds are no fun :( | 16:36 |
| Sam-I-Am | well, i guess my morning is a bit better | 16:37 |
| openstackgerrit | Merged openstack/networking-ovn: doc: Use ovn-sbctl show. https://review.openstack.org/274235 | 16:37 |
| Sam-I-Am | although my countertop and part of the floor is well caffeinated, and beginning to say "feeeeed meeeeee" | 16:37 |
| russellb | Sam-I-Am: https://review.openstack.org/#/c/274948/3 might need a rebase, it says "Cannot Merge" | 16:45 |
| Sam-I-Am | russellb: i just saw that | 16:45 |
| *** armax has joined #openstack-neutron-ovn | 16:46 | |
| Sam-I-Am | its good when docs have merge conflicts, it means people are working on them :) | 16:46 |
| russellb | nice perspective | 16:46 |
| Sam-I-Am | hmm this rebase is nasty | 16:52 |
| Sam-I-Am | git and docs dont get along well | 16:52 |
| russellb | at least it's plain text docs and not xml? | 16:52 |
| Sam-I-Am | YES | 16:52 |
| * russellb sadly has docbook experience | 16:53 | |
| Sam-I-Am | docbook is a good part of the reason why openstack docs suck | 16:53 |
| Sam-I-Am | no one wants any part of that | 16:53 |
| Sam-I-Am | we've mostly converted to rst, but it was a day late and a dollar short... and the api stuff (which uses wadl... docbook+worse) is still in shambles | 16:54 |
| * russellb can provide a dollar if that helps | 16:55 | |
| *** allan_h has joined #openstack-neutron-ovn | 17:16 | |
| *** yamamoto has quit IRC | 17:17 | |
| Mic22 | russelb: the behaviour I had yesterday | 17:25 |
| Mic22 | happend only when you add a new compute node after having networks already configured | 17:25 |
| Mic22 | if i add a compute and then do all the network configurations I have no problems. In the moment I add a compute with already network config nothing works | 17:27 |
| Mic22 | after some minutes it start to works | 17:33 |
| *** roeyc has quit IRC | 17:45 | |
| openstackgerrit | Matthew Kassawara proposed openstack/networking-ovn: Docs: Implement doc8 linter https://review.openstack.org/274948 | 17:45 |
| Sam-I-Am | lets try this again | 17:49 |
| russellb | Mic22: cool, sounds like you're getting closer to a clear way to reproduce it! | 17:49 |
| russellb | that's great | 17:49 |
| Mic22 | I think i can reproduce it, if you have time tomorrow we can reproduce it together so you can see what really happends | 17:50 |
| russellb | ok | 17:51 |
| russellb | you could also document the info you have in a bug report | 17:51 |
| russellb | so we start gathering info | 17:51 |
| Mic22 | you can tell me exacly what you need as outputs and i will do it | 17:51 |
| russellb | let's see ... | 17:52 |
| russellb | i suppose i could look now | 17:52 |
| Mic22 | now it's working and i have to go home, I live in london+1 time... | 17:52 |
| russellb | ah, ok | 17:53 |
| russellb | sorry about that | 17:53 |
| russellb | tomorrow morning works | 17:53 |
| Mic22 | yes ofc | 17:53 |
| russellb | when it's in the broken state, a general dump would be something like ... ovn-nbctl show | 17:53 |
| russellb | ovn-sbctl show | 17:53 |
| russellb | ovn-sbctl lflow-list | 17:53 |
| russellb | ovs-ofctl -O OpenFlow13 dump-flows br-int (on every host) | 17:54 |
| russellb | ovs-ofctl -O OpenFlow13 show br-int (on every host) | 17:54 |
| russellb | something like that | 17:54 |
| russellb | have a nice evening! | 17:54 |
| Mic22 | thanks | 17:57 |
| Mic22 | where do you want this stuff posted? | 17:57 |
| *** gangil has joined #openstack-neutron-ovn | 18:40 | |
