| *** HenryG has quit IRC | 00:04 | |
| *** HenryG has joined #openstack-neutron | 00:04 | |
| *** yfujioka has joined #openstack-neutron | 00:16 | |
| openstackgerrit | Jenkins proposed a change to openstack/neutron: Updated from global requirements https://review.openstack.org/60974 | 00:17 |
|---|---|---|
| *** aymenfrikha has joined #openstack-neutron | 00:19 | |
| *** aymenfrikha1 has quit IRC | 00:21 | |
| openstackgerrit | Aaron Rosen proposed a change to openstack/neutron: Bump api_workers from 0 to 4 https://review.openstack.org/59787 | 00:23 |
| *** aymenfrikha has quit IRC | 00:24 | |
| *** Alienyyg_ has joined #openstack-neutron | 00:36 | |
| *** aymenfrikha has joined #openstack-neutron | 00:36 | |
| *** yamahata_ has joined #openstack-neutron | 00:39 | |
| *** pcm__ has joined #openstack-neutron | 00:45 | |
| *** Alienyyg has quit IRC | 00:45 | |
| *** kashyap has quit IRC | 00:46 | |
| *** pasquier-s has quit IRC | 00:46 | |
| *** kashyap has joined #openstack-neutron | 00:47 | |
| *** pasquier-s has joined #openstack-neutron | 00:47 | |
| *** openstack has joined #openstack-neutron | 00:49 | |
| *** aymenfrikha has quit IRC | 00:52 | |
| openstackgerrit | berlin proposed a change to openstack/neutron: Fix VPNaaS Service driver should check whether router has gateway https://review.openstack.org/60461 | 00:55 |
| *** changbl has quit IRC | 00:57 | |
| *** pcm_ has quit IRC | 00:57 | |
| *** jorisroovers has quit IRC | 01:00 | |
| *** yongli_away is now known as yongli | 01:01 | |
| *** changbl has joined #openstack-neutron | 01:02 | |
| *** SumitNaiksatam has quit IRC | 01:06 | |
| *** djbkd has quit IRC | 01:17 | |
| *** SumitNaiksatam has joined #openstack-neutron | 01:30 | |
| *** Jianyong has joined #openstack-neutron | 01:35 | |
| *** thedodd has quit IRC | 01:37 | |
| *** jianingy_afk is now known as jianingy | 01:41 | |
| jianingy | i try to generate a db migration by alembic but it turns out that the generated script removes many tables | 01:43 |
| *** ljjjustin has joined #openstack-neutron | 01:43 | |
| jianingy | i ran something like, neutron-db-manage --config-file /etc/neutron/neutron.conf --config-file /etc/neutron/plugins/ml2/ml2_conf.ini revision -m "portforwarding" --autogenerate | 01:43 |
| jianingy | i know that i can manually correct the generated script. just wanna if it possible to make it correct by itself | 01:44 |
| *** yamahata_ has quit IRC | 01:51 | |
| *** julim has quit IRC | 01:53 | |
| *** julim has joined #openstack-neutron | 01:56 | |
| *** gdubreui has quit IRC | 02:01 | |
| *** rwsu has quit IRC | 02:12 | |
| *** enikanorov__ has joined #openstack-neutron | 02:14 | |
| *** enikanorov has quit IRC | 02:16 | |
| *** rwsu has joined #openstack-neutron | 02:26 | |
| *** unicell has left #openstack-neutron | 02:28 | |
| *** banix has joined #openstack-neutron | 02:32 | |
| *** marun has quit IRC | 02:32 | |
| *** marun has joined #openstack-neutron | 02:33 | |
| *** gongysh has joined #openstack-neutron | 02:47 | |
| *** vkozhukalov has joined #openstack-neutron | 02:49 | |
| *** alagalah has joined #openstack-neutron | 03:01 | |
| openstackgerrit | dekehn proposed a change to openstack/neutron: extra_dhcp_opt add checks for empty strings https://review.openstack.org/59858 | 03:10 |
| *** yamahata_ has joined #openstack-neutron | 03:36 | |
| *** jgrimm has quit IRC | 03:37 | |
| openstackgerrit | Dazhao Yu proposed a change to openstack/neutron: Calculate stateless IPv6 address https://review.openstack.org/56184 | 03:38 |
| openstackgerrit | Dazhao Yu proposed a change to openstack/neutron: Create a new attribute for subnets, to store v6 dhcp options https://review.openstack.org/52983 | 03:38 |
| *** amotoki has joined #openstack-neutron | 03:39 | |
| *** yamahata_ has quit IRC | 03:40 | |
| *** x86brandon has quit IRC | 03:40 | |
| *** banix has quit IRC | 03:41 | |
| *** chandankumar__ has quit IRC | 03:43 | |
| *** chandankumar has joined #openstack-neutron | 03:52 | |
| *** suresh12 has quit IRC | 03:57 | |
| openstackgerrit | Maru Newby proposed a change to openstack/neutron: Add post-mortem debug option for tests https://review.openstack.org/43776 | 03:57 |
| *** pcm__ has quit IRC | 04:00 | |
| *** aveiga has quit IRC | 04:02 | |
| *** banix has joined #openstack-neutron | 04:05 | |
| *** yamahata_ has joined #openstack-neutron | 04:19 | |
| *** x86brandon has joined #openstack-neutron | 04:21 | |
| enikanorov__ | gongysh: hi | 04:30 |
| enikanorov__ | sorry for delay, I had to sleep :) | 04:31 |
| *** nati_ueno has quit IRC | 04:42 | |
| *** julim has quit IRC | 04:44 | |
| *** yamahata_ has quit IRC | 04:50 | |
| *** banix has quit IRC | 04:51 | |
| *** suresh12 has joined #openstack-neutron | 05:08 | |
| *** yamahata_ has joined #openstack-neutron | 05:11 | |
| *** suresh12 has quit IRC | 05:13 | |
| *** banix has joined #openstack-neutron | 05:20 | |
| *** x86brandon has quit IRC | 05:21 | |
| *** banix has quit IRC | 05:22 | |
| *** yamahata_ has quit IRC | 05:23 | |
| *** yamahata_ has joined #openstack-neutron | 05:23 | |
| *** banix has joined #openstack-neutron | 05:24 | |
| *** banix has quit IRC | 05:26 | |
| *** otherwiseguy has quit IRC | 05:36 | |
| *** x86brandon has joined #openstack-neutron | 05:40 | |
| openstackgerrit | Dazhao Yu proposed a change to openstack/neutron: Create a new attribute for subnets, to store v6 dhcp options https://review.openstack.org/52983 | 05:58 |
| *** suresh12 has joined #openstack-neutron | 05:58 | |
| *** alagalah has quit IRC | 06:02 | |
| *** vkozhukalov has quit IRC | 06:12 | |
| *** jlibosva has joined #openstack-neutron | 06:31 | |
| openstackgerrit | Jenkins proposed a change to openstack/neutron: Imported Translations from Transifex https://review.openstack.org/61021 | 06:33 |
| openstackgerrit | gongysh proposed a change to openstack/neutron: move rpc_setup to the last step of __init__ https://review.openstack.org/60805 | 06:34 |
| *** x86brandon has quit IRC | 06:40 | |
| openstackgerrit | gongysh proposed a change to openstack/neutron: change variable name from plugin into agent https://review.openstack.org/61022 | 06:40 |
| *** alex_klimov has joined #openstack-neutron | 06:58 | |
| *** amritanshu_RnD has joined #openstack-neutron | 06:59 | |
| *** gongysh has quit IRC | 07:11 | |
| *** gongysh has joined #openstack-neutron | 07:20 | |
| *** yfried has joined #openstack-neutron | 07:32 | |
| openstackgerrit | oda-g proposed a change to openstack/neutron: Enable to set dhcp port attributes. https://review.openstack.org/61026 | 07:34 |
| *** SushilKM has joined #openstack-neutron | 07:37 | |
| openstackgerrit | jun xie proposed a change to openstack/neutron: Change from eager load to lazy load https://review.openstack.org/61027 | 07:40 |
| *** vkozhukalov has joined #openstack-neutron | 07:52 | |
| *** SushilKM has quit IRC | 07:53 | |
| *** dcahill1 has quit IRC | 07:56 | |
| *** nati_ueno has joined #openstack-neutron | 07:56 | |
| *** yfried has quit IRC | 08:02 | |
| openstackgerrit | Oleg Bondarev proposed a change to openstack/neutron: LBaaS: unify haproxy-on-host plugin driver and agent https://review.openstack.org/40381 | 08:09 |
| *** yfujioka has quit IRC | 08:11 | |
| gongysh | salv-orlando: ping | 08:12 |
| enikanorov__ | salv-orlando: ping | 08:14 |
| enikanorov__ | gongysh: hi | 08:14 |
| *** dcahill has joined #openstack-neutron | 08:15 | |
| gongysh | <enikanorov__> are u enikanorov? | 08:15 |
| enikanorov__ | yes, it's me. i'm usually sitting from 2 places | 08:15 |
| enikanorov__ | and changin them from time to time | 08:15 |
| enikanorov__ | (sorry for inconvenience) | 08:15 |
| *** suresh12 has quit IRC | 08:17 | |
| openstackgerrit | Dazhao Yu proposed a change to openstack/python-neutronclient: Add code for support dhcp mode in neutronClient https://review.openstack.org/54250 | 08:18 |
| *** ljjjustin has quit IRC | 08:26 | |
| openstackgerrit | jun xie proposed a change to openstack/neutron: Change from eager load to lazy load https://review.openstack.org/61027 | 08:28 |
| *** ihrachyska has quit IRC | 08:31 | |
| openstackgerrit | Dazhao Yu proposed a change to openstack/neutron: Calculate stateless IPv6 address https://review.openstack.org/56184 | 08:34 |
| *** rossella_s has joined #openstack-neutron | 08:34 | |
| *** rossella_s has quit IRC | 08:39 | |
| openstackgerrit | oda-g proposed a change to openstack/neutron: Enable to set dhcp port attributes https://review.openstack.org/61026 | 08:40 |
| *** ygbo has joined #openstack-neutron | 08:43 | |
| *** fouxm has joined #openstack-neutron | 08:44 | |
| *** ihrachyska has joined #openstack-neutron | 08:46 | |
| openstackgerrit | Oleg Bondarev proposed a change to openstack/neutron: LBaaS: agent monitoring and instance rescheduling https://review.openstack.org/59743 | 08:47 |
| *** jistr has joined #openstack-neutron | 09:05 | |
| *** ihrachyska has quit IRC | 09:07 | |
| openstackgerrit | Dazhao Yu proposed a change to openstack/python-neutronclient: Add code for support dhcp mode in neutronClient https://review.openstack.org/54250 | 09:09 |
| *** safchain has joined #openstack-neutron | 09:09 | |
| *** jpich has joined #openstack-neutron | 09:10 | |
| *** suresh12 has joined #openstack-neutron | 09:15 | |
| *** ihrachyska has joined #openstack-neutron | 09:19 | |
| salv-orlando | enikanorov__ and gongysh: good morning, anything I can do for you? | 09:19 |
| *** suresh12 has quit IRC | 09:19 | |
| gongysh | salv-orlando: I forgot why I ping u, but could u please have a look at this patch: https://review.openstack.org/#/c/61022/ | 09:25 |
| gongysh | salv-orlando: and this https://review.openstack.org/#/c/60756/ | 09:25 |
| gongysh | salv-orlando: sorry for that. :) | 09:26 |
| salv-orlando | np. | 09:26 |
| *** netavenger-jr has joined #openstack-neutron | 09:27 | |
| openstackgerrit | berlin proposed a change to openstack/neutron: Fix VPNaaS Service driver should check whether router has gateway https://review.openstack.org/60461 | 09:29 |
| enikanorov__ | salv-orlando: and I wanted to get your opinion on the question about data model. I'm currently working on adding loadbalancer instance to the lbaas, it includes the change: provider is specified for the loadbalancer, not for the pool. | 09:30 |
| enikanorov__ | Provider attribute is used for dispatching rest calls to the driver. Currently in order to process calls to work with members/healthmonitors, we need to fetch the parent pool and use it's provider attribute. | 09:31 |
| salv-orlando | is this again about moving the providers in the database? | 09:31 |
| enikanorov__ | with loadbalancer this will require additional step: fetch pool -> fetch it's loadbalancer and use it's provider | 09:31 |
| enikanorov__ | salv-orlando: no | 09:32 |
| salv-orlando | ah ok sorry | 09:32 |
| salv-orlando | hearing then | 09:32 |
| * salv-orlando to be more precise reading | 09:32 | |
| enikanorov__ | So I'm adding loadbalancer_id to members and healthmonitors | 09:32 |
| enikanorov__ | to save one step | 09:32 |
| enikanorov__ | while the attribute itself is deductible from another attributes | 09:32 |
| enikanorov__ | such as member belongs to a pool, and pool has loadbalancer_id | 09:33 |
| enikanorov__ | so member gets it loadbalancer_id | 09:33 |
| enikanorov__ | so in other words i'm adding data duplication to the model | 09:33 |
| enikanorov__ | to avoid additional queries | 09:33 |
| enikanorov__ | fix: loadbalancer_id to health monitor association, not healthmonitor model itself | 09:34 |
| enikanorov__ | so the question is whether it is acecptable | 09:34 |
| salv-orlando | in principle I'd say that denormalization is not good, but I need to look better at this specific use case before having an opinion | 09:34 |
| salv-orlando | Is the goal to avoid both extra db queries and rest API calls from clients, right? | 09:35 |
| salv-orlando | ? | 09:35 |
| openstackgerrit | Isaku Yamahata proposed a change to openstack/neutron: dhcp_agent: ignore notification during syncing https://review.openstack.org/61057 | 09:35 |
| enikanorov__ | salv-orlando: right | 09:35 |
| enikanorov__ | extra queries during call proccessing | 09:35 |
| salv-orlando | sorry about the double question mark, my fingers slipped on the keyboard. I guess the coffee has not yet kicked in. | 09:35 |
| salv-orlando | the loadbalancer_id attribute, is that returned to the client as well? | 09:36 |
| enikanorov__ | yes. but for members and health monitor associations it can't be modified by user | 09:38 |
| enikanorov__ | so currently we use the follwoing method to dispatch the call to the driver: https://github.com/openstack/neutron/blob/master/neutron/services/loadbalancer/plugin.py#L93 | 09:38 |
| jianingy | hello, I try to follow https://wiki.openstack.org/wiki/Neutron/DatabaseMigration to create a db migration. but it turns out there are many drop tables in upgrade function. my command was neutron-db-manage --config-file /etc/neutron/neutron.conf --config-file /etc/neutron/plugins/ml2/ml2_conf.ini revision -m "portforwarding" --autogenerate | 09:38 |
| jianingy | need someone to shed some light :) | 09:39 |
| *** dcahill has quit IRC | 09:39 | |
| salv-orlando | enikanorov__: vip and pool can instead be switched to a different load balancer instance? | 09:40 |
| enikanorov__ | jianingy: seems to me that this way of generatin migrations had an issue of swapping upgrade and downgrade. and also I would not rely on autogeneration as it has another issues | 09:40 |
| jianingy | ok. got it. I will do some manually change on the script | 09:41 |
| enikanorov__ | salv-orlando: theoretically, but i don't think it worth adding such ability. It creates more problems then solves | 09:41 |
| salv-orlando | indeed | 09:41 |
| *** dcahill has joined #openstack-neutron | 09:41 | |
| *** markmcclain has joined #openstack-neutron | 09:42 | |
| enikanorov__ | oh, i god the idea how to avoid the duplication of data in other models. Looks like I could leave 'provider' attribute in the pool model, but not as a db field, but as a relationship that will fetch provider from parent loadbalancer | 09:42 |
| jianingy | markmcclain: hi, mark. would you please help have a look at https://review.openstack.org/#/c/60195/. thank you ;) | 09:42 |
| enikanorov__ | *i got | 09:43 |
| salv-orlando | Seems like the right approach to me; this is why I asked if the field could be edited too. | 09:43 |
| salv-orlando | Perhaps you should be all right using joined loads, what do you reckon? | 09:44 |
| enikanorov__ | salv-orlando: yeah. that should do | 09:44 |
