06:30:18 <anil_rao> #startmeeting taas 06:30:19 <openstack> Meeting started Wed Apr 13 06:30:18 2016 UTC and is due to finish in 60 minutes. The chair is anil_rao. Information about MeetBot at http://wiki.debian.org/MeetBot. 06:30:20 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 06:30:22 <openstack> The meeting name has been set to 'taas' 06:30:30 <anil_rao> Hi 06:30:31 <yamamot__> hi 06:30:32 <soichi> hi 06:30:37 <kaz> hi 06:31:23 <anil_rao> #topic Preparation status for the summit... 06:32:00 <anil_rao> We are working on the TaaS Tech Talk presentation slides 06:32:40 <anil_rao> I recorded the TaaS demo yesterday. I'll upload it somewhere and send out a link. 06:33:14 <soichi> anil_rao: thank you 06:33:15 <reedip> youtube ? 06:33:39 <reedip> if nothing proprietary is there 06:33:46 <anil_rao> reedip: The final presentation will eventually get on YouTube. 06:34:13 <reedip> Keeping the demo on Youtube can also help during the summit 06:34:18 <anil_rao> reedip: The one I will upload doesn't have any voice because we will be speaking when the demo is played at the Summit. 06:34:47 <reedip> Okay, then whatever method seems correct :) 06:35:08 <anil_rao> I will put it up someone and provide you all with a link. May be tomorrow. 06:35:33 <anil_rao> I have also added a section on TaaS for consideration as a topic for the Design Summit 06:35:44 <anil_rao> Let's see if it gets picked up. 06:36:23 <reedip> as per yesterday's meeting 06:36:30 <reedip> for networking 06:36:43 <reedip> there is a cross-project and *-aas Session on tuesday 06:36:45 <reedip> wait 06:37:31 <yamamot__> #link https://etherpad.openstack.org/p/newton-neutron-summit-ideas 06:37:41 <reedip> https://etherpad.openstack.org/p/newton-neutron-future-adv-services 06:37:49 <soichi> i think *aaS session will be on 27th Wednesday 06:38:17 <reedip> soichi: yes 06:38:56 <reedip> anil_rao, yamamot__ ( how many nics do you have ??? ) fawadkhaliq, we can start putting our point of discussion on this etherpad: https://etherpad.openstack.org/p/newton-neutron-future-adv-services 06:39:51 <yamamot__> it seems empty 06:39:58 <anil_rao> reedip: Didn't follow the question on how many nics? 06:40:33 <reedip> it was for yamamot__ 06:40:41 <reedip> yes, its empty right now 06:41:14 <fawadkhaliq> reedip maybe neutron guys already have something like this 06:41:20 <fawadkhaliq> can you please double check 06:41:37 <yamamot__> reedip: it somehow changes on every reconnect 06:42:41 <reedip> I dont know the technical glitches affecting it , I got the link from the summit page .. 06:43:54 <yamamot__> #link https://www.openstack.org/summit/austin-2016/summit-schedule/events/9109?goback=1 06:44:32 <yamamot__> i guess this session is not intended to cover taas. 06:46:04 <soichi> i think so, too 06:47:43 <anil_rao> OK, let's move to the next topic 06:47:58 <anil_rao> #topic The start time of weekly TaaS IRC 06:48:10 <anil_rao> Any objections? 06:48:15 <reedip> *-aas is not limited to LB/FW and VPN. As Taas is also a part of the Neutron Family, I would like to see TaaS participating here as well 06:48:32 <reedip> no objection 06:48:56 <yamamot__> reedip: sure. let's ping dougwig/sridar to confirm 06:49:08 <yamamot__> anil_rao: no objection 06:49:16 <yamamot__> #link https://review.openstack.org/#/c/304383/ 06:50:03 <soichi> anil_rao: no objection 06:50:23 <yamamot__> i submitted irc-meetings change. please vote +1 if it's ok for you. i'll remove workflow -1 if i see no -1 votes. 06:50:43 <anil_rao> yamamot__: Will do. 06:51:31 <anil_rao> done 06:51:49 <reedip> +1 06:52:42 <fawadkhaliq> +! 06:52:43 <fawadkhaliq> +1 06:53:37 <anil_rao> #topic Open Discussion 06:53:52 <reedip> someone had a question ( simhon ) on #openstack-meeting-4 06:54:04 <reedip> I asked him to come here and ask the Taas telated query 06:54:15 <simhon> yes, 06:54:27 <reedip> hi simhon, go ahead 06:54:33 <simhon> I run Liberty on Ubuntu with ovs 2.0.2 06:54:33 <simhon> <simhon> configured taas service and taas flow 06:54:33 <simhon> <simhon> the ingress (in) to a vm does not work 06:54:33 <simhon> <simhon> since the rule on the ovs use vlan as a match but traffic comes without vlan 06:54:51 <simhon> packets=0, n_bytes=0, idle_age=259, priority=20,dl_vlan=3,dl_dst=fa:16:3e:d3:60:16 actions=NORMAL,mod_vlan_vid:3901,output:11 06:54:51 <simhon> <simhon> as you can see this rule never hit 06:54:51 <simhon> <simhon> because the match expects vlan 06:54:51 <simhon> <simhon> any reason for the vlan ??? 