17:06:30 <boris-42> #startmeeting rally
17:06:31 <openstack> Meeting started Tue Jan 20 17:06:30 2015 UTC and is due to finish in 60 minutes.  The chair is boris-42. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:06:33 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:06:35 <openstack> The meeting name has been set to 'rally'
17:06:47 <boris-42> #topic VMTask benchmark updates and future steps
17:07:01 <boris-42> amaretskiy: could you please provide infomration related to this topic
17:07:07 <boris-42> waht we did what w are going to do
17:07:21 <amaretskiy> here is a patch that was recently merged: https://review.openstack.org/#/c/144337/
17:07:51 <amaretskiy> there are a lot of improvements for scenario VMTasks.boot_runcommand_delete
17:08:18 <amaretskiy> first main improvement is neutron support
17:08:40 <amaretskiy> second one in simple network configuration
17:09:05 <amaretskiy> configuration was a head ache before
17:09:46 <amaretskiy> so some configuration parameters were removed, other were auto-configured
17:10:40 <amaretskiy> so if we use network context (which support was also added with this patch), we can leave all parameters blank - the networking configuration will be auto-complete
17:10:47 <amaretskiy> eom
17:11:29 <amaretskiy> about future steps
17:12:01 <amaretskiy> now i'm going to switch to another part of rally - my next aim is html report
17:12:31 <amaretskiy> there are a lot of fixes/features planned for the report
17:12:41 <amaretskiy> eom
17:13:03 <boris-42> amaretskiy: so what kinds of bugs?
17:13:17 <boris-42> amaretskiy: and improvements would you like to make
17:13:23 <boris-42> could you be more verbose please
17:16:47 <boris-42> amaretskiy: so?
17:17:13 <amaretskiy> * failures aggregation (do not repeat similar errors)
17:17:13 <amaretskiy> * fix css/html layout for charts if there are a lot of parameters (especially for pie chart)
17:17:13 <amaretskiy> * show broken iterations with red stripes on stacked chart
17:17:13 <amaretskiy> * add pretty tooltips
17:17:14 <amaretskiy> * fix: add scroll bar for input task json (if too wide)
17:17:14 <amaretskiy> * add cluster information
17:17:14 <amaretskiy> * reimplement templates with jinja2
17:17:14 <amaretskiy> * etc...
17:17:58 <boris-42> amaretskiy: reimplemnt templates with jinja2 ???
17:17:59 <boris-42> what
17:18:00 <boris-42> ?
17:18:15 <amaretskiy> replace mako with jinja2
17:18:21 <boris-42> amaretskiy: ahh
17:18:22 <amaretskiy> and get rid of mako
17:18:25 <boris-42> amaretskiy: yep that is good one
17:18:30 <boris-42> amaretskiy: mako sux
17:18:37 <amaretskiy> yes :)
17:18:39 <boris-42> it was my bad to use it..
17:18:54 <amaretskiy> reimