07:33:12 <yujunz> #startmeeting qtip
07:33:12 <collabot> Meeting started Mon Aug  7 07:33:12 2017 UTC.  The chair is yujunz. Information about MeetBot at http://wiki.debian.org/MeetBot.
07:33:12 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
07:33:12 <collabot> The meeting name has been set to 'qtip'
07:33:50 <yujunz> #topic roll call
07:33:53 <yujunz> #info Yujun Zhang
07:33:56 <zhihui_wu> #info zhihui
07:34:00 <Sridhar> #info Sridhar
07:34:44 <yujunz> Today's agenda, Euphrates schedule, weekly update and demo from Sridhar
07:34:56 <yujunz> #topic Euphrates schedule
07:34:58 <yujunz> #link https://etherpad.opnfv.org/p/qtip-euphrates
07:35:16 <yujunz> As we all know, the schedule is postponed for two weeks.
07:35:25 <Sridhar> @yujunz, I'll share the update today. Next call, I'll provide the demo.
07:35:25 <collabot> Sridhar: Error: "yujunz," is not a valid command.
07:35:41 <yujunz> OK, Sridhar
07:37:57 <yujunz> In this release, we shall focus on storage performance indicator.
07:38:08 <yujunz> And other thing is keep compute testing up to date
07:38:38 <yujunz> Since the installer has updated, there are some work to be done to get the compute test suite working correctly.
07:39:03 <yujunz> #topic status update
07:39:26 <yujunz> My update, trying to get storage testing run on zte-pod4
07:39:43 <yujunz> #link https://build.opnfv.org/ci/job/qtip-storage-master/
07:40:01 <yujunz> Still failing, will try to fix it as soon as possible
07:40:35 <yujunz> That's it.
07:41:14 <Sridhar> My Updates: (1) Initiated Spirent virtual solutions deployment at ZTE pod.
07:42:02 <yujunz> I saw email from your colleague in Beijing, Sridhar
07:42:21 <Sridhar> (2) Found no overlap between QTIP network performance tests and other projects in OPNFV.
07:43:16 <Sridhar> (3) Working on network performance tests - starting with VM2VM. Need discussion on what should be "Indicator"
07:43:56 <Sridhar> For compute and storage - the algorithm for "indicator" is available?
07:44:38 <Sridhar> For network we can consider the performance 'relative' to the theoretical max and comeup with the indicators
07:45:23 <yujunz> For compute and storage, it is a comparison to baseline and the weighted average of various metrics
07:46:12 <Sridhar> Great, then we can stick to the same approach for network too - comparison to 'theoretical max' and weighted average of different metrics..
07:46:20 <yujunz> Theoretical max you mean the hardware limitation?
07:47:18 <Sridhar> Based on Link-Bandwidth and flow-configuration,
07:48:04 <yujunz> Make sense. So it means we need collect such information to know the theoretical max.
07:48:26 <Sridhar> Yes. The Traffic-Generators should do provide them
07:50:06 <Sridhar> I'm Sorry. In next call, I'll be more detailed.
07:50:25 <yujunz> OK. That would be good.
07:50:27 <Sridhar> I'll start documenting at the place you mentioned
07:50:33 <yujunz> Thanks
07:50:44 <Sridhar> Thats all from my side.
07:52:18 <yujunz> Anything else? zhihui_wu
07:54:27 <zhihui_wu> The storperf.sh only supports running on the host, I am in progress of making it works in qtip container.
07:55:32 <zhihui_wu> I will run the storperf containers in qtip container. It looks not a good way.
07:56:16 <zhihui_wu> But I don't find the better one now. Any suggestion is welcome~
07:56:48 <zhihui_wu> That's all
07:57:28 <yujunz> OK. Thanks.
07:57:54 <yujunz> #topic AoB
07:58:02 <yujunz> Anything else to discuss?
07:59:44 <yujunz> Thanks all. See you next week.
07:59:47 <yujunz> #endmeeting