| *** gangil has joined #openstack-neutron-ovn | 18:40 | |
| *** jckasper has quit IRC | 18:41 | |
| *** jckasper has joined #openstack-neutron-ovn | 18:42 | |
| *** fzdarsky is now known as fzdarsky|afk | 18:46 | |
| *** jckasper has quit IRC | 18:46 | |
| rtheis | Anyone else seeing "/opt/stack/new/networking-ovn/tools/tox_install.sh: 21: /opt/stack/new/networking-ovn/tools/tox_install.sh: Bad substitution" on devstack deployment ? | 18:58 |
| russellb | Mic22: good question ... paste.openstack.org for now i suppose | 19:00 |
| russellb | or gist.github.com, or your paste site of choice | 19:00 |
| russellb | Sam-I-Am: vagrant bug for you https://bugs.launchpad.net/bugs/1541542 | 19:00 |
| Sam-I-Am | russellb: what version of tox? | 19:00 |
| openstack | Launchpad bug 1541542 in networking-ovn "vagrant netmask parameter can not use the length" [Undecided,New] | 19:00 |
| Sam-I-Am | err | 19:01 |
| Sam-I-Am | rtheis: what version of tox? | 19:01 |
| Sam-I-Am | russellb: awesome, so apparently vagrant doesnt support cidr notation :/ | 19:02 |
| Sam-I-Am | that should be a no-brainer these days | 19:02 |
| Sam-I-Am | it also doesnt support working around missing shared directories either :/ | 19:02 |
| *** s3wong has joined #openstack-neutron-ovn | 19:03 | |
| rtheis | Sam-I-Am: worked a couple days ago with tox 2.3.1... retrying now and will let you know | 19:04 |
| Sam-I-Am | russellb: heyyy... should i be able to manage networking-ovn bugs? | 19:04 |
| russellb | Sam-I-Am: if you can't, let's fix that | 19:04 |
| Sam-I-Am | rtheis: ok, i've seen similar problems and turns out they were not using 2.3.1 | 19:04 |
| rtheis | I just saw gate-tempest-dsvm-networking-ovn fail on a couple patch sets with same error | 19:04 |
| Sam-I-Am | rtheis: did you pastebin? | 19:05 |
| *** jckasper has joined #openstack-neutron-ovn | 19:05 | |
| rtheis | Sam-I-Am: no, what do I do for that | 19:05 |
| russellb | mestery: how do i make it so Sam-I-Am can manage networking-ovn bugs? | 19:06 |
| Sam-I-Am | rtheis: go to paste.openstack.org and paste the full tox error... and how you got there. | 19:06 |
| rtheis | Sam-I-Am: okay | 19:07 |
| Sam-I-Am | rtheis: is this a similar failure? | 19:08 |
| Sam-I-Am | http://logs.openstack.org/92/275392/2/check/gate-tempest-dsvm-networking-ovn/20e990f/ | 19:08 |
| Sam-I-Am | i havent looked closely at it | 19:08 |
| rtheis | Sam-I-Am: yes it is | 19:09 |
| rtheis | http://logs.openstack.org/92/275392/2/check/gate-tempest-dsvm-networking-ovn/20e990f/logs/devstacklog.txt.gz | 19:09 |
| rtheis | 2016-02-03 18:48:00.314 | /opt/stack/new/networking-ovn/tools/tox_install.sh: 21: /opt/stack/new/networking-ovn/tools/tox_install.sh: Bad substitution | 19:09 |
| rtheis | It is the exact same failure that I'm seeing locally. | 19:10 |
| russellb | looks like something has changed that broke networking-ovn to me | 19:11 |
| mestery | russellb: You need to be in this group https://launchpad.net/~networking-ovn-bugs | 19:14 |
| mestery | russellb: I added you and Sam-I-Am | 19:14 |
| russellb | oh, hm | 19:14 |
| russellb | i was already able to do everything though | 19:14 |
| mestery | Weird | 19:14 |
| * russellb shrugs | 19:14 | |
| russellb | moving on! | 19:15 |
| mestery | lol | 19:15 |
| *** chandrav has joined #openstack-neutron-ovn | 19:18 | |
| *** azbiswas has joined #openstack-neutron-ovn | 19:31 | |
| regXboi | well... this certainly *looks* right | 20:19 |