06:55:33 <simhon> the egress to the vm works just fine. problem is on the ingress only 06:55:41 <anil_rao> simhon: Are you running in a single node DevStack or on the Network node? 06:55:52 <simhon> single node DevStack 06:56:11 <anil_rao> I am guessing you have a br-ex in your OVS 06:56:37 <simhon> yes. but the ovs rules are running in br-int only 06:57:08 <anil_rao> OK, we currently have a problem with traffic coming in from br-ex because the vlan tag is not being set there. 06:57:41 <anil_rao> To answer your question, current ML2 logic 06:58:08 <anil_rao> uses vlan ids to separate out traffic between different virtual networks on a host. 06:58:45 <anil_rao> We (TaaS) do the same thing to separate out mirrored traffic belonging to a tap-service instance from regular (production) 06:58:58 <anil_rao> traffic and from traffic beloning to other tap-service instances. 06:59:17 <yamamot__> simhon: do you think it's same issue as this? https://bugs.launchpad.net/tap-as-a-service/+bug/1544176 06:59:18 <openstack> Launchpad bug 1544176 in tap-as-a-service "br-int with Normal action" [Undecided,New] 07:00:02 <simhon> Anil, thanks a lot for your answer. 07:00:40 <anil_rao> yamamot__: That might be the same problem. 07:00:52 <anil_rao> We are seeing vlan tags in only one direction with br-ex. 07:01:10 <anil_rao> As I had mentioned last week or the week before I'll look into it. 07:01:48 <simhon> Anil: I work at imVision Tech. where we do security app for NFV using BA and machine learning - we work together with Intel and we have some requirements from Taas that we would like to discuss with you 07:02:41 <simhon> Anil: could we have a open conf call together to discuss these requirements and see how they can be fullfilled using taas 07:03:28 <anil_rao> simhon: I am ok with the conf call but if you don't mind can you also please write something up on the Neutron mailing list. 07:03:52 <anil_rao> We typically use the [taas] tag following [openstack-dev][neutron] 07:04:13 <simhon> Anil: Sure. 07:04:20 <anil_rao> Perfect 07:05:01 <anil_rao> soichi,kaz: Can I discuss some dashboard items with you. 07:05:32 <soichi> sure 07:06:08 <reedip> I would not suggest openstack-dev for now .. openstack would be better 07:06:17 <reedip> openstack-dev is mainly for developers 07:06:26 <reedip> but thats my opinion 07:06:49 <anil_rao> Firstly, the TaaS dashboard looks and works great! 07:06:59 <soichi> thank you 07:07:04 <anil_rao> I did run into some issues that I would like to bring to your attention. 07:07:20 <soichi> yes 07:07:56 <anil_rao> The main problem was getting it installed. The instructions provided seem to have a compatibility issue because -egg-info doesn't get created. 07:08:42 <anil_rao> If we unstack, replace the original horizon with the new one and add the TaaS patch and then stack. 07:09:51 <anil_rao> Seems like a DevStack problem perhaps. When I replaced the original horizon tree with the replacement containing the TaaS patch before the first stack.sh completed, everything worked fine. 07:10:16 <anil_rao> Just thought I'd mention it. 07:10:30 <soichi> Kaz, do you have any comments? 07:11:15 <kaz> I did not the reason yet. 07:11:32 <kaz> i did not find ... 07:11:41 <anil_rao> Its something that we can look into. No hurry. 07:11:48 <anil_rao> Another issue... 07:12:33 <reedip> anil_rao : maybe horizon is not pointing to the correct location in devstack 07:12:33 <anil_rao> When I try to create a tap-flow from the topology page, I see a warning indicating that the form didn't get posted correctly and to try again. However, there is also a success popup. 