| *** jerrygb has quit IRC | 00:03 | |
| *** slowe has quit IRC | 00:06 | |
| *** carl_baldwin has quit IRC | 00:27 | |
| *** salv-orlando has quit IRC | 00:27 | |
| *** jerrygb has joined #openstack-neutron-ovn | 01:38 | |
| *** azbiswas has quit IRC | 01:43 | |
| *** yamamoto has joined #openstack-neutron-ovn | 01:51 | |
| *** yamamoto has quit IRC | 02:05 | |
| *** yamamoto has joined #openstack-neutron-ovn | 02:10 | |
| *** yamamoto has quit IRC | 02:40 | |
| *** gangil has quit IRC | 02:42 | |
| *** yamamoto has joined #openstack-neutron-ovn | 02:46 | |
| *** yamamoto has quit IRC | 02:46 | |
| *** chandrav has joined #openstack-neutron-ovn | 02:49 | |
| *** yamamoto has joined #openstack-neutron-ovn | 02:50 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 02:51 | |
| *** yamamoto has quit IRC | 02:55 | |
| *** yamamoto has joined #openstack-neutron-ovn | 03:01 | |
| *** chandrav has quit IRC | 03:05 | |
| *** azbiswas has joined #openstack-neutron-ovn | 03:07 | |
| *** salv-orlando has quit IRC | 03:12 | |
| *** yamamoto has quit IRC | 03:24 | |
| *** gangil has joined #openstack-neutron-ovn | 03:25 | |
| *** gangil has joined #openstack-neutron-ovn | 03:25 | |
| *** armax has quit IRC | 03:47 | |
| *** carl_baldwin has joined #openstack-neutron-ovn | 03:53 | |
| *** azbiswas has quit IRC | 04:05 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 04:09 | |
| *** salv-orlando has quit IRC | 04:34 | |
| *** azbiswas has joined #openstack-neutron-ovn | 04:34 | |
| *** carl_baldwin has quit IRC | 04:37 | |
| *** carl_baldwin has joined #openstack-neutron-ovn | 04:38 | |
| *** arosen has quit IRC | 04:40 | |
| *** shettyg1 has quit IRC | 04:43 | |
| *** yamamoto has joined #openstack-neutron-ovn | 04:51 | |
| *** yamamoto has quit IRC | 04:57 | |
| *** nate_gone is now known as njohnston | 04:58 | |
| *** njohnston is now known as nate_gone | 05:10 | |
| *** carl_baldwin has quit IRC | 05:19 | |
| *** carl_baldwin has joined #openstack-neutron-ovn | 05:21 | |
| *** ajo has quit IRC | 05:22 | |
| *** ajo has joined #openstack-neutron-ovn | 05:27 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 05:31 | |
| *** salv-orlando has quit IRC | 05:55 | |
| *** azbiswas has quit IRC | 06:13 | |
| *** jerrygb has quit IRC | 06:14 | |
| *** carl_baldwin has quit IRC | 06:18 | |
| *** azbiswas has joined #openstack-neutron-ovn | 06:48 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 06:48 | |
| *** openstackstatus has quit IRC | 06:49 | |
| *** nate_gone is now known as njohnston | 06:49 | |
| *** openstackstatus has joined #openstack-neutron-ovn | 06:50 | |
| *** ChanServ sets mode: +v openstackstatus | 06:50 | |
| *** azbiswas has quit IRC | 06:52 | |
| *** njohnston is now known as nate_gone | 06:58 | |
| *** jerrygb has joined #openstack-neutron-ovn | 07:15 | |
| *** jerrygb has quit IRC | 07:19 | |
| *** salv-orlando has quit IRC | 07:46 | |
| *** gangil has quit IRC | 08:07 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 08:32 | |
| *** subscope has joined #openstack-neutron-ovn | 08:39 | |
| *** yamamoto has joined #openstack-neutron-ovn | 08:43 | |
| *** yamamoto has quit IRC | 08:50 | |
| *** yamamoto has joined #openstack-neutron-ovn | 08:52 | |
| *** subscope has quit IRC | 09:01 | |
| *** subscope has joined #openstack-neutron-ovn | 09:04 | |
| *** yamamoto has quit IRC | 09:07 | |
| *** jerrygb has joined #openstack-neutron-ovn | 09:16 | |
| *** asuvvari has joined #openstack-neutron-ovn | 09:16 | |
| *** asuvvari has quit IRC | 09:20 | |
| *** jerrygb has quit IRC | 09:21 | |