| enikanorov__ | salv-orlando: thanks! | 09:45 |
| salv-orlando | np | 09:45 |
| markmcclain | jianingy: approved | 09:46 |
| jianingy | markmcclain: thank you very much | 09:46 |
| *** wenjianhn has joined #openstack-neutron | 09:54 | |
| *** netavenger-jr is now known as networkstatic | 09:55 | |
| *** jlibosva has quit IRC | 09:55 | |
| *** networkstatic has quit IRC | 09:56 | |
| *** networkstatic has joined #openstack-neutron | 09:56 | |
| *** jlibosva has joined #openstack-neutron | 09:58 | |
| *** chandankumar has quit IRC | 10:00 | |
| kashyap | Hello, any hints on why isn't the "--router:external=True" option listed under `neutron help net-create`? | 10:03 |
| *** networkstatic has quit IRC | 10:05 | |
| kashyap | From here - https://github.com/openstack/python-neutronclient/blob/master/neutronclient/neutron/v2_0/network.py#L95 | 10:06 |
| kashyap | markmcclain, Hello, when you get a sec, maybe you have a hint here? | 10:07 |
| *** chandankumar has joined #openstack-neutron | 10:07 | |
| markmcclain | kashyap: it's an extension | 10:08 |
| markmcclain | I don't think any of our help shows extension attribute suggestions | 10:09 |
| kashyap | markmcclain, I see, how does a user/admin discover them? | 10:09 |
| openstackgerrit | Dazhao Yu proposed a change to openstack/neutron: Calculate stateless IPv6 address https://review.openstack.org/56184 | 10:10 |
| *** jorisroovers has joined #openstack-neutron | 10:13 | |
| kashyap | No worries, I found the documentatoin here -- http://docs.openstack.org/admin-guide-cloud/content/ch_networking.html#section_networking-adv-features | 10:14 |
| marun | salv-orlando: ping | 10:14 |
| salv-orlando | hi marun | 10:16 |
| marun | salv-orlando: Profiling eventlet sucks :( | 10:18 |
| marun | salv-orlando: I'm going to try an eventlet-specific profiler, maybe I'll have more luck: https://github.com/colinhowe/eventlet_profiler | 10:18 |
| openstackgerrit | A change was merged to openstack/neutron: change variable name from plugin into agent https://review.openstack.org/61022 | 10:19 |
| marun | salv-orlando: though the short-term emphasis should remain on making the best of a bad situation, the lack of explicit coordination with nova would appear our biggest hole | 10:20 |
| marun | salv-orlando: we're at the mercy of timing issues so long as vm boot is not directly coordinated port/network configuration (l2, dhcp, and l3) | 10:21 |
| salv-orlando | by explicit coordination you mean synchronising nova events with neutron to make a vm active only when the network is up and running? | 10:21 |
| marun | salv-orlando: correct. i'm starting to think that a vm should not be booted unless networking resources have been preallocated | 10:21 |
| salv-orlando | meaning at least that the port has been wired by the l3 agent, and the dhcp entry added into the hostsfile | 10:22 |
| marun | salv-orlando: correct. and we have to have some assurance that the l2 agent is going to respond fast enough to vif plug | 10:22 |
| openstackgerrit | Dazhao Yu proposed a change to openstack/neutron: Calculate stateless IPv6 address https://review.openstack.org/56184 | 10:22 |
| marun | salv-orlando: is there any allowance for this kind of arrangement in nova at present? | 10:22 |
| salv-orlando | marun: I think I brought down that time a lot with the experimental patch I have | 10:22 |
| salv-orlando | nope, but it's been talked about | 10:22 |
| marun | salv-orlando: which patch? | 10:23 |
| salv-orlando | The first thing to do is to have a way in neutron to tell other services, or directly to users, whether a port is really really working | 10:23 |
| salv-orlando | 57420 | 10:23 |
| salv-orlando | but it is a mash-up of several patches at the moment | 10:23 |
| salv-orlando | I am working on extracting 3/4 patches out of it targeting distinct problems | 10:24 |
| *** gongysh has quit IRC | 10:24 | |
| salv-orlando | I've entered a bug on this matter; basically I was observing the main loop in the agent taking even 600 seconds to wire a non-impressive number of interfaces | 10:24 |
| marun | that's pretty crazy | 10:25 |
| salv-orlando | Now I brought down these figures to 60 seconds in the worst case. | 10:25 |
| sgran | that's still a while, but it's definitely better :) | 10:25 |
| salv-orlando | Trying to do better than that on the gate is hard however, because ovs-vsctl goes crazy | 10:25 |
| salv-orlando | and this is because the dhcpagent and the l3 agent are on the same host too, in my opinion | 10:25 |
| marun | salv-orlando: i've seen ovs bottlenecking too | 10:26 |
| salv-orlando | if one would run gate tests on a 2 or 3 hosts setup, I think it will be much faster, but this is just a guess at the moment | 10:26 |
| marun | salv-orlando: if i boot even 40 VM's, the ovs commands start stacking up | 10:26 |
| *** jp_at_hp has joined #openstack-neutron | 10:26 | |
| salv-orlando | marun: Yeah, I wanted to give a shot to ovs2.0 which has multi-threading in vswitchd but never found time to do that | 10:27 |
| marun | salv-orlando: well, I'm happy to try to work through reviewing this stuff with you | 10:28 |
| salv-orlando | The other things would be to run less ovs commands; we might be able to spare some 'get' operations, but they're seldom slow; the slow ones are the ones for setting the tag on the port and for configuring the flows | 10:28 |
| salv-orlando | yeah I will start pushing patches today | 10:28 |
| marun | salv-orlando: I see you're using all kinds of debug logging | 10:29 |
| marun | salv-orlando: it gets expensive quick, we'll probably want to figure out how to instrument things generically and then profile specific paths as they are seen to be bottlenecks | 10:29 |
| *** Jianyong has quit IRC | 10:30 | |
| marun | salv-orlando: we definitely need more data as to what is going on | 10:30 |
| *** markmcclain has quit IRC | 10:30 | |
| openstackgerrit | Dazhao Yu proposed a change to openstack/python-neutronclient: Add code for support dhcp mode in neutronClient https://review.openstack.org/54250 | 10:30 |
| *** jlibosva1 has joined #openstack-neutron | 10:30 | |
| marun | salv-orlando: have you considered other ways of collecting data? say, logging on rpc dequeue/notification processing, wsgi call, etc. the rally project looks interesting, too | 10:31 |
| salv-orlando | marun: I have been going down the path of the poor man's profiling so far, because we needed to act quick. | 10:32 |
| salv-orlando | But I too was hoping to add some serious profiling for the next release, so that at least we would able to understand immediately what's going on when things go wrong | 10:32 |
| *** jlibosva has quit IRC | 10:33 | |
| *** markmcclain has joined #openstack-neutron | 10:33 | |
| marun | salv-orlando: at least when we start seeing things fall over we can quickly determine why | 10:33 |
| marun | salv-orlando: at present the ad-hoc nature makes it difficult for a non-specialist to participate | 10:33 |
| marun | salv-orlando: ^ presuming we get serious profiling in | 10:33 |
| *** wenjianhn has quit IRC | 10:34 | |
| marun | salv-orlando: anyway, I'll let you know if the eventlet-specific profiling proves useful. | 10:34 |
| salv-orlando | marun: right. Even with the crappy instrumentation I added for doing this work, It still take me about 10 minutes for each failure to grep data I need from the logs, pinpoint where the problem is located, and then finally look at the details | 10:34 |
| marun | salv-orlando: ugh | 10:35 |
| marun | salv-orlando: less than ideal | 10:35 |
| marun | salv-orlando: another potential option: http://codeascraft.com/2011/02/15/measure-anything-measure-everything/ | 10:35 |
| salv-orlando | marun: factor in those 10 minutes the fact that I'm dumber than average ;) | 10:35 |
| marun | salv-orlando: I'm hoping we can learn from others example, at least. We're not exactly breaking new ground. | 10:35 |
| salv-orlando | right. I totally do not want to re-invent the wheel | 10:36 |
| marun | salv-orlando: nah, debugging with logs sucks :) | 10:36 |
| salv-orlando | Another thing that baffles me is that we've been totally carelessness when it comes to manage this operational status | 10:37 |
| marun | operational status? | 10:37 |
| openstackgerrit | Sylvain Afchain proposed a change to openstack/neutron: Dnsmasq uses all agent IPs as nameservers https://review.openstack.org/61067 | 10:38 |
| salv-orlando | we put a port to ACTIVE when it gets wired on the agent for instance, but what about the DHCP state? can a port be considered "active" even if it won't be able to receive the ip addresses configured on it? | 10:38 |
| salv-orlando | marun: also, as another example, why don't floating IPs have a state as well? | 10:39 |
| salv-orlando | so far clients assume they're ready to use as soon as they're created, but this is not true at all. | 10:39 |
| marun | salv-orlando: ah, gotcha. and yes. | 10:40 |
| marun | salv-orlando: completely agree. | 10:40 |
| marun | salv-orlando: at present we're not really keeping track of the true status of ports, and not communicating the status to nova. | 10:41 |
| salv-orlando | marun: I | 10:41 |
| salv-orlando | would say we are optimistically assuming everything will magically work fine in an eventually consistent way | 10:41 |
| marun | salv-orlando: magically being the operative word! | 10:41 |
| salv-orlando | which really depends on what you mean as "eventually". | 10:42 |
| salv-orlando | marun: so say if Neutron was working at any italian public office, its "eventual consistency" would be totally acceptable | 10:42 |
| marun | hah | 10:42 |
| marun | salv-orlando: so 'eventual' in the sense of 'occassionaly never'? ;) | 10:43 |
| *** gdubreui has joined #openstack-neutron | 10:43 | |
| salv-orlando | yes, if you get into an issue I've seen with the dhcp agent. | 10:43 |
| marun | (or 'often', as the case may be) | 10:44 |
| salv-orlando | basically port_update might interfere with sync_state and the final result being your port being added and then removed from the hosts file | 10:44 |
| salv-orlando | and your vm never getting an address | 10:44 |
| marun | salv-orlando: yeah, or the port not being added to the hosts file in time. | 10:45 |
| marun | salv-orlando: or the port not being wired in time for the vm to get a dhcp address | 10:45 |
| marun | salv-orlando: *sigh* | 10:45 |
| *** pbeskow has quit IRC | 10:45 | |
| marun | salv-orlando: wrt the sync state problem, have you followed the ml discussion at all? | 10:45 |
| salv-orlando | yes I have but it has deviated on a RPC discussion which I did not found very relevent | 10:46 |
| marun | salv-orlando: What do you think of my suggestion of setting a timestamp on sync state and discarding rpc notifications previous to that timestamp? | 10:46 |
| openstackgerrit | A change was merged to openstack/neutron: Imported Translations from Transifex https://review.openstack.org/61021 | 10:47 |
| marun | salv-orlando: isn't rpc important, though? since port_update and sync_state are mutually exclusive (as per the file lock), the only problem I can see is processing stale notifications. | 10:48 |
| salv-orlando | That might work, but I'm not sure it would be 100% safe. The problemI think is that sync_state reads your network from the cache, and then starts doing its job. Then port_update arrives in, preempts sync_state, and configures a port | 10:48 |
| marun | salv-orlando: what am I missing? | 10:48 |
| marun | salv-orlando: there is no pre-emption, though | 10:48 |
| marun | salv-orlando: there is a file lock | 10:48 |
| marun | (as per armax's recent patches) | 10:49 |
| salv-orlando | marun: indeed the corollary is that armax patches fixed it I think | 10:49 |
| salv-orlando | at least after merging his patches in the code I'm running I've not been seeing this problem anymore. | 10:49 |
| *** yamahata_ has quit IRC | 10:49 | |
| marun | salv-orlando: so assuming no pre-emption, ensuring that stale rpc messages are discarded would seem the next bug to squash | 10:49 |
| *** cyeoh has quit IRC | 10:50 | |
| salv-orlando | marun: Agreed. But I have just one more little thing that it's not yet very clear. | 10:51 |
| salv-orlando | What makes these notifications stale? | 10:51 |
| marun | salv-orlando: service disruption | 10:51 |
| marun | salv-orlando: sorry, that's not clear | 10:51 |
| salv-orlando | marun: nope that was clear | 10:52 |
| marun | salv-orlando: I'm a bit fuzzy on this, to be honest. | 10:52 |
| salv-orlando | I was just to make sure we're on the same page - so service disruption, which might always happen, triggers sync_state | 10:53 |
| salv-orlando | and sync_state then makes the notifications (or at least part of them) stale | 10:53 |
| marun | salv-orlando: right. and the order in which sync_state is called wrt notifications could cause problems | 10:53 |
| marun | salv-orlando: ideally sync_state would always be called first on agent startup and cause all previous notifications to be ignored. | 10:53 |
| salv-orlando | that's where my preemption issue started (it was really about the cache, not the locked routine) | 10:54 |
| salv-orlando | marun: I am growing to think that perhaps processing notification reactively while it looks good for some reasons is actually bad for others - see the problems we're discussing here | 10:54 |
| marun | salv-orlando: we also need to have the service announce when it has started so agents can know to sync themselves and, again, ensure that sync_state causes previous notifications to be ignored. | 10:54 |
| salv-orlando | this was causing mayhem in the l3 agent, and we now do things differently there, and I quite like it | 10:55 |
| *** bvandenh has joined #openstack-neutron | 10:55 | |
| marun | salv-orlando: what was the nature of the l3 change? | 10:55 |
| marun | salv-orlando: I'm not sure I understand how we should be processing notifications (other than reactively) | 10:56 |
| salv-orlando | marun: Basically a notification is interpreted as something that sets a trigger that says "look something has changed, and this is what changed exactly" | 10:57 |
| salv-orlando | then you do nothing until the "main processing loop" starts | 10:57 |
| *** pbeskow has joined #openstack-neutron | 10:57 | |
| salv-orlando | this will operate on all the notifications which were received and process them | 10:58 |
| salv-orlando | the benefits are that you can have a lot more control over concurrency, avoid interferences such as the ones which were causing processing of stale notification, and also coalesce repeated notifications | 10:58 |