| russellb | :shipit: | 20:22 |
| regXboi | russelb: waiting for the test to finish | 20:23 |
| regXboi | russellb: but this looks hopeful: | 20:24 |
| regXboi | 2016-02-03T20:24:43.845Z|85669|lflow|DBG|start lflow_run | 20:24 |
| regXboi | 2016-02-03T20:24:43.857Z|85670|lflow|DBG|finished lflow_run. 9650 flows seen, 764 flows worked, 756 matches. | 20:24 |
| regXboi | poor man's incremental processing :) | 20:25 |
| Sam-I-Am | rtheis: have you looked any more into the tox problem? | 21:24 |
| Sam-I-Am | looks like its breaking on genconfig, which is odd because that works fine locally | 21:24 |
| rtheis | Sam-I-Am: I have not | 21:24 |
| mestery | It fails for me too | 21:31 |
| mestery | With all the latest | 21:32 |
| mestery | #sadpanda | 21:32 |
| * mestery shakes his fist at tox and genconfig | 21:32 | |
| Sam-I-Am | mestery: locally? | 21:32 |
| mestery | Yes, in my vagrant | 21:32 |
| Sam-I-Am | weird, it works in my regular source tree | 21:32 |
| Sam-I-Am | just by running tox -e genconfig | 21:32 |
| mestery | Yes | 21:32 |
| openstackgerrit | Matthew Kassawara proposed openstack/networking-ovn: Fix gate problem with tox https://review.openstack.org/275939 | 21:40 |
| Sam-I-Am | errr | 21:44 |
| openstackgerrit | litong01 proposed openstack/networking-ovn: Vagrant: Increase the size of the provider network subnet https://review.openstack.org/275748 | 21:50 |
| *** salv-orlando has joined #openstack-neutron-ovn | 21:54 | |
| openstackgerrit | Doug Hellmann proposed openstack/networking-ovn: Devstack: Provide metadata route via DHCP https://review.openstack.org/275392 | 21:55 |
| *** thumpba has quit IRC | 21:55 | |
| openstackgerrit | litong01 proposed openstack/networking-ovn: Vagrant: Increase the size of the provider network subnet https://review.openstack.org/275748 | 21:56 |
| *** salv-orl_ has quit IRC | 21:56 | |
| mestery | Wow, dhellman re-proposed this one for you Sam-I-Am? https://review.openstack.org/#/c/275392/ | 21:57 |
| mestery | What did I miss? | 21:57 |
| Sam-I-Am | mestery: git andd i were not getting along | 21:57 |
| mestery | lol | 21:57 |
| Sam-I-Am | was trying to insert the tox patch under the devstack patch and ended up in bad places | 21:58 |
| mestery | So, "tox -egenconfig" works fine AFTER it fails | 22:01 |
| mestery | weird | 22:01 |
| Sam-I-Am | mestery: that 275748... | 22:02 |
| Sam-I-Am | not sure if the benefits outweigh the added complexity | 22:03 |
| Sam-I-Am | the vagrant nets are local to the host, so there's no good reason to change them outside of expanding the provider net to /16 for scale testing | 22:03 |
| Sam-I-Am | if we were connecting these nets outside of the host running the VMs, thats another story | 22:04 |
| mestery | Oh, it's worth it :) | 22:05 |
| Sam-I-Am | the implementation is... | 22:05 |
| Sam-I-Am | its not easy to follow | 22:05 |
| *** jckasper has quit IRC | 22:07 | |
| *** jckasper has joined #openstack-neutron-ovn | 22:10 | |
| openstackgerrit | litong01 proposed openstack/networking-ovn: Vagrant: Increase the size of the provider network subnet https://review.openstack.org/275748 | 22:19 |
| *** rtheis has quit IRC | 22:30 | |
| *** shettyg has quit IRC | 23:01 | |
| *** regXboi has quit IRC | 23:03 | |
| *** chandrav has quit IRC | 23:04 | |
| *** azbiswas has quit IRC | 23:39 | |
| *** allan_h has quit IRC | 23:51 | |
| *** allan_h has joined #openstack-neutron-ovn | 23:51 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!