| *** yamamoto has joined #openstack-neutron-ovn | 09:30 | |
| *** subscope has quit IRC | 09:40 | |
| *** yamamoto has quit IRC | 09:41 | |
| *** salv-orlando has quit IRC | 09:41 | |
| *** subscope has joined #openstack-neutron-ovn | 09:59 | |
| *** subscope has quit IRC | 10:20 | |
| *** subscope has joined #openstack-neutron-ovn | 10:20 | |
| *** nate_gone is now known as njohnston | 10:25 | |
| *** njohnston is now known as nate_gone | 10:35 | |
| *** yamamoto has joined #openstack-neutron-ovn | 10:41 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 10:43 | |
| *** yamamoto has quit IRC | 10:46 | |
| *** salv-orlando has quit IRC | 10:50 | |
| *** azbiswas has joined #openstack-neutron-ovn | 10:50 | |
| *** subscope has quit IRC | 10:53 | |
| *** azbiswas has quit IRC | 10:55 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 10:59 | |
| *** salv-orlando has quit IRC | 11:02 | |
| *** yamamoto has joined #openstack-neutron-ovn | 11:34 | |
| *** yamamoto has quit IRC | 11:45 | |
| *** yamamoto has joined #openstack-neutron-ovn | 11:50 | |
| *** jerrygb has joined #openstack-neutron-ovn | 12:03 | |
| *** yamamoto has quit IRC | 12:05 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 12:06 | |
| *** subscope has joined #openstack-neutron-ovn | 12:10 | |
| *** nate_gone is now known as njohnston | 12:14 | |
| *** salv-orl_ has joined #openstack-neutron-ovn | 12:15 | |
| *** salv-orlando has quit IRC | 12:15 | |
| *** njohnston is now known as nate_gone | 12:24 | |
| *** subscope has quit IRC | 12:24 | |
| *** subscope has joined #openstack-neutron-ovn | 12:26 | |
| *** salv-orl_ has quit IRC | 12:29 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 12:31 | |
| *** rtheis has joined #openstack-neutron-ovn | 12:38 | |
| *** subscope has quit IRC | 12:50 | |
| *** subscope has joined #openstack-neutron-ovn | 12:58 | |
| *** subscope has quit IRC | 13:00 | |
| *** jerrygb has joined #openstack-neutron-ovn | 13:01 | |
| *** salv-orlando has quit IRC | 13:37 | |
| *** subscope has joined #openstack-neutron-ovn | 13:38 | |
| *** yamamoto has joined #openstack-neutron-ovn | 13:42 | |
| *** yamamoto has quit IRC | 13:48 | |
| *** subscope has quit IRC | 13:55 | |
| *** nate_gone is now known as njohnston | 14:03 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 14:05 | |
| *** subscope has joined #openstack-neutron-ovn | 14:12 | |
| *** njohnston is now known as nate_gone | 14:13 | |
| *** regXboi has joined #openstack-neutron-ovn | 14:28 | |
| *** salv-orlando has quit IRC | 14:39 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 14:39 | |
| *** azbiswas has joined #openstack-neutron-ovn | 14:54 | |
| *** gongysh_ has joined #openstack-neutron-ovn | 14:56 | |
| *** azbiswas has quit IRC | 14:59 | |
| *** carl_baldwin has joined #openstack-neutron-ovn | 15:00 | |
| gongysh_ | hi, does ovn plugin support provider network, or network type such as vlan, vxlan, flat? | 15:01 |
|---|---|---|
| gsagie | gongysh_ : russellb added provider network support | 15:05 |
| russellb | flat and vlan provider networks are supported | 15:05 |
| gongysh_ | russellb, do we have doc about the flow table arrangement in ovn? | 15:06 |
| russellb | yes | 15:06 |
| russellb | http://www.openvswitch.org/support/dist-docs/ovn-architecture.7.html is the best place to start | 15:07 |
| russellb | "Life Cycle of a Packet" talks about each OpenFlow table | 15:07 |
| *** slowe has joined #openstack-neutron-ovn | 15:08 | |
| russellb | and ovn-northd's man page discusses the logical flow tables http://www.openvswitch.org/support/dist-docs/ovn-northd.8.html | 15:08 |