| marun | salv-orlando: ah, so basically we take control of the event loop in the agent instead of just creating methods to handle events | 10:59 |
| marun | +1 | 10:59 |
| salv-orlando | right, that's the idea. this adds some lag in request processing | 10:59 |
| salv-orlando | the lag is negligible in low-workload conditions | 10:59 |
| marun | salv-orlando: that would allow us to choose how to coordinate the sync state with the notification processing, so I think the lag is a small price to pay | 10:59 |
| salv-orlando | marun: I totally agree | 10:59 |
| *** Jianyong has joined #openstack-neutron | 11:00 | |
| *** Jianyong has left #openstack-neutron | 11:00 | |
| marun | salv-orlando: what do you think the timeline should be on implementing this? And will it be straightforward enough to backport? | 11:00 |
| salv-orlando | for the ovs agent I think it's even back portable; same should apply for the dhcp agent. I think we can easily shoot for a I-2 timeline | 11:01 |
| marun | salv-orlando: ok, cool. I'm definitely up for helping with that. Are you focused on moving either the ovs or l2 agent in that direction? | 11:02 |
| *** markmcclain has quit IRC | 11:05 | |
| salv-orlando | marun: I am already doing that for the ovs agent. But I would really appreciate if somebody else took care of the dhcp agent. Are you up for that? | 11:05 |
| marun | salv-orlando: yes | 11:06 |
| marun | salv-orlando: is there a bug/blueprint at present? | 11:06 |
| salv-orlando | not for the dhcp agent not yet | 11:06 |
| salv-orlando | I have a bug for the ovs agent, not a blueprint. since i am not really doing major changes nor adding new features, I think a bug is good enough | 11:06 |
| marun | salv-orlando: do you think that will be true of the dhcp agent changes as well? | 11:07 |
| salv-orlando | marun: I am not sure of that, you should probably do an assessment first of what's needed, but honestly I really do not care a lot of whether its a bug or a blueprint | 11:08 |
| marun | salv-orlando: I'd appreciate a pointer to the bug in question, in any case, so I can derive from that. | 11:08 |
| salv-orlando | marun: this is the bug I created: https://bugs.launchpad.net/neutron/+bug/1253993 | 11:08 |
| marun | salv-orlando: ok, understood. | 11:08 |
| marun | salv-orlando: ok, thank you. | 11:09 |
| marun | salv-orlando: so, tangential question. I haven't yet gotten around to posting a patch to send agent notifications even when an agent is reported to be down. | 11:10 |
| marun | salv-orlando: if an agent is not available for a dhcp notification, should that report an exception? | 11:10 |
| marun | salv-orlando: and for that matter, if an agent cannot be scheduled for a network, should that be reported as an exception? | 11:10 |
| marun | salv-orlando: I'm assuming the answer should be 'yes' in both cases, but I don't know that the calling code is going to know how to handle things. | 11:11 |
| salv-orlando | The agent is processed synchronously, is that correct? | 11:11 |
| marun | salv-orlando: it's not like we have any capability to rollback port creation if we can't fully configure things. | 11:11 |
| marun | salv-orlando: in both cases (scheduling and sending of notification), it is synchronous, yes. | 11:11 |
| salv-orlando | the we should fail it. | 11:12 |
| marun | salv-orlando: ok. | 11:12 |
| salv-orlando | I think it's tantamount to the case where you have not enough memory for a vm | 11:12 |
| salv-orlando | only that in nova this is processed asynchronously from the scheduler | 11:12 |
| marun | salv-orlando: right. it changes the failure mode from silent to noisy | 11:13 |
| marun | salv-orlando: The crappy thing, though, is that just because the notification couldn't be sent doesn't mean the port won't be configured (via sync_state). | 11:13 |
| marun | salv-orlando: but as you say, fail now, ask questions later. | 11:13 |
| *** jorisroovers has quit IRC | 11:14 | |
| *** gdubreui has quit IRC | 11:17 | |
| *** heyongli has joined #openstack-neutron | 11:29 | |
| *** Sreedhar has joined #openstack-neutron | 11:33 | |
| *** pcm_ has joined #openstack-neutron | 11:37 | |
| *** pcm_ has quit IRC | 11:37 | |
| *** alagalah has joined #openstack-neutron | 11:37 | |
| amotoki | marun: could you take a look at https://review.openstack.org/#/c/60786/? | 11:38 |
| *** pcm_ has joined #openstack-neutron | 11:38 | |
| marun | amotoki: can do | 11:38 |
| amotoki | marun: request-id middlware patch on oslo. I would like to know your opinion whether the middleware should be split. | 11:38 |
| marun | amotoki: it's a bit pedantic | 11:39 |
| openstackgerrit | Marios Andreou proposed a change to openstack/neutron: Validate CIDR given as ip-prefix in security-group-rule-create https://review.openstack.org/59212 | 11:39 |
| marun | amotoki: But I guess I would agree with them | 11:40 |
| amotoki | marun: request-id middleware without no error handling? | 11:41 |
| *** alagalah has left #openstack-neutron | 11:42 | |
| marun | amotoki: I would submit 2 patches, once for each middleware (error and request id) | 11:42 |
| amotoki | marun: makes sense. I will update. thanks for the suggestion. | 11:43 |
| marun | amotoki: np :) | 11:43 |
| *** jorisroovers has joined #openstack-neutron | 11:46 | |
| *** SushilKM has joined #openstack-neutron | 11:47 | |
| openstackgerrit | Svetlana Dobogoeva proposed a change to openstack/neutron: Added unit tests for module neutron/plugins/nicira/api_client/client.py https://review.openstack.org/59948 | 11:50 |
| *** amotoki has quit IRC | 11:53 | |
| *** SushilKM has quit IRC | 11:54 | |
| *** HenryG has quit IRC | 11:55 | |
| *** pcm_ has quit IRC | 12:04 | |
| openstackgerrit | A change was merged to openstack/neutron: Fix str2dict and dict2str's incorrect behavior https://review.openstack.org/60195 | 12:06 |
| *** cyeoh has joined #openstack-neutron | 12:27 | |
| *** alex_klimov has quit IRC | 12:33 | |
| *** jianingy is now known as jianingy_afk | 12:35 | |
| *** cyeoh has quit IRC | 12:38 | |
| *** simon-AS559 has joined #openstack-neutron | 12:44 | |
| openstackgerrit | Wu Wenxiang proposed a change to openstack/neutron: Remove start index 0 in range() https://review.openstack.org/61100 | 12:44 |
| *** netavenger-jr has joined #openstack-neutron | 12:48 | |
| openstackgerrit | Wu Wenxiang proposed a change to openstack/python-neutronclient: Remove start index 0 in range() https://review.openstack.org/61102 | 12:49 |
| openstackgerrit | Oleg Bondarev proposed a change to openstack/neutron: Delete disassociated floating ips on external network deletion https://review.openstack.org/53364 | 12:54 |
| *** SushilKM has joined #openstack-neutron | 12:57 | |
| *** netavenger-jr has quit IRC | 12:59 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Make timeout for ovs-vsctl configurable https://review.openstack.org/61105 | 13:00 |
| *** alagalah has joined #openstack-neutron | 13:00 | |
| *** Jianyong has joined #openstack-neutron | 13:00 | |
| *** SushilKM has quit IRC | 13:04 | |
| *** SushilKM has joined #openstack-neutron | 13:05 | |
| *** SushilKM has quit IRC | 13:09 | |
| *** alagalah has left #openstack-neutron | 13:10 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Make timeout for ovs-vsctl configurable https://review.openstack.org/61105 | 13:13 |
| *** pcm_ has joined #openstack-neutron | 13:16 | |
