05:40:16 <soichi> #startmeeting taas
05:40:17 <openstack> Meeting started Wed Aug 24 05:40:16 2016 UTC and is due to finish in 60 minutes.  The chair is soichi. Information about MeetBot at http://wiki.debian.org/MeetBot.
05:40:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
05:40:21 <openstack> The meeting name has been set to 'taas'
05:40:24 <soichi> hi
05:40:28 <kaz> hi
05:40:41 <yamamoto> hi
05:40:55 <soichi> #topic: Open Discussion
05:41:16 <kaz> #link https://review.openstack.org/#/c/324367/
05:41:30 <kaz> yamamoto: Thank you for your comment for Patch Set 6.
05:41:47 <kaz> But Jenkins put error messages, So I don't know how to treat this issue.
05:42:28 <yamamoto> it's complaining about migration vs model mismatch. you should fix either of them.
05:43:33 <yamamoto> here's some explanation about the error messages
05:43:35 <yamamoto> #link https://github.com/openstack/neutron/blob/master/neutron/tests/functional/db/test_migrations.py#L82
05:45:23 <yamamoto> i guess you need to put server_default to models.
05:48:30 <kaz> thank you. i will check it.
05:49:26 <soichi> kaz: i think it is better to leave a message so that anil_rao can know our current status about performance measurement
05:49:55 <kaz> I'm trying to measure the performance of taas when increasing the source VMs of iperf.
05:50:06 <kaz> But, i found a mistake in the measuring procedure.
05:50:17 <kaz> I used the '-t' option of iperf which set transmit time in seconds.
05:50:26 <kaz> In this case, when the source VMs are increasing, it could't stop transmit at the exact time.
05:50:41 <kaz> So, I will measure by another method.
05:51:47 <kaz> I would like to share the result next week.
05:52:58 <reedip> hi
05:53:11 <soichi> we are also planning to measure in case of pairs of comjunication VAs are increased
05:53:15 <soichi> reedip: hi
05:53:54 <reedip> I saw the model ve migration mismatch
05:53:55 <soichi> comjunication -> communication, VAs -> VMs
05:55:03 <soichi> reedip: can you give us any advice about the mismatch?
05:56:18 <reedip> No, I am still a bit confused. I am seeing yamamoto's advice right now
05:56:34 <soichi> okay
05:58:51 <soichi> yamamoto, reedip: do you have any update or topics for today's meeting?
05:59:55 <yamamoto> i want to ping reviewers about spec https://review.openstack.org/#/c/256210/
06:00:05 <yamamoto> i don't see any reason to keep it unmerged
06:00:12 <reedip> soichi : I shared the devstack error with anil_rao last week. TaaS is not working in the single node devstack ( not sure why )
06:00:58 <soichi> yamamoto: okay, i will take a look it
06:01:27 <reedip> yamamoto : The TaaS spec right???  Yeah, that has been stalled for no reason. Can you send a message to the Neutron Core teams to help in its review and inclusion of the same?
06:02:19 <yamamoto> reedip: taas is not neutron subproject.  i don't see why it needs neutron core team involved.
06:02:56 <reedip> yamamoto : Then what should we be waiting for ?
06:03:13 <yamamoto> reedip: nothing i'm aware of.
06:03:20 <reedip> hehehe
06:03:33 <reedip> yamamoto : then should we contact Infra team for the same ?
06:03:46 <yamamoto> ?
06:04:07 <reedip> for inclusion of Tap into the Big tent ?
06:04:28 <yamamoto> i guess it's a separate topic.
06:05:02 <yamamoto> i have no strong opinion either ways (big tent or neutron)
06:06:47 <yamamoto> speaking about stadium, armando is working on updating guide
06:06:48 <yamamoto> #link https://review.openstack.org/#/c/335739/
06:06:56 <reedip> okay
06:07:31 <yamamoto> anyway having the spec in the repo for now shouldn't hurt either options
06:08:09 <soichi> yamamoto: agree
06:08:19 <reedip> yes
06:10:05 <yamamoto> ta
06:10:08 <yamamoto> that's all from me
06:11:44 <yamamoto> (ta is just a mistype)
06:12:06 <soichi> reedip: kaz has a question
06:12:18 <reedip> ??
06:12:51 <kaz> reedip: what is a devstack error?
06:13:10 <reedip> when I launch devstack , kaz, then devstack launch fails
06:13:44 <reedip> the reason for the  failure is http://paste.openstack.org/show/560712/
06:13:53 <reedip> The devstack/local.conf has the following parameters for taas:
06:13:53 <reedip> enable_service taas
06:13:53 <reedip> enable_service taas_openvswitch_agent\
06:13:53 <reedip> enable_plugin tap-as-a-service https://github.com/openstack/tap-as-a-service
06:14:31 <kaz> is this about OpenFlow version mismatch ?
06:15:58 <yamamoto> reedip: taas ini is not specified for neutron server, right?
06:17:04 <reedip> taas.ini ?
06:17:08 <reedip> no I dont think so
06:17:19 <reedip> kaz : I am not sure if that would be the problem
06:17:21 <yamamoto> taas_plugin.ini
06:17:33 <reedip> can you tell me how to decode this issue?
06:17:56 <yamamoto> decode?
06:18:18 <reedip> yamamoto : decode -> understand and resolve :)
06:18:42 <yamamoto> there's a related patch
06:18:44 <yamamoto> #link https://review.openstack.org/#/c/342469/
06:19:29 <yamamoto> and my effort to clean up things
06:19:31 <yamamoto> #link https://review.openstack.org/#/c/342493/
06:19:56 <reedip> yamamoto: taas_plugin.ini is there in the system but I dont think it is mentioned in the neutron.conf
06:20:24 <yamamoto> i meant neutron server's command line argument
06:20:38 <yamamoto> --config-file /etc/neutron/taas_plugin.ini or whatever
06:21:27 <reedip> yamamoto: shouldnt devstack do it automatically?
06:21:48 <reedip> I mean when I am trying to launch devstack, shouldnt it pickup taas_plugin.ini for Neutron Server?
06:22:52 <kaz> there are configuratins in ./devstack/README.rst
06:22:52 <yamamoto> if you tell it so appropriately, yes.  but for some combination of adv services, it's unintuitive at best.
06:24:00 <reedip> Oh ok , let me check that once ... by the way how have you all enabled taas ?
06:24:07 <kaz> Q_PLUGIN_EXTRA_CONF_FILES=(taas_plugin.ini)
06:24:16 <reedip> okay
06:27:28 <soichi> if we have no more topics, i'd like to close today's meeting
06:27:38 <yamamoto> nothing from me
06:27:50 <kaz> i have no more topics
06:28:19 <soichi> okay, thank you. see you next week
06:28:21 <soichi> #endmeeting