| gongysh_ | that's great. I will have a look. | 15:08 |
| slowe | Morning all, quick question---in the DevStack-based test environment for OVN, is that using OVN ACLs for security groups? | 15:09 |
| russellb | slowe: yes | 15:09 |
| slowe | russellb: Thought so, but wasn't 100% sure. Still using the Neutron L3 agent, though, right? | 15:09 |
| russellb | slowe: it grabs a special branch of ovn that includes a backport of the kernel changes required to make that work. it loads a custom openvswitch kernel module that it builds | 15:09 |
| russellb | yes, it still uses l3 agent for now | 15:09 |
| slowe | russellb: Perfect. That's exactly what I thought, but needed to confirm. | 15:10 |
| russellb | np | 15:10 |
| gongysh_ | russellb, OVN ACLs needs kernel patch, right? | 15:10 |
| russellb | correct | 15:11 |
| russellb | those changes have been accepted into the upstream kernel and are in Linux 4.3 | 15:11 |
| russellb | some distros will backport those changes | 15:11 |
| russellb | otherwise, you'll need to use the openvswitch module maintained out of the linux tree in the ovs git tree, which will have the backport | 15:11 |
| russellb | that backport has not been merged into the ovs git tree yet though | 15:11 |
| *** azbiswas has joined #openstack-neutron-ovn | 15:17 | |
| *** tongli has joined #openstack-neutron-ovn | 15:29 | |
| *** jimchou has joined #openstack-neutron-ovn | 15:29 | |
| *** subscope has quit IRC | 15:33 | |
| *** asuvvari has joined #openstack-neutron-ovn | 15:38 | |
| *** HenryG has quit IRC | 15:40 | |
| *** subscope has joined #openstack-neutron-ovn | 15:41 | |
| *** HenryG has joined #openstack-neutron-ovn | 15:41 | |
| openstackgerrit | Russell Bryant proposed openstack/networking-ovn: Check valid option combos in _get_ovn_port_options https://review.openstack.org/241256 | 15:46 |
| *** slowe has quit IRC | 15:47 | |
| openstackgerrit | Merged openstack/networking-ovn: Pull out ovn config options to helper method https://review.openstack.org/241013 | 15:50 |
| openstackgerrit | Merged openstack/networking-ovn: Add helper method _update_port_in_ovn() https://review.openstack.org/241015 | 15:50 |
| *** yamamoto has joined #openstack-neutron-ovn | 15:51 | |
| *** yamamoto has quit IRC | 15:51 | |
| *** nate_gone is now known as njohnston | 15:52 | |
| *** yamamoto has joined #openstack-neutron-ovn | 15:53 | |
| *** azbiswas has quit IRC | 15:56 | |
| openstackgerrit | Merged openstack/networking-ovn: Remove duplicate call to _get_allowed_mac_addresses_from_port() https://review.openstack.org/241020 | 15:57 |
| *** njohnston is now known as nate_gone | 16:02 | |
| *** jimchou_ has joined #openstack-neutron-ovn | 16:02 | |
| *** jimchou has quit IRC | 16:04 | |
| *** jimchou_ has quit IRC | 16:09 | |
| *** Guest35592 has joined #openstack-neutron-ovn | 16:10 | |
| *** azbiswas has joined #openstack-neutron-ovn | 16:13 | |
| *** asuvvari has quit IRC | 16:21 | |
| *** asuvvari has joined #openstack-neutron-ovn | 16:21 | |
| *** asuvvari has quit IRC | 16:26 | |
| *** asuvvari has joined #openstack-neutron-ovn | 16:27 | |
| *** asuvvari has quit IRC | 16:37 | |
| *** asuvvari has joined #openstack-neutron-ovn | 16:38 | |
| *** asuvvari has quit IRC | 16:42 | |
| *** asuvvari has joined #openstack-neutron-ovn | 16:42 | |
| openstackgerrit | Merged openstack/networking-ovn: Remove unused method _delete_ports() https://review.openstack.org/241024 | 16:53 |
| *** asuvvari has quit IRC | 16:53 | |
| *** asuvvari has joined #openstack-neutron-ovn | 16:54 | |
| *** asuvvari has quit IRC | 16:56 | |