| openstackgerrit | enikanorov proposed a change to openstack/neutron: Introduce Loadbalancer instance https://review.openstack.org/60207 | 13:19 |
| *** salv-orlando has quit IRC | 13:23 | |
| *** HenryG has joined #openstack-neutron | 13:29 | |
| *** openstackgerrit has quit IRC | 13:37 | |
| *** openstackgerrit has joined #openstack-neutron | 13:37 | |
| *** dims has quit IRC | 13:39 | |
| *** jianingy_afk is now known as jianingy | 13:40 | |
| *** armax has joined #openstack-neutron | 13:41 | |
| *** dims has joined #openstack-neutron | 13:41 | |
| *** jianingy is now known as jianingy_afk | 13:44 | |
| *** irenab__ has joined #openstack-neutron | 13:59 | |
| *** JoeHazzers has joined #openstack-neutron | 13:59 | |
| JoeHazzers | is it possible to somehow restrict or associate an IP range with a (group of) nodes? | 14:01 |
| openstackgerrit | Wu Wenxiang proposed a change to openstack/neutron: Remove start index 0 in range() https://review.openstack.org/61100 | 14:05 |
| *** jp_at_hp has quit IRC | 14:06 | |
| *** jp_at_hp has joined #openstack-neutron | 14:07 | |
| *** SushilKM has joined #openstack-neutron | 14:07 | |
| *** amritanshu_RnD has quit IRC | 14:09 | |
| *** aymenfrikha has joined #openstack-neutron | 14:11 | |
| *** dsockwell has quit IRC | 14:11 | |
| *** aveiga has joined #openstack-neutron | 14:11 | |
| *** litong has joined #openstack-neutron | 14:13 | |
| *** dsockwell has joined #openstack-neutron | 14:17 | |
| openstackgerrit | Marios Andreou proposed a change to openstack/neutron: Validate CIDR given as ip-prefix in security-group-rule-create https://review.openstack.org/59212 | 14:21 |
| *** amritanshu_RnD has joined #openstack-neutron | 14:23 | |
| *** julim has joined #openstack-neutron | 14:28 | |
| *** banix has joined #openstack-neutron | 14:28 | |
| openstackgerrit | A change was merged to openstack/neutron: LBaaS: unify haproxy-on-host plugin driver and agent https://review.openstack.org/40381 | 14:34 |
| *** salv-orlando has joined #openstack-neutron | 14:39 | |
| *** salv-orlando_ has joined #openstack-neutron | 14:44 | |
| *** beagles is now known as beagles_brb | 14:45 | |
| *** salv-orlando has quit IRC | 14:46 | |
| *** salv-orlando_ is now known as salv-orlando | 14:46 | |
| *** clev has joined #openstack-neutron | 14:46 | |
| *** jecarey has joined #openstack-neutron | 14:48 | |
| *** peristeri has joined #openstack-neutron | 14:52 | |
| *** rkukura has quit IRC | 14:56 | |
| *** salv-orlando has quit IRC | 14:56 | |
| *** armax has quit IRC | 15:00 | |
| *** irenab__ has quit IRC | 15:04 | |
| *** julim has quit IRC | 15:05 | |
| openstackgerrit | Sylvain Afchain proposed a change to openstack/neutron: Dnsmasq uses all agent IPs as nameservers https://review.openstack.org/61067 | 15:06 |
| openstackgerrit | enikanorov proposed a change to openstack/neutron: Introduce Loadbalancer instance https://review.openstack.org/60207 | 15:07 |
| *** julim has joined #openstack-neutron | 15:07 | |
| *** jgrimm has joined #openstack-neutron | 15:12 | |
| *** reaper has joined #openstack-neutron | 15:12 | |
| *** reaper has quit IRC | 15:12 | |
| *** heyongli has quit IRC | 15:13 | |
| *** rkukura has joined #openstack-neutron | 15:18 | |
| *** ywu has joined #openstack-neutron | 15:21 | |
| *** jgrimm has quit IRC | 15:22 | |
| *** beagles_brb is now known as beagles | 15:23 | |
| *** otherwiseguy has joined #openstack-neutron | 15:23 | |
| *** Jianyong has quit IRC | 15:24 | |
| *** jgrimm has joined #openstack-neutron | 15:35 | |
| *** amritanshu_RnD has quit IRC | 15:38 | |
| *** salv-orlando has joined #openstack-neutron | 15:40 | |
| marios | obondarev: enikanorov__ : safchain : SumitNaiksatam: thanks for the review and comments - latest version is up if you have a minute: https://review.openstack.org/#/c/59212/ thanks! | 15:46 |
| *** salv-orlando has quit IRC | 15:48 | |
| *** nati_ueno has quit IRC | 15:48 | |
| *** changbl has quit IRC | 15:54 | |
| *** changbl has joined #openstack-neutron | 15:55 | |
| SumitNaiksatam | marios: thanks for your work, will take a look | 15:59 |
| marios | SumitNaiksatam: thanks! | 15:59 |
| *** jorisroovers has quit IRC | 16:02 | |
| *** jorisroovers has joined #openstack-neutron | 16:02 | |
| *** HenryG_ has joined #openstack-neutron | 16:05 | |
| *** vkozhukalov has quit IRC | 16:05 | |
| *** HenryG has quit IRC | 16:05 | |
| *** bvandenh has quit IRC | 16:06 | |
| *** jroovers has joined #openstack-neutron | 16:06 | |
| *** SushilKM has quit IRC | 16:09 | |
| *** jorisroovers has quit IRC | 16:10 | |
| *** jistr has quit IRC | 16:16 | |
| *** yfried has joined #openstack-neutron | 16:16 | |
| *** jroovers has quit IRC | 16:16 | |
| *** djbkd has joined #openstack-neutron | 16:17 | |
| *** HenryG_ has quit IRC | 16:21 | |
| *** HenryG has joined #openstack-neutron | 16:23 | |
| *** roeyc has joined #openstack-neutron | 16:31 | |
| openstackgerrit | Maru Newby proposed a change to openstack/neutron: Send DHCP notifications regardless of agent status https://review.openstack.org/61168 | 16:37 |
| openstackgerrit | Sylvain Afchain proposed a change to openstack/neutron: Dnsmasq uses all agent IPs as nameservers https://review.openstack.org/61067 | 16:40 |
| *** mlavalle has joined #openstack-neutron | 16:43 | |
| *** marun has quit IRC | 16:46 | |
| *** HenryG has quit IRC | 16:47 | |
| *** SumitNaiksatam has quit IRC | 16:50 | |
| *** bvandenh has joined #openstack-neutron | 16:56 | |
| *** SushilKM has joined #openstack-neutron | 16:56 | |
| *** nati_ueno has joined #openstack-neutron | 16:57 | |
| openstackgerrit | Jon Grimm proposed a change to openstack/neutron: Openvswitch update_port should return updated port info https://review.openstack.org/58847 | 16:59 |
| *** SushilKM has quit IRC | 17:01 | |
| *** clev has quit IRC | 17:01 | |
| openstackgerrit | Jon Grimm proposed a change to openstack/neutron: Openvswitch update_port should return updated port info https://review.openstack.org/58847 | 17:05 |
| *** clev has joined #openstack-neutron | 17:06 | |
| openstackgerrit | Jon Grimm proposed a change to openstack/neutron: Fix ml2 & nec plugins for allowedaddresspairs tests https://review.openstack.org/58896 | 17:08 |
| *** clev has quit IRC | 17:08 | |
| *** simon-AS559 has quit IRC | 17:10 | |
| *** SumitNaiksatam has joined #openstack-neutron | 17:12 | |
| *** SushilKM has joined #openstack-neutron | 17:16 | |
| *** yfried has quit IRC | 17:18 | |
| anteaya | mestery: the -infra meeting is today at 1900 utc, do you think that multi-node testing still needs to be on the agenda? https://wiki.openstack.org/wiki/Meetings/InfraTeamMeeting | 17:21 |
| mestery | anteaya: Lets take it off the agenda for today. I setup an IRC meeting to discuss on Thursday with folks in the community, so perhaps next week we will have more to say on it. | 17:22 |
| mestery | anteaya: Sound ok? | 17:22 |
| anteaya | sure | 17:23 |
| anteaya | do you want to amend the agenda? | 17:23 |
| mestery | anteaya: Sure, I'll take care of it. | 17:23 |
| anteaya | thanks | 17:23 |
| *** ygbo has quit IRC | 17:28 | |
| mestery | anteaya: Done | 17:32 |
| anteaya | thanks | 17:32 |
| *** yfried has joined #openstack-neutron | 17:37 | |
| *** clev has joined #openstack-neutron | 17:39 | |
| *** safchain has quit IRC | 17:41 | |
