| *** saneax is now known as saneax-_-|AFK | 00:01 | |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: [NOT_FOR_REVIEW] Test py35 devstack https://review.openstack.org/419683 | 00:19 |
|---|---|---|
| *** hongbin has quit IRC | 00:19 | |
| *** limao has joined #openstack-kuryr | 00:37 | |
| *** yedongcan has joined #openstack-kuryr | 01:15 | |
| *** salv-orlando has joined #openstack-kuryr | 01:19 | |
| *** tonanhngo has quit IRC | 01:20 | |
| *** salv-orlando has quit IRC | 01:24 | |
| *** yamamoto_ has joined #openstack-kuryr | 01:57 | |
| *** yuanying has quit IRC | 02:52 | |
| *** hongbin has joined #openstack-kuryr | 02:59 | |
| *** hongbin_ has joined #openstack-kuryr | 03:03 | |
| *** hongbin has quit IRC | 03:03 | |
| *** salv-orlando has joined #openstack-kuryr | 03:21 | |
| *** salv-orlando has quit IRC | 03:25 | |
| *** limao has quit IRC | 03:50 | |
| *** yuanying has joined #openstack-kuryr | 03:52 | |
| *** limao has joined #openstack-kuryr | 04:19 | |
| *** janki has joined #openstack-kuryr | 04:28 | |
| *** limao has quit IRC | 04:28 | |
| *** limao has joined #openstack-kuryr | 05:00 | |
| *** hongbin_ has quit IRC | 05:02 | |
| openstackgerrit | Xu Ao proposed openstack/fuxi: RBD:Make the volume to do local attach after created https://review.openstack.org/419767 | 05:26 |
| *** v1k0d3n has joined #openstack-kuryr | 05:26 | |
| *** limao has quit IRC | 05:37 | |
| *** yamamoto_ has quit IRC | 05:37 | |
| *** limao has joined #openstack-kuryr | 05:38 | |
| *** yuanying has quit IRC | 06:06 | |
| *** salv-orlando has joined #openstack-kuryr | 06:07 | |
| *** yuanying has joined #openstack-kuryr | 06:07 | |
| *** yuanying has quit IRC | 06:11 | |
| *** yuanying_ has joined #openstack-kuryr | 06:11 | |
| *** yuanying_ has quit IRC | 06:12 | |
| *** yuanying has joined #openstack-kuryr | 06:12 | |
| *** v1k0d3n has quit IRC | 06:12 | |
| *** yuanying has quit IRC | 06:17 | |
| *** yamamoto_ has joined #openstack-kuryr | 06:21 | |
| *** salv-orlando has quit IRC | 06:43 | |
| *** yuanying has joined #openstack-kuryr | 07:19 | |
| *** salv-orlando has joined #openstack-kuryr | 07:27 | |
| *** portdirect has quit IRC | 07:30 | |
| *** portdirect has joined #openstack-kuryr | 07:32 | |
| *** janki has quit IRC | 07:42 | |
| *** janki has joined #openstack-kuryr | 07:43 | |
| *** yedongcan1 has joined #openstack-kuryr | 07:44 | |
| *** yedongcan has quit IRC | 07:47 | |
| *** tonanhngo has joined #openstack-kuryr | 07:49 | |
| *** tonanhngo has quit IRC | 07:53 | |
| *** tonanhngo has joined #openstack-kuryr | 08:02 | |
| *** tonanhngo has quit IRC | 08:03 | |
| *** salv-orlando has quit IRC | 08:43 | |
| *** salv-orlando has joined #openstack-kuryr | 08:44 | |
| *** garyloug has joined #openstack-kuryr | 09:43 | |
| openstackgerrit | Xuepeng Ji proposed openstack/fuxi: Removes unnecessary utf-8 encoding https://review.openstack.org/419859 | 09:53 |
| *** devvesa has joined #openstack-kuryr | 09:53 | |
| *** limao has quit IRC | 10:03 | |
| *** openstackgerrit has quit IRC | 10:18 | |
| *** salv-orl_ has joined #openstack-kuryr | 10:23 | |
| *** salv-orlando has quit IRC | 10:25 | |
| mchiappero | apuimedo: yes, Kural and I are part of the same team | 10:33 |
| mchiappero | (sorry for lagging, but had some other stuff to do this week) | 10:33 |
| mchiappero | apuimedo: BTW, I need to have a quick chat with you, I'm sure whether you can and I'll be able today, but if possible soon :) | 10:54 |
| *** mchiappero has quit IRC | 10:55 | |
| *** mchiappero has joined #openstack-kuryr | 10:57 | |
| apuimedo | mchiappero: no worries | 11:14 |
| apuimedo | mchiappero: I can today | 11:14 |
| *** limao has joined #openstack-kuryr | 11:31 | |
| *** limao has quit IRC | 11:36 | |
| *** limao_ has joined #openstack-kuryr | 11:36 | |
| *** garyloug has quit IRC | 11:41 | |
| mchiappero | apuimedo: ok, I'll try to ping you later after lunch :) | 11:46 |
| mchiappero | thanks | 11:46 |
| apuimedo | very well | 11:46 |
| *** garyloug has joined #openstack-kuryr | 11:50 | |
| *** salv-orl_ has quit IRC | 11:55 | |
| *** salv-orlando has joined #openstack-kuryr | 11:59 | |
| *** tonanhngo has joined #openstack-kuryr | 12:04 | |
| *** tonanhngo has quit IRC | 12:05 | |
| *** yamamoto_ has quit IRC | 12:11 | |
| *** openstackgerrit has joined #openstack-kuryr | 12:41 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi: Updated from global requirements https://review.openstack.org/419929 | 12:41 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-libnetwork: Updated from global requirements https://review.openstack.org/419932 | 12:43 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-kubernetes: Updated from global requirements https://review.openstack.org/419933 | 12:43 |
| *** garyloug has quit IRC | 12:55 | |
| *** yedongcan1 has left #openstack-kuryr | 13:00 | |
| *** neiljerram has quit IRC | 13:04 | |
| *** yamamoto has joined #openstack-kuryr | 13:04 | |
| *** yamamoto_ has joined #openstack-kuryr | 13:05 | |
| dougbtv | mornin' apuimedo :) I dragged in another guy from the nfvpe team leifmadsen and we talked quite a bit about a sample NFV work-load to use as a demo VNF for Kuryr | 13:07 |
| dougbtv | we're still thinking generic SIP trunking situation, and one of our considerations is to make a sample instead of a direct partner requirement | 13:07 |
| dougbtv | we think this will insulate everyone from the potential for failure -- the partner, kuryr, and the NFV partner engineering team... which is better than disappointing everyone :) | 13:08 |
| *** neiljerram has joined #openstack-kuryr | 13:08 | |
| dougbtv | ...but we still need to run this by our leadership to make sure we're on the right track | 13:08 |
| dougbtv | we're pretty sure it's going to be a go-ahead, we're all kind of rooting for kuryr, so if we can help... we'll be psyched :) | 13:08 |
| *** yamamoto has quit IRC | 13:09 | |
| apuimedo | dougbtv: cool | 13:09 |
| apuimedo | can't wait! | 13:09 |
| dougbtv | awesome! cool, so that sounds like it's in the ballpark for what might be helpful for you guys for an NFV demo? | 13:10 |
| apuimedo | I think so | 13:10 |
| apuimedo | the most important thing is the requirements | 13:10 |
| apuimedo | and to define what we want to see | 13:11 |
| apuimedo | of the kuryr-kubernetes demo example | 13:11 |
| apuimedo | if it's cool we can show it at the Boston summit demo | 13:11 |
| janki | apuimedo, dougbtv Sorry to pitch in mid-discussion. Heard NFV and couldnot resist. I would like to help out in any way possible | 13:11 |
| apuimedo | The more, the merrier | 13:11 |
| dougbtv | awesome, I'm going to talk to leifmadsen about putting those together for you from our side, maybe early-ish next week we can have some reqs together | 13:12 |
| ivc_ | dougbtv apuimedo do we expect DPDK to work with kuryr/kubernetes? | 13:12 |
| ivc_ | for NFV | 13:12 |
| dougbtv | definitely janki !! | 13:12 |
| janki | dougbtv, I contribute to Tacker - incase this info helps :) | 13:12 |
| dougbtv | ...this particular situation wouldn't use DPDK right off the bat, but, leifmadsen and I did discuss it, so, potentially could be integrated later, or if people are interested in that earlier rather than later... we can go for it | 13:12 |
| * dougbtv wags his eyebrows | 13:13 | |
| apuimedo | ivc_: That's something I definitely would like to look at | 13:13 |
| dougbtv | Tacker integration sounds neat | 13:13 |
| apuimedo | I think jerms did some dpdk on containers | 13:14 |
| dougbtv | also that'd be rad regarding Boston summit :) | 13:14 |
| apuimedo | but IIRC it needed the container to run in privileged mode | 13:14 |
| apuimedo | which if possible I'd like to avoid | 13:14 |
| janki | dougbtv, I have registered a BP to containarise VNF - https://blueprints.launchpad.net/tacker/+spec/containerised-vnfs | 13:14 |
| *** garyloug has joined #openstack-kuryr | 13:15 | |
| dougbtv | janki, nice! thanks for the pointer to that doc. let's definitely make sure to keep in touch | 13:16 |
| janki | dougbtv, sounds cool :) | 13:16 |
| dougbtv | really pumped about the whole she-bang :D | 13:17 |
| *** dougbtv is now known as dougbtv|bbiaf | 13:36 | |
| *** yamamoto_ has quit IRC | 13:37 | |
| *** yamamoto has joined #openstack-kuryr | 13:40 | |
| *** dougbtv|bbiaf is now known as dougbtv | 13:42 | |
| leifmadsen | hell yea | 13:46 |
| leifmadsen | dougbtv can attest to my excitement in our discussion yesterday :) | 13:46 |
| dougbtv | +1 | 13:47 |
| leifmadsen | so let me lay out one of the possible scenarios I threw out, and see if this sounds at all like we're in push in the right direction: | 13:47 |
| leifmadsen | A pair of containers utilizing kuryr for the networking aspects, with a couple of VNF components: 1) Asterisk container for SIP/RTP traffic, 2) HoMER for monitoring and stats gathering of the traffic, 3) SIPp for traffic generation across a network fabric, and 4) a stupid simple controller that would push information into the Asterisk VNF components in order to setup the authentication for SIP, | 13:49 |
| leifmadsen | and the routes for the RTP between a pair of Asterisk boxes | 13:49 |
| leifmadsen | For the networking aspect, we could do something like a virtual MPLS circuit that has guarantees for latency and jitter, load it up, and then do some other analysis about what happens when we start overloading the circuit, and do validations to understand if traffic is moving as expected | 13:50 |
| leifmadsen | we could even throw in something like skydive for an additional view of the network analysis | 13:50 |
| leifmadsen | EOL | 13:50 |
| leifmadsen | that's just my spit balling some ideas, but like what dougbtv said earlier, we'd like to work with some VoIP traffic, because that's our backgrounds, so that overhead for dealing with the traffic generation and routing itself is trivial for us | 13:51 |
| apuimedo | leifmadsen: ok | 13:51 |
| apuimedo | there may be some issues with VoIP, since I think that neutron-lbaasv2 is limitted for udp traffic at the api level | 13:52 |
| leifmadsen | we can always build up additional complexity, but we're thinking this would be a nice VNF that has a lot of uses, easy to generate a large amount of PPS and traffic, etc | 13:52 |
| leifmadsen | that's perfect | 13:52 |
| apuimedo | but if we use a single pod for demo we should be fine | 13:52 |
| leifmadsen | oh wait, you mean limited in terms of not being good for UDP | 13:52 |
| apuimedo | xD | 13:52 |
| leifmadsen | we can always encapsulate it in TCP | 13:52 |
| leifmadsen | it's not ideal for voip usually, but Asterisk can at least deal with the TCP aspects | 13:52 |
| leifmadsen | or like you said, if we can setup an environment where UDP is fine, then that's even better | 13:53 |
| dougbtv | yeah TCP encap is a good stop-gap for a demo for sure | 13:53 |
| leifmadsen | the idea also being that we have other use cases where this VNF could be useful in other areas like OPNFV, etc | 13:53 |
| leifmadsen | anyways, something we can discuss further and make sure we're meeting whatever needs were wanted | 13:54 |
| apuimedo | in terms of interfaces, what do you usually put on the VMs that work as vnfs now? | 13:55 |
| apuimedo | one sriov and one on the overlay? | 13:55 |
| leifmadsen | I think we have many options there | 13:55 |
| leifmadsen | but I think that makes some sense | 13:55 |
| leifmadsen | probably SR-IOV for the ingress/egress traffic, then overlay for the virtual circuits between VNFs | 13:56 |
| apuimedo | which are the most typical options? | 13:56 |
| leifmadsen | I've never done this before lol | 13:56 |
| leifmadsen | (has anyone?) | 13:56 |
| leifmadsen | we could talk with an expert on our team to make sure we're architecting this the "typical" way though | 13:57 |
| leifmadsen | but I think it makes sense to me that you'd have some ingress/egress circuits to the black box of virtual goodness | 13:57 |
| leifmadsen | basically trusted and untrusted circuits, and the VNF straddling that | 13:57 |
| leifmadsen | that's how I'm visualizing it | 13:58 |
| apuimedo | leifmadsen: with VMs I'm sure it's been done | 13:58 |
| leifmadsen | SR-IOV would be your "untrusted" or public side | 13:58 |
| dougbtv | yeah we should run it by some experts on our team in that arena, we're the integrator guys | 13:58 |
| leifmadsen | apuimedo: yea, and dougbtv and I talked, and really the VNF components could be built for both VM and containers, and how the network is setup could be done either way | 13:58 |
| leifmadsen | the topology wouldn't need to change really | 13:58 |
| leifmadsen | just replace containers with VMs, but everything else the same | 13:59 |
| leifmadsen | like I said, I'm mostly spitballing on this, but a typical voip scenario would have trusted and untrusted circuits straddled by a proxy/SBC/B2BUA | 13:59 |
| leifmadsen | so we have a lot of options to intermixing circuits and tech, just need to talk to our subject matter experts and make sure we build it in a manner that would match best practices | 14:00 |
| leifmadsen | I'm not too worried about that part though | 14:00 |
| apuimedo | good | 14:04 |
| apuimedo | :-) | 14:04 |
| openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr-kubernetes: docs: Fix image rendering https://review.openstack.org/419981 | 14:11 |
| *** v1k0d3n has joined #openstack-kuryr | 14:11 | |
| openstackgerrit | Merged openstack/kuryr-libnetwork: Add subnetpool_id tag for existing Neutron subnet https://review.openstack.org/414455 | 14:24 |
| openstackgerrit | Merged openstack/kuryr-kubernetes: Kubernetes Service support design reference document https://review.openstack.org/416228 | 14:28 |
| openstackgerrit | Merged openstack/kuryr-kubernetes: Update requirements https://review.openstack.org/419281 | 14:33 |
| *** yamamoto has quit IRC | 14:35 | |
| *** limao has joined #openstack-kuryr | 14:37 | |
| *** salv-orlando has quit IRC | 14:38 | |
| *** limao_ has quit IRC | 14:39 | |
| *** limao has quit IRC | 14:41 | |
| ltomasbo | this sounds really cool, I would like to be involved too! | 14:42 |
| dougbtv | o/ | 14:42 |
| ltomasbo | janki, as regards to this: https://review.openstack.org/#/c/397233/ | 14:42 |
| ltomasbo | does this means that tacker will be able to handle both VMs and Containers? | 14:43 |
| janki | ltomasbo, yes.. | 14:43 |
| ltomasbo | that would fit really well as a VNF-Manager then | 14:43 |
| janki | ltomasbo, yes, it will handle both vm and container | 14:43 |
| *** limao has joined #openstack-kuryr | 14:43 | |
| ltomasbo | I've been looking at OPNFV and one of the problems they have is the lack of joint VMs and container management tools | 14:44 |
| ltomasbo | I've seen some proposals to use tacker for VMs and kubernetes for containers | 14:45 |
| ltomasbo | but that would be a great abstraction for both of them | 14:45 |
| ltomasbo | on the topic of VNFs being always VMs, there is already some virtual network functions running in containers, such as clearwater | 14:46 |
| ltomasbo | which is a VNF router | 14:46 |
| apuimedo | ltomasbo: you must xD | 14:47 |
| ltomasbo | apuimedo, :DDD | 14:47 |
| ltomasbo | apuimedo, I'll definitely do! | 14:47 |
| apuimedo | great! | 14:47 |
| janki | ltomasbo, sounds interesting.. | 14:48 |
| janki | ltomasbo, apuimedo please share your thoughts on the spec too :) | 14:48 |
| janki | dougbtv, ^ | 14:48 |
| dougbtv | +1! | 14:49 |
| dougbtv | idea is really promising | 14:49 |
| *** yamamoto has joined #openstack-kuryr | 14:50 | |
| openstackgerrit | Merged openstack/fuxi: Provide Keystone session for Nova with its own group https://review.openstack.org/419224 | 14:52 |
| *** janki has quit IRC | 14:52 | |
| openstackgerrit | Merged openstack/fuxi: Use the default executor in os-brick https://review.openstack.org/408845 | 14:52 |
| *** yamamoto has quit IRC | 14:54 | |
| *** hongbin has joined #openstack-kuryr | 15:00 | |
| *** tonanhngo has joined #openstack-kuryr | 15:01 | |
| *** tonanhngo has quit IRC | 15:02 | |
| openstackgerrit | dengshaolin proposed openstack/fuxi: Replace hardcode values with OptGroup in config setting https://review.openstack.org/415378 | 15:09 |
| *** limao_ has joined #openstack-kuryr | 15:31 | |
| *** limao has quit IRC | 15:32 | |
| *** limao_ has quit IRC | 16:04 | |
| *** devvesa has quit IRC | 16:30 | |
| *** pcaruana has joined #openstack-kuryr | 16:36 | |
| *** tonanhngo has joined #openstack-kuryr | 17:12 | |
| *** dougbtv is now known as dougbtv|lunch | 17:36 | |
| *** salv-orlando has joined #openstack-kuryr | 17:45 | |
| *** garyloug has quit IRC | 18:02 | |
| *** huikang has joined #openstack-kuryr | 18:10 | |
| *** huikang has quit IRC | 18:22 | |
| *** huikang has joined #openstack-kuryr | 18:26 | |
| *** pcaruana has quit IRC | 18:26 | |
| *** dougbtv|lunch is now known as dougbtv | 18:39 | |
| *** pcaruana has joined #openstack-kuryr | 18:39 | |
| *** huikang has quit IRC | 18:41 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi: Updated from global requirements https://review.openstack.org/419929 | 18:42 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-libnetwork: Updated from global requirements https://review.openstack.org/419932 | 18:44 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-kubernetes: Updated from global requirements https://review.openstack.org/419933 | 18:44 |
| *** huikang has joined #openstack-kuryr | 18:47 | |
| *** huikang has quit IRC | 18:54 | |
| *** openstack has joined #openstack-kuryr | 18:58 | |
| *** huikang_ has joined #openstack-kuryr | 18:58 | |
| *** huikang_ has quit IRC | 19:00 | |
| *** pcaruana has quit IRC | 19:02 | |
| *** huikang has joined #openstack-kuryr | 19:08 | |
| *** openstack has joined #openstack-kuryr | 19:57 | |
| *** david-lyle is now known as baling-wire | 20:00 | |
| *** huikang has quit IRC | 20:06 | |
| *** huikang has joined #openstack-kuryr | 20:10 | |
| *** openstack has joined #openstack-kuryr | 20:27 | |
| *** vikasc has quit IRC | 20:29 | |
| *** vikasc has joined #openstack-kuryr | 20:32 | |
| *** salv-orlando has quit IRC | 20:32 | |
| *** huikang has quit IRC | 20:33 | |
| *** huikang has joined #openstack-kuryr | 20:34 | |
| *** huikang_ has joined #openstack-kuryr | 20:35 | |
| *** huikang_ has quit IRC | 20:37 | |
| *** huikang has quit IRC | 20:38 | |
| *** dougbtv_ has joined #openstack-kuryr | 20:40 | |
| *** dougbtv has quit IRC | 20:40 | |
| *** dougbtv_ is now known as dougbtv | 20:40 | |
| *** huikang has joined #openstack-kuryr | 21:38 | |
| *** huikang has quit IRC | 21:42 | |
| *** dougbtv has quit IRC | 21:47 | |
| *** huikang has joined #openstack-kuryr | 21:49 | |
| *** salv-orlando has joined #openstack-kuryr | 21:55 | |
| *** yamamoto has joined #openstack-kuryr | 21:56 | |
| *** yamamoto has quit IRC | 22:11 | |
| *** yamamoto has joined #openstack-kuryr | 22:13 | |
| *** yamamoto has quit IRC | 22:15 | |
| *** yamamoto has joined #openstack-kuryr | 22:20 | |
| *** yamamoto has quit IRC | 22:22 | |
| *** salv-orl_ has joined #openstack-kuryr | 22:22 | |
| *** yamamoto has joined #openstack-kuryr | 22:24 | |
| *** salv-orlando has quit IRC | 22:25 | |
| *** baling-wire is now known as david-lyle | 22:26 | |
| *** yamamoto has quit IRC | 22:26 | |
| *** huikang has quit IRC | 22:45 | |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: [NOT_FOR_REVIEW] Test py35 devstack https://review.openstack.org/419683 | 22:53 |
| *** pmannidi has quit IRC | 22:57 | |
| *** salv-orl_ has quit IRC | 23:26 | |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: [WIP] Make py35 devstack working https://review.openstack.org/419683 | 23:32 |
| *** huikang has joined #openstack-kuryr | 23:38 | |
| *** huikang has quit IRC | 23:40 | |
| *** huikang has joined #openstack-kuryr | 23:40 | |
| *** huikang has quit IRC | 23:44 | |
| *** huikang has joined #openstack-kuryr | 23:44 | |
| *** huikang has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!