| *** asuvvari has joined #openstack-neutron-ovn | 16:56 | |
| *** jimchou has joined #openstack-neutron-ovn | 16:57 | |
| openstackgerrit | Aaron Rosen proposed openstack/networking-ovn: DONT" REVIEW https://review.openstack.org/241296 | 16:57 |
| *** asuvvari_ has joined #openstack-neutron-ovn | 17:04 | |
| *** shettyg has quit IRC | 17:05 | |
| *** shettyg has joined #openstack-neutron-ovn | 17:05 | |
| mestery | russellb: I think I've got a multi-machine Vagrantfile which sets up a 3-node OVN setup (1 control node, 2 compute) | 17:06 |
| mestery | Once it's working, would that be something interesting to add to networking-ovn for people to use? | 17:06 |
| *** asuvvari has quit IRC | 17:07 | |
| *** subscope has quit IRC | 17:10 | |
| russellb | mestery: absolutely ... technically we have a 2-node vagrantfile setup we already merged | 17:17 |
| *** azbiswas_ has joined #openstack-neutron-ovn | 17:17 | |
| russellb | but i haven't tried it | 17:17 |
| mestery | russellb: Coolio :) | 17:17 |
| russellb | but in general, i do think it makes sense to include that stuff in tree | 17:17 |
| mestery | ++ | 17:17 |
| *** yamamoto has quit IRC | 17:19 | |
| *** azbiswas has quit IRC | 17:19 | |
| *** carl_baldwin has quit IRC | 17:21 | |
| switchcade | mestery: ooh, keep in touch on how that goes. May also be relevant for kernel module testing | 17:27 |
| mestery | switchcade: ++ :) | 17:27 |
| switchcade | particularly on the windows side, since they don't have namespaces or containers yet | 17:28 |
| russellb | i really need to debug this dhcp + ACLs thing | 17:30 |
| russellb | if i can get out from under this pile of email | 17:30 |
| *** nate_gone is now known as njohnston | 17:41 | |
| switchcade | russellb: I'm inclined to think that your quick-fix before is not that far off the real solution. | 17:43 |
| russellb | switchcade: maybe just tighten it up? | 17:43 |
| switchcade | yeah. | 17:44 |
| russellb | i at least need to better wrap my head around exactly what's happening so I can justify it properly :) | 17:44 |
| switchcade | I mean, DHCP isn't a point-to-point connection as such | 17:44 |
| russellb | right, i wonder how neutron handles this today though | 17:44 |
| russellb | i should probably start there | 17:44 |
| switchcade | yeah, I'd like to know that too. | 17:44 |
| switchcade | it may be one of these things which magically works because we have a linux stack in the path | 17:45 |
| russellb | will be in touch :) | 17:45 |
| russellb | maybe *shrug* | 17:45 |
| russellb | in theory we're replicating what we used to do with iptables | 17:45 |
| russellb | so i'll look at exactly what's in there again | 17:45 |
| switchcade | right, but iptables has implicit behaviour, where OVS doesn't | 17:45 |
| switchcade | and looking at the DHCP protocol request/response, it's not immediately apparent how you'd "connection-track" it | 17:46 |
| *** njohnston is now known as nate_gone | 17:50 | |
| *** carl_baldwin has joined #openstack-neutron-ovn | 17:53 | |
| openstackgerrit | Aaron Rosen proposed openstack/networking-ovn: Add etc to .gitignore https://review.openstack.org/241319 | 17:54 |
| *** gangil has joined #openstack-neutron-ovn | 17:57 | |
| *** gangil has joined #openstack-neutron-ovn | 17:57 | |
| russellb | switchcade: if it's just implicit iptables behavior, i wonder how many more cases we'll hit :/ | 18:00 |
| russellb | anyway, i'm just stabbing in the dark, i'll do more proper research before complaining to you :) | 18:00 |
| *** azbiswas_ has quit IRC | 18:03 | |