| *** clev has quit IRC | 17:42 | |
| *** fouxm has quit IRC | 17:45 | |
| *** nati_ueno has quit IRC | 17:45 | |
| openstackgerrit | Edward Hope-Morley proposed a change to openstack/neutron: Adds optional timeout for neutronclient https://review.openstack.org/61183 | 17:45 |
| *** fouxm has joined #openstack-neutron | 17:45 | |
| *** jdev789 has joined #openstack-neutron | 17:49 | |
| *** jp_at_hp has quit IRC | 17:49 | |
| *** HenryG has joined #openstack-neutron | 17:56 | |
| *** jlibosva1 has quit IRC | 17:57 | |
| *** clev has joined #openstack-neutron | 17:59 | |
| *** ihrachyska has quit IRC | 18:01 | |
| *** jlibosva has joined #openstack-neutron | 18:03 | |
| *** SushilKM has quit IRC | 18:07 | |
| *** nati_ueno has joined #openstack-neutron | 18:08 | |
| *** yfried has quit IRC | 18:12 | |
| *** x86brandon has joined #openstack-neutron | 18:16 | |
| *** ihrachyska has joined #openstack-neutron | 18:16 | |
| *** vkozhukalov has joined #openstack-neutron | 18:18 | |
| anteaya | mestery: I'd like to attend the Thursday third party testing meeting | 18:20 |
| anteaya | just thought I | 18:20 |
| anteaya | 'd let you know | 18:20 |
| openstackgerrit | dekehn proposed a change to openstack/neutron: extra_dhcp_opt add checks for empty strings https://review.openstack.org/59858 | 18:22 |
| *** Abhishek_ has joined #openstack-neutron | 18:23 | |
| *** jistr has joined #openstack-neutron | 18:26 | |
| *** fouxm has quit IRC | 18:27 | |
| mestery | anteaya: Great! Would be great to have you there, thanks! | 18:28 |
| *** yfried has joined #openstack-neutron | 18:29 | |
| anteaya | I might be able to offer something useful, who knows | 18:30 |
| *** Shri_ has joined #openstack-neutron | 18:34 | |
| anteaya | mlavalle: hello | 18:45 |
| anteaya | anything new to report regarding https://bugs.launchpad.net/tempest/+bug/1259282 ? | 18:45 |
| *** peristeri has quit IRC | 18:45 | |
| *** jdev789 has quit IRC | 18:46 | |
| *** roeyc has quit IRC | 18:50 | |
| *** jprovazn has joined #openstack-neutron | 18:51 | |
| *** jlibosva has quit IRC | 18:52 | |
| *** ihrachyska has quit IRC | 18:52 | |
| anteaya | armax: this bug seems to still show up https://bugs.launchpad.net/neutron/+bug/1244255 would you be able to converse with SpamapS to see if you can update the bug report? | 18:52 |
| *** mlavalle has quit IRC | 18:55 | |
| *** salv-orlando has joined #openstack-neutron | 18:56 | |
| *** mlavalle has joined #openstack-neutron | 18:58 | |
| *** jlibosva has joined #openstack-neutron | 19:04 | |
| *** ihrachyska has joined #openstack-neutron | 19:04 | |
| *** Abhishek_ has quit IRC | 19:07 | |
| *** salv-orlando has quit IRC | 19:13 | |
| mlavalle | anteaya: nothing new to report in regards to https://bugs.launchpad.net/tempest/+bug/1259282 nati_ueno offered yesterday during the meeting to submit a fix for it. I just went through the Gerrit submissions of today and it's not there yet. As soon as he send the patchset, I will review it | 19:18 |
| anteaya | mlavalle: can you ping him and ask how he is doing? | 19:19 |
| mlavalle | sure | 19:19 |
| mlavalle | i'll keep you posted | 19:19 |
| *** salv-orlando has joined #openstack-neutron | 19:19 | |
| nati_ueno | anteaya: I'm working on 1259282. I'll send patch today | 19:20 |
| mlavalle | neti_ueno: thanks…. :-) | 19:20 |
| mlavalle | nati_ueno | 19:21 |
| anteaya | nati_ueno: awesome, thank you | 19:21 |
| *** suresh12 has joined #openstack-neutron | 19:24 | |
| mlavalle | anteaya, nati_ueno: as soon as that patchset merges, the behaviour of https://bugs.launchpad.net/neutron/+bug/1251448 will change…. I'll go back to it and review what the impact is | 19:24 |
| anteaya | mlavalle: great, thank you | 19:24 |
| openstackgerrit | Marios Andreou proposed a change to openstack/neutron: Validate CIDR given as ip-prefix in security-group-rule-create https://review.openstack.org/59212 | 19:25 |
| *** mordred has quit IRC | 19:26 | |
| *** mordred has joined #openstack-neutron | 19:26 | |
| openstackgerrit | Justin Hammond proposed a change to openstack/neutron: Update common network type consts to same origin https://review.openstack.org/44942 | 19:30 |
| *** jlibosva has quit IRC | 19:32 | |
| openstackgerrit | Praneet Bachheti proposed a change to openstack/neutron: Juniper Contrail plug-in implementation for core resources https://review.openstack.org/43793 | 19:33 |
| *** salv-orlando has quit IRC | 19:34 | |
| *** bvandenh has quit IRC | 19:41 | |
| *** gdubreui has joined #openstack-neutron | 20:00 | |
| anteaya | roaet: yes | 20:13 |
| roaet | sorry, guess that was the wrong place to ask. | 20:13 |
| anteaya | I will reply in a bit, need to pull myself away from the keyboard and go for a walk | 20:13 |
| anteaya | np | 20:13 |
| anteaya | deva and dkranz were just in a good conversation | 20:14 |
| anteaya | roaet so SlickNik and romcheg are also creating tempest tests for their projects | 20:15 |
| anteaya | they are in a bit of a different situation than -neutron since they are incubated but not integrated | 20:15 |
| anteaya | -neutron is integrated | 20:15 |
| anteaya | but it will still be helpful to talk with them in -qa | 20:15 |
| *** suresh12 has quit IRC | 20:16 | |
| roaet | Ah ok. I'll keep that in mind. | 20:16 |
| roaet | I will write their names down. | 20:16 |
| anteaya | devananda is PTL for ironic and is a good person to talk to as well | 20:16 |
| anteaya | :D | 20:16 |
| *** suresh12 has joined #openstack-neutron | 20:18 | |
| *** jistr has quit IRC | 20:25 | |
| *** Sreedhar has quit IRC | 20:26 | |
| *** Shri_ has quit IRC | 20:28 | |
| *** ywu has quit IRC | 20:28 | |
| *** alex_klimov has joined #openstack-neutron | 20:36 | |
| *** suresh12 has quit IRC | 20:38 | |
| *** jprovazn has quit IRC | 20:42 | |
| *** suresh12 has joined #openstack-neutron | 20:55 | |
| *** pcm_ has quit IRC | 21:00 | |
| roaet | my https://review.openstack.org/#/c/44942/ review is failing gate-neutron-docs and I have never run into that gate before. What should I do to fix this? | 21:02 |
| *** HenryG has quit IRC | 21:05 | |
| *** afazekas has quit IRC | 21:06 | |
| annegentle | roaet: We had to pin to a version of sphinx <1.2, I think you can just recheck no bug | 21:11 |
| roaet | annegentle: thanks I"ll do that | 21:11 |
| *** julim has quit IRC | 21:12 | |
| *** jecarey has quit IRC | 21:15 | |
| *** carl_baldwin has joined #openstack-neutron | 21:19 | |
| *** otherwiseguy has quit IRC | 21:20 | |
| openstackgerrit | Carl Baldwin proposed a change to openstack/neutron: Remove release_lease from the DHCP driver interface https://review.openstack.org/56285 | 21:26 |
| *** otherwiseguy has joined #openstack-neutron | 21:32 | |
| anteaya | or do we need to update the version of sphinx neutron uses? | 21:33 |
| anteaya | I remember reading that as a possiblity in backscroll somewhere | 21:33 |
| openstackgerrit | Nachi Ueno proposed a change to openstack/neutron: Improve vif attributes related with firewalling https://review.openstack.org/21946 | 21:34 |
| annegentle | roaet: anteaya: yes, actually, pin the version of sphinx for nuetron, I thought there was a global setting but there's not | 21:34 |
| annegentle | er, neutron | 21:34 |
| anteaya | annegentle: great thanks | 21:35 |
| anteaya | roaet: do you need more than that? | 21:35 |
| roaet | nope | 21:35 |
| anteaya | very good | 21:35 |
| roaet | That sounds like a different change | 21:35 |
| anteaya | yes | 21:35 |
| anteaya | it is | 21:35 |
| roaet | or a bug or something | 21:35 |
| roaet | Maybe I"ll make a bug | 21:35 |
| anteaya | sounds good | 21:35 |
| *** alex_klimov has quit IRC | 21:42 | |
| openstackgerrit | Morgan Fainberg proposed a change to openstack/neutron: Sync global requirements to pin sphinx to sphinx>=1.1.2,<1.2 https://review.openstack.org/61245 | 22:03 |
| anteaya | roaet: you had some concerns about Montreal in January: http://montreal.about.com/od/Montreal-Weather/ss/montreal-january-weather-climate-temperatures-average.htm | 22:05 |
| anteaya | roaet: what was the url for your sphnix version bug? | 22:05 |
| anteaya | I don't see one that mentions sphnix | 22:09 |
| anteaya | I was going to ask morganfainberg to link to in his commit message for 61245 | 22:10 |
| morganfainberg | anteaya, hi! | 22:10 |
| morganfainberg | anteaya, if you guys don't merge that before infra gets the auto-proposal job fixed it's all good. just figured i'd help since i had neutron cloned already. | 22:13 |
| anteaya | morganfainberg: grateful for your help, thank you | 22:14 |
| anteaya | what time line for auto-proposal job fixed? do you know? | 22:14 |
| morganfainberg | anteaya, they are working on it. no ETA afaik | 22:15 |
| anteaya | the current -infra backscroll is all maven plugin | 22:15 |
| anteaya | I can feel jeblair's frustration from here | 22:15 |
| anteaya | k | 22:15 |
| morganfainberg | anteaya, clarkb asked me to leave at least one project w/o the proposed fix | 22:15 |
| morganfainberg | he isn't sure everyone will have it merged by then | 22:15 |
| morganfainberg | so, today would be my guess (ish) | 22:16 |
| morganfainberg | i only did patchsets for what i had cloned locally | 22:16 |
| anteaya | thanks | 22:22 |
| *** litong has quit IRC | 22:23 | |
| *** aymenfrikha has quit IRC | 22:24 | |
| *** x86brandon has quit IRC | 22:30 | |
| anteaya | if jlibosvar is around, is he able to proceed with https://review.openstack.org/#/c/58695/ ? | 22:39 |
| *** mlavalle has quit IRC | 22:43 | |
| *** mlavalle has joined #openstack-neutron | 22:47 | |
| anteaya | looks like we need the sphnix pin in before any of our check or gate jobs pass: https://review.openstack.org/#/c/61245/ who is up markmcclain nati_ueno enikanorov__ obondarev arosen armax mestery marun rkukura SumitNaiksatam garyk ? | 22:56 |
| SumitNaiksatam | anteaya: yeah, i am here | 22:56 |
| nati_ueno | anteaya: me too | 22:56 |
| nati_ueno | what's up? | 22:56 |
| anteaya | awesome thanks | 22:57 |
| mestery | Who's on first? | 22:57 |
| mestery | :) | 22:57 |
| mestery | Checking anteaya. | 22:57 |
| anteaya | can you review 61245 please | 22:57 |
| anteaya | thanks | 22:57 |
| *** x86brandon has joined #openstack-neutron | 22:57 | |
| *** pcm_ has joined #openstack-neutron | 22:57 | |
| nati_ueno | mestery is 1st :P | 22:57 |
| anteaya | we need it for our check and gate tests to pass | 22:57 |
| nati_ueno | anteaya: so we should review it ? | 22:57 |
| anteaya | please | 22:57 |
| mestery | Looks ok to me | 22:57 |
| SumitNaiksatam | +2'ed | 22:57 |
| anteaya | great | 22:57 |
| anteaya | do we have enough for an approval? | 22:58 |
| *** pcm_ has quit IRC | 22:58 | |
| * mestery nods and is trying to approve now | 22:58 | |
| anteaya | thank you | 22:58 |
| SumitNaiksatam | we have confirmed with openstack global requirements, right? | 22:58 |
| mestery | approved | 22:58 |
| anteaya | SumitNaiksatam: I have not | 22:58 |
| *** pcm_ has joined #openstack-neutron | 22:58 | |
| mestery | yikes | 22:58 |
| anteaya | just following the conversation in -infra | 22:58 |
| SumitNaiksatam | anteaya: ok i guess that should be authoritative :-) | 22:59 |
| SumitNaiksatam | let me check | 22:59 |
| anteaya | SumitNaiksatam: yes please | 22:59 |
| * mestery ducks out to head to a choir concert his daughter is signing in ... | 23:00 | |
| nati_ueno | mestery: youtube it :P | 23:01 |
| anteaya | http://git.openstack.org/cgit/openstack/requirements/tree/global-requirements.txt#n144 | 23:01 |
| anteaya | mestery: have a great time | 23:01 |
| SumitNaiksatam | anteaya: yeah | 23:01 |
| mestery | nati_ueno: Ha! Don't tempt me, I just may.. Thanks anteaya! | 23:01 |
| SumitNaiksatam | it has gone it today: https://github.com/openstack/requirements/commit/0e8862e958813a031fe7398440e8a3866b42a8e4 | 23:01 |
| nati_ueno | mestery: ha ha have a great time! | 23:01 |
| anteaya | SumitNaiksatam: awesome we are good to go | 23:01 |
| SumitNaiksatam | mestery: don't get excited, nati_ueno did not mean record your singing :-P | 23:02 |
| anteaya | thanks all for the fast response | 23:02 |
| nati_ueno | anteaya: Thank you for working on this | 23:02 |
| anteaya | :D | 23:03 |
| SumitNaiksatam | anteaya: any time! | 23:03 |
| anteaya | SumitNaiksatam: :D | 23:03 |
| SumitNaiksatam | :-) | 23:03 |
| *** jpich has quit IRC | 23:05 | |
| anteaya | it appears that neutron had been subscribed to global requirements changes (see Dec. 9th) but was not subscribed for this change: https://review.openstack.org/#/q/topic:openstack/requirements,n,z | 23:09 |
| anteaya | does anyone have any ideas why that might be? | 23:10 |
| *** jgrimm has quit IRC | 23:10 | |
| *** clev has quit IRC | 23:12 | |
| *** yamahata_ has joined #openstack-neutron | 23:12 | |
| nati_ueno | anteaya: I'm not sure. But it was working, but they have some bug. so may be it is not fixed yet | 23:13 |
| anteaya | okay, yes I posed a question in -infra, this is the list of subscribed projects: http://git.openstack.org/cgit/openstack/requirements/tree/projects.txt | 23:17 |
| anteaya | and we are listed | 23:17 |
| anteaya | and the list is used by requirements_update.sh: http://git.openstack.org/cgit/openstack-infra/config/tree/modules/jenkins/files/slave_scripts/propose_requirements_update.sh | 23:18 |
| anteaya | so I am uncertain where the error is | 23:18 |
| *** dims has quit IRC | 23:21 | |
| nati_ueno | hmm | 23:25 |
| *** HenryG has joined #openstack-neutron | 23:27 | |
| *** heyongli has joined #openstack-neutron | 23:33 | |
| openstackgerrit | Shiv Haris proposed a change to openstack/neutron: Testing tempest with neutron https://review.openstack.org/61273 | 23:39 |
| anteaya | nati_ueno: if you are following the conversation in -infra it isn't just us affected, nova, keystone, heat were as well | 23:39 |
| anteaya | something different about how the job runs `git review -d` and how clarkb ran `git diff --quiet` manually | 23:40 |
| anteaya | so this just got merged: https://review.openstack.org/#/c/61268/ | 23:40 |
| anteaya | so we will see what difference that makes | 23:40 |
| openstackgerrit | Jenkins proposed a change to openstack/neutron: Updated from global requirements https://review.openstack.org/60974 | 23:42 |
| openstackgerrit | Jenkins proposed a change to openstack/python-neutronclient: Updated from global requirements https://review.openstack.org/56654 | 23:45 |
| anteaya | so we appear to be fixed again for auto-magic global requirements from Jenkins | 23:45 |
| *** aymenfrikha has joined #openstack-neutron | 23:53 | |
| *** mlavalle has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!