05:30:09 <anil_rao> #startmeeting taas
05:30:10 <openstack> Meeting started Wed May 11 05:30:09 2016 UTC and is due to finish in 60 minutes.  The chair is anil_rao. Information about MeetBot at http://wiki.debian.org/MeetBot.
05:30:11 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
05:30:13 <openstack> The meeting name has been set to 'taas'
05:30:18 <yamamot__> hi
05:30:20 <kaz> hi
05:30:21 <soichi> hi
05:30:21 <anil_rao> Hi
05:31:40 <anil_rao> Hope everyone who attended the summit had a safe trip back.
05:32:11 <soichi> thank you for meet up in the summit
05:33:05 <anil_rao> The only item in today's agenda is the gate blocking bug.
05:33:27 <yamamot__> #link https://wiki.openstack.org/wiki/Meetings/taas
05:34:21 <anil_rao> I complete the review shortly.
05:34:26 <yamamot__> #link https://review.openstack.org/#/c/306390/
05:35:26 <anil_rao> Any other items folks want to discuss?
05:35:49 <yamamot__> any update on lp permission tweak?
05:36:19 <anil_rao> Not yet. Waiting for Vinay on this. Not sure what is going on
05:37:49 <anil_rao> we had talked in Austin about me putting up the TaaS Driver detailed design doc.
05:38:02 <anil_rao> Where is a good place to park such a document?
05:38:28 <yamamot__> doc/ in the repo
05:39:21 <yamamot__> incase it's inappropriate for some reason, maybe put on wiki and have a link there.
05:39:54 <anil_rao> Ok.
05:40:30 <yamamot__> common flow classifier meeting will start on may 17
05:40:35 <yamamot__> #link http://lists.openstack.org/pipermail/openstack-dev/2016-May/094538.html
05:40:53 <anil_rao> Thanks for the reminder.
05:41:23 <yamamot__> and neutron stadium
05:41:25 <yamamot__> #link https://review.openstack.org/#/c/312199/
05:43:51 <anil_rao> Looks like lots of activity in the Neutron Statdium Evolution review
05:45:40 <anil_rao> I'll look into adding some error handling code in the TaaS driver.
05:46:06 <anil_rao> We will also need to design the status field for some of the TaaS API calls.
05:47:03 <soichi> anil_rao: agree (to support async API)
05:47:21 <anil_rao> soichi: Yes.
05:51:16 <kaz> I can try to this.
05:52:03 <anil_rao> kaz: Sure.
05:52:59 <anil_rao> kaz: The first step would be to just add the status field in the API and just set it to success by default.
05:53:17 <anil_rao> When the driver error handling is completed we can properly update the status field.
05:53:18 <kaz> i see
05:53:45 <soichi> anil_rao: +1
05:56:31 <anil_rao> I am headed back to Texas (Dallas) next week for a couple of days. When I get back I'll set up a h/w based DevStack environment and collect some performance data for TaaS.
05:56:45 <anil_rao> Will share the results on the mailing list.
05:58:08 <soichi> anil_rao: that's sounds great. we really need performance data.
05:58:48 <kaz> anil_rao: +1
06:01:58 <soichi> in our experiment, we found that the virt-io running on a host can be a bottleneck
06:06:09 <anil_rao> Sorry, compute problem
06:06:48 <anil_rao> soichi: virt-io?
06:08:58 <soichi> virt io process running on a host seems to comsume CPU 100%
06:09:30 <soichi> comsume -> consume
06:09:36 <anil_rao> soichi: I see.
06:10:12 <anil_rao> soichi: What kind of load were you putting on the VM being monitored.
06:10:56 <yamamot__> soichi: it isn't taas-specific, right?
06:12:01 <soichi> yamamotot_: you are right. it isn't taas-specific.
06:13:19 <anil_rao> I am still wondering what kind of load was resulting in virt-io consuming 100% cpu
06:13:39 <soichi> anil_rao: we put load by using iperf
06:14:25 <yamamot__> small frames i guess
06:14:41 <anil_rao> soichi: I guess iperf is measuring the max speed so it will push the system until it maxes out.
06:15:32 <anil_rao> soichi: Was this on hardware based DevStack? Multi-node?
06:15:46 <kaz> i guess so.
06:16:48 <kaz> anil_rao: yes
06:16:58 <anil_rao> kaz: Thanks.
06:18:07 <soichi> anil_rao: we would like to share our result on next week
06:18:25 <anil_rao> soichi: That would be nice.
06:18:48 <anil_rao> However, I will have to skip next week's meeting since I am going to Texax on a business trip.
06:18:58 <anil_rao> I can read up on the meeting minutes.
06:19:10 <soichi> okay
06:21:10 <soichi> i have no additional topic for today.
06:21:28 <anil_rao> Anyone else has any updates?
06:21:38 <yamamot__> nothing from me
06:21:48 <kaz> me too
06:21:59 <anil_rao> Ok, I think we can end for today then.
06:23:21 <anil_rao> #endmeeting