| mestery | russellb: I've got my multi-node devstack up, VMs splatter across compute nodes and get IP addresses, but I can't ping between them. | 18:03 |
| mestery | And yes, I have a SG setup to allow ping. | 18:03 |
| mestery | I can ping only the VM on the control node. | 18:03 |
| mestery | I'll have time to debug this afer this meeting I'm in is over, just FYI and a heads up | 18:04 |
| russellb | k, i can help debug ... | 18:05 |
| russellb | first thing i'd try is probably updating all of the ports to have no security group applied | 18:05 |
| russellb | to rule that out | 18:05 |
| russellb | i haven't actually tested security groups on a multi-node setup yet | 18:06 |
| *** salv-orl_ has joined #openstack-neutron-ovn | 18:07 | |
| switchcade | russellb: well, we'll no doubt see :-) | 18:09 |
| *** salv-orlando has quit IRC | 18:09 | |
| *** yamamoto has joined #openstack-neutron-ovn | 18:19 | |
| *** yamamoto has quit IRC | 18:24 | |
| Guest35592 | russellb: fwiw the neutron-l2-agent adds rules to allow dhcp automatically https://github.com/openstack/neutron/blob/master/neutron/agent/linux/iptables_firewall.py#L387 | 18:31 |
| Guest35592 | not sure if that's helpful info to know | 18:31 |
| *** Guest35592 is now known as arosen | 18:31 | |
| russellb | arosen: yep, just found that | 18:31 |
| russellb | we'll need to add something similar, but it doesn't explain my problem | 18:32 |
| russellb | default security group allows all outbound IP traffic + related return traffic already | 18:32 |
| russellb | so that would have allowed the client request already | 18:32 |
| arosen | russellb: The initial dhcp-request is sent out with a src.ip of 0.0.0.0 (i guess we are already allowing that)? | 18:34 |
| russellb | default security group rules just allow anything that's IP traffic outbound | 18:34 |
| russellb | i think it's the response getting dropped | 18:35 |
| arosen | russellb: though I think some special handling might need to be done for sending traffic with 0.0.0.0 ip addresses maybe? | 18:41 |
| arosen | since 0.0.0.0 doesn't match the fixed_ip on the port | 18:41 |
| russellb | right, but i don't think we have any flows that would get upset about that | 18:42 |
| switchcade | yeah, default security group still sends things through conntrack right? And if conntrack doesn't like that.. | 18:42 |
| arosen | russellb: k | 18:42 |
| russellb | arosen: or did you mean conntrack just pukes on that? | 18:43 |
| russellb | which i think is what switchcade is saying | 18:43 |
| arosen | russellb: no i was just throwing out ideas. | 18:43 |
| arosen | I know in the nvp plugin we needed to specify this rule here: | 18:43 |
| russellb | ah k :) | 18:43 |
| arosen | https://github.com/openstack/vmware-nsx/blob/master/vmware_nsx/nsxlib/mh/switch.py#L52 | 18:43 |
| switchcade | My reading of that iptables rule is to skip iptables entirely for that traffic | 18:43 |
| arosen | that says allow 0.0.0.0 + mac | 18:43 |
| switchcade | err, skip conntrack. | 18:44 |
| russellb | switchcade: ah | 18:44 |
| russellb | well that's easy enough for the client request | 18:46 |
| *** salv-orl_ has quit IRC | 18:46 | |
| russellb | wonder what's proper for the response then | 18:46 |
| russellb | we could allow everything that looks like a dhcp response too | 18:46 |
| russellb | and then we've basically arrived at the hack i already have, heh | 18:46 |
| arosen | I think it would be better if the neutron plugin adds this rule | 18:47 |
| *** salv-orlando has joined #openstack-neutron-ovn | 18:47 | |
| russellb | agree | 18:47 |
| russellb | but neutron can't add any rules for things to skip conntrack | 18:47 |
| russellb | right now anyway | 18:47 |
| russellb | it can only write rules for things to accept or drop after going through conntrack | 18:48 |
| * russellb has ideas to try .. | 18:51 | |
| openstackgerrit | Merged openstack/networking-ovn: Add etc to .gitignore https://review.openstack.org/241319 | 18:53 |
| mestery | russellb: I'm back now, trying to remove SGs on my multi-node and trying again. | 18:54 |
| russellb | cool | 18:54 |
| mestery | No such luck when removing SGs from two VMs on different hosts. Ping still failing :( | 18:56 |
| russellb | switchcade: so general debugging question, when a packet is marked as invalid, what's the best way to debug why it thought something was invalid? | 18:56 |
| arosen | I get an ip and the security-group stuff is working on single node | 18:57 |
| arosen | so it must be related when the dhcp-request needs to come in over the overlay. | 18:57 |
| russellb | mestery: OK, do all the VMs get addresses via DHCP? | 18:58 |
| russellb | or do you know? | 18:58 |
| mestery | russellb : Indeed they do! Which is what makes this ... odd o_O | 18:58 |
| russellb | ok, interesting | 18:58 |
| russellb | arosen: are you using the ovn branch that our plugin defaults to? | 18:58 |
| russellb | that has a patch to make dhcp bypass conntrack | 18:59 |
| arosen | ah yea | 18:59 |
| arosen | i am | 18:59 |
| mestery | russellb: I don't have a special branch specified anywhere | 18:59 |
| russellb | mestery: sorry that was to arosen | 18:59 |
| mestery | russellb: OK :) | 18:59 |
| russellb | arosen: if you want to look at when it fails, change the ovs tree to remove the top commit (the workaround) | 18:59 |
| russellb | recompile, reinstall, restart ovn-controller | 19:00 |
| russellb | mestery: ok soooooo ... um. | 19:00 |
| mestery | :) | 19:00 |
| russellb | mestery: you're able to ping the ones local to the controller | 19:00 |
| russellb | mestery: right? | 19:00 |
| mestery | russellb: Ack, I can only ping the VM runniung on the controller itself from the DHCP namespace | 19:00 |
| * mestery goes to look at if packets are making it across the tunnel | 19:00 | |
| russellb | mestery: but not elsewhere. and you ssh into a local VM and try to ping from there? | 19:01 |
| mestery | russellb: I was using the console of a VM | 19:01 |
| russellb | ah ok | 19:01 |
| *** asuvvari_ has quit IRC | 19:01 | |
| russellb | tunnels should be working if they all got dhcp responses | 19:01 |
| *** asuvvari has joined #openstack-neutron-ovn | 19:02 | |
| mestery | Yup ... | 19:02 |
| russellb | mestery: i think my next step would be to see what flows are getting hit. | 19:02 |
| mestery | russellb: I'm looking at that now in fact ;) | 19:03 |
| russellb | mestery: a cool trick i got from switchcade was to run something like ... watch -d -n 1 ovs-ofctl -O OpenFlow13 dump-flows br-int | 19:03 |
| russellb | to watch what counters are updating once a second | 19:03 |
| mestery | russellb: Cool! | 19:03 |
| russellb | and hopefully you can see where the ping is going if that's the only traffic you have | 19:03 |
| mestery | russellb: Look here http://paste.openstack.org/show/477888/ | 19:05 |
| mestery | That looks wrong to me | 19:05 |
| mestery | That's a dump from the control node, but the "remote_ip" for both tunnel endpoints points back to itself! | 19:05 |
| russellb | yeah, that's wrong :-) | 19:05 |
| russellb | https://github.com/openstack/networking-ovn/blob/master/devstack/plugin.sh#L253 | 19:06 |
| russellb | is what sets that on each host | 19:06 |
| russellb | should be HOST_IP on each node | 19:06 |
| *** asuvvari has quit IRC | 19:06 | |
| mestery | russellb: OK, let me debug that first | 19:06 |
| russellb | also make sure OVN_UUID is different on each host, too | 19:07 |
| russellb | basically everything is devstack's fault | 19:07 |
| russellb | this watch command seems to be a little better ... $ sudo watch -d -n 1 "ovs-ofctl -O OpenFlow13 dump-flows br-int | cut -f4- -d','" | 19:09 |
| russellb | anyway, something like that | 19:09 |
| mestery | russellb: Good call on that! | 19:10 |
| * russellb feels sorry for ops | 19:10 | |
| mestery | lol | 19:11 |
| mestery | :) | 19:11 |
| russellb | maybe they're all smarter than me | 19:11 |
| *** asuvvari has joined #openstack-neutron-ovn | 19:11 | |
| switchcade | russellb: woops, intermittently in here:) uh there's not a great way to debug INVALID at the moment really:( | 19:16 |
| switchcade | some of the cases are caught and logged to dmesg | 19:17 |
| russellb | switchcade: OK, just making sure I wasn't missing something | 19:17 |
| switchcade | there's a few suggestions in ovs-ofctl I think | 19:17 |
| *** azbiswas has joined #openstack-neutron-ovn | 19:21 | |
| *** asuvvari has quit IRC | 19:21 | |
| russellb | arosen: i mis-spoke, you need to restart ovn-northd after reverting that commit, not ovn-controller | 19:21 |
| *** asuvvari has joined #openstack-neutron-ovn | 19:22 | |
| *** nate_gone is now known as njohnston | 19:29 | |
| *** salv-orlando has quit IRC | 19:30 | |
| mestery | russellb: I re-provisioned my vagrant cluster, and now the tunnel IPs look much more sane. | 19:36 |
| mestery | russellb: Going to spam the system with VMs and see what breaks :) | 19:36 |
| russellb | cool, progress :) | 19:36 |
| mestery | ++ | 19:37 |
| russellb | yo dawg, i heard you liked VMs | 19:38 |
| russellb | so I launched a cloud in a cloud so you could boot some VMs in your VMs | 19:38 |
| mestery | WOOT! | 19:39 |
| mestery | Success! | 19:39 |
| mestery | Yay! | 19:39 |
| * mestery happy dance | 19:39 | |
| russellb | :) | 19:39 |
| mestery | Nice work russellb and team :) | 19:39 |
| russellb | now ... how was it working before, heh | 19:39 |
| mestery | Heh | 19:39 |
| mestery | ;) | 19:39 |
| russellb | let's ignore that | 19:39 |
| mestery | exactly | 19:39 |
| *** jerrygb has quit IRC | 19:39 | |
| *** njohnston is now known as nate_gone | 19:39 | |
| mestery | So, let me push out a patch with the vagrant config now | 19:40 |
| mestery | I think others may find it useful | 19:40 |
| russellb | are you *sure* it used dhcp? and not config drive? | 19:40 |
| mestery | Yeah .... but I have no idea how that worked | 19:40 |
| russellb | how are you going to propose it with the existing vagrantfile | 19:40 |
| * mestery shrug | 19:40 | |
| mestery | Where is the existing Vagrantfile? I looked but didn't see it. Let me look again. | 19:40 |
| russellb | root of the repo i think | 19:40 |
| mestery | Ah, that's why my find wasn't working :P | 19:41 |
| mestery | OK | 19:41 |
| mestery | I see it | 19:41 |
| * mestery looks | 19:41 | |
| russellb | replacing it is fine if what you have is better in some way | 19:41 |
| russellb | if they're different, then we could have vagrant options | 19:41 |
| mestery | Well, I think I can merge the two perhaps | 19:41 |
| russellb | ok | 19:41 |
| mestery | But let me look | 19:41 |
| * russellb offers cake to whoever makes it work on vagrant with libvirt+kvm provider and a centos7 image | 19:42 | |
| russellb | or a fedora image | 19:42 |
| russellb | or box | 19:42 |