07:13:12 <anil_rao> The tap-flow gets created properly but I was wondering why that warning appears. It comes every time 07:13:27 <soichi> okay, we will check on our site. 07:13:39 <reedip> I will also try the patch tomorrow once 07:14:14 <anil_rao> soichi: This only happens when creating a tap-flow from the topology page. The warning is not there if I create a tap-flow from the tap-services tab. 07:14:39 <soichi> thank you for your information 07:15:38 <kaz> thank you , i will check. 07:15:54 <anil_rao> soichi,kaz: Thanks. 07:16:06 <reedip> anil_rao : one point for future , it seems ( I got some positive news from my own end as well ) , that the interest in TaaS is growing a bit 07:16:36 <reedip> if a lot of users and devs are interested in it, having a dedicated channel for discussion may be useful 07:16:52 <reedip> once it gathers enough momentum 07:18:16 <reedip> yamamot__ once the patch merges, can you also send a mail on the ML, so that everyone gets to know of the time change ( people may forget it by next week ...) 07:18:34 <yamamot__> it brings us back to the channel naming discussion :-) 07:18:36 <yamamot__> sure 07:20:23 <anil_rao_> How do we co-ordinate the merge of Takashi's patch that will remove network from tap-service-create? 07:20:24 <reedip> yamamot__ lets see the response in the Austin summit ( hoping for a good turnout) 07:20:45 <anil_rao_> The TaaS dashboard has a dependency... 07:20:51 <reedip> anil_rao_ : Let it merge! nd then gather up the pieces :) 07:21:04 <reedip> yamamot__ 's patch is more important 07:21:21 <reedip> because other wise a lot of dependant patches would break if that patch is not merged 07:21:21 <anil_rao_> OK 07:21:33 <yamamot__> is it difficult to update the dashboard not to pass the network? anyway no hurry. 07:22:05 <reedip> anil_rao_ , yamamot_ If the TaaS demo is working fine, then merging yamamot__ 's patch would not be an issue 07:22:30 <anil_rao_> reedip: I have already recorded the demo, so we are good for the Summit Presentation. 07:22:32 <yamamot__> sure, we don't need to use the latest master for demo 07:22:53 <soichi> yamamot_: we'd like to answer until next week 07:23:02 <reedip> so we can go ahead ... 07:24:06 <yamamot__> considering that dashboard needs a major work (ie rebase/rewrite for master) anyway, i guess removing network-id is trivial :-) 07:24:41 <soichi> agree 07:24:46 <reedip> 1minute 07:24:49 <reedip> and counting 07:25:15 <anil_rao_> The good thing is that the TaaS Dashboard doesn't ask the user to input the network id. 07:25:33 <yamamot__> anil_rao_: can you give me a permission to triage LP bugs? 07:25:44 <anil_rao_> reedip: We still have 5 min. 07:25:55 <reedip> damn my watch! 07:26:02 <anil_rao_> ;-) 07:26:12 <reedip> happens everytime, sorry 07:27:15 <anil_rao_> yamamot__: I am not sure what is the problem with the permissions. Let's ask Vinay 07:27:54 <reedip> Currently the owner of LP for Taas is Vinay 07:27:55 <anil_rao_> Any other issues that need discussion? 07:28:10 <reedip> it should ideally be a group which includes Vinay, Anil, Soichi and Yamamoto 07:28:13 <soichi> i have no topic 07:28:26 <yamamot__> nothing from me 07:28:31 <anil_rao_> reedip: We'll fix that 07:28:41 <reedip> I have started investigating the L2 migration of Taas to OVS , with some help from davidsha and the current QoS model 07:28:56 <reedip> just an update ^^ 07:28:58 <yamamot__> reedip: great 07:29:18 <anil_rao_> Nice 07:29:21 <reedip> and also working with rtheis for migration to Openstackclient ( migration would be finished sooner than expected) 07:29:34 <reedip> thats it from my end! 07:29:36 <anil_rao_> reedip: +1 07:29:58 <anil_rao_> soichi,kaz: You will be coming to Austin, right? 07:30:04 <soichi> yes 07:30:12 <kaz> yes 07:30:32 <anil_rao_> Look forward to seeing all of you there. 07:30:39 <soichi> sure 07:30:40 <anil_rao_> reedip: Get your visa soon. 07:30:49 <reedip> not in my hands :D 07:31:07 <reedip> wish it was :( 07:31:10 <anil_rao_> reedip: I hope it comes in time. 07:31:22 <yamamot__> anil_rao_: +1 07:31:26 <anil_rao_> okay we are out of time. talk to you all next week. 07:31:27 <reedip> I hope it comes before the presentation, worst case I would be travelling on 28th 07:31:36 <anil_rao_> #endmeeting