| mestery | I like cake | 19:42 |
| russellb | or whatever the heck you call images | 19:42 |
| *** armax has joined #openstack-neutron-ovn | 19:48 | |
| *** armax has quit IRC | 19:49 | |
| *** gangil has quit IRC | 19:55 | |
| *** jerrygb has joined #openstack-neutron-ovn | 19:57 | |
| *** shettyg has left #openstack-neutron-ovn | 20:00 | |
| *** azbiswas has quit IRC | 20:40 | |
| *** azbiswas has joined #openstack-neutron-ovn | 20:40 | |
| *** azbiswas has quit IRC | 20:45 | |
| *** gangil has joined #openstack-neutron-ovn | 20:45 | |
| *** gangil has joined #openstack-neutron-ovn | 20:45 | |
| *** azbiswas has joined #openstack-neutron-ovn | 20:55 | |
| *** jerrygb has quit IRC | 21:16 | |
| *** nate_gone is now known as njohnston | 21:18 | |
| russellb | mestery: and you have a OFFLINE devstack fix too right? | 21:28 |
| mestery | russellb: Working on it, but yes, it shouldn't take too much | 21:28 |
| russellb | oh ok | 21:28 |
| russellb | thought it was just something you needed to push | 21:28 |
| russellb | no rush :) | 21:28 |
| *** njohnston is now known as nate_gone | 21:28 | |
| mestery | Heh, well, I've gotta write it, but it's super simple, I have a call in a few minutes, will push it after that | 21:28 |
| azbiswas | arosen: Regarding https://review.openstack.org/#/c/237820/6/devstack/plugin.sh, if we make OVN_L3_MODE boolean are you saying the check can be removed i.e. a blank boolean value will be treated as default? | 21:36 |
| *** nate_gone is now known as njohnston | 21:45 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 21:58 | |
| *** rtheis has quit IRC | 22:13 | |
| *** tongli has quit IRC | 22:13 | |
| *** jerrygb has joined #openstack-neutron-ovn | 22:17 | |
| *** jerrygb has quit IRC | 22:22 | |
| *** salv-orlando has quit IRC | 22:26 | |
| *** jerrygb has joined #openstack-neutron-ovn | 22:32 | |
| *** azbiswas has quit IRC | 22:37 | |
| *** azbiswas has joined #openstack-neutron-ovn | 22:38 | |
| *** armax has joined #openstack-neutron-ovn | 22:39 | |
| *** azbiswas_ has joined #openstack-neutron-ovn | 22:45 | |
| *** azbiswas has quit IRC | 22:49 | |
| arosen | azbiswas_: i'm saying that we can have it just write the bool value each time | 22:50 |
| arosen | this way the config is always correct | 22:50 |
| *** gangil has quit IRC | 22:53 | |
| *** azbiswas_ is now known as azbiswas | 22:55 | |
| *** gangil has joined #openstack-neutron-ovn | 22:56 | |
| *** gangil has joined #openstack-neutron-ovn | 22:56 | |
| azbiswas | If the bool is not defined then it translate to ovn_l3_mode= | 22:58 |
| azbiswas | i.e. blank | 22:58 |
| azbiswas | will the default value kick-in in that case? | 22:58 |
| arosen | azbiswas: Then define it to false like the other options work | 23:08 |
| arosen | OVN_L3_MODE=${OVN_L3_mode:- False} | 23:09 |
| azbiswas | arosen: Now I get it, thanks | 23:10 |
| *** asuvvari has quit IRC | 23:11 | |
| *** asuvvari has joined #openstack-neutron-ovn | 23:12 | |
| *** jimchou has quit IRC | 23:15 | |
| *** asuvvari has quit IRC | 23:16 | |
| *** regXboi has quit IRC | 23:17 | |
| *** salv-orlando has joined #openstack-neutron-ovn | 23:25 | |
| *** asuvvari has joined #openstack-neutron-ovn | 23:26 | |
| *** gangil has quit IRC | 23:27 | |
| *** gangil has joined #openstack-neutron-ovn | 23:29 | |
| *** gangil has joined #openstack-neutron-ovn | 23:29 | |
| *** asuvvari has quit IRC | 23:31 | |
| *** armax has quit IRC | 23:40 | |
| *** jerrygb has quit IRC | 23:43 | |
| *** salv-orlando has quit IRC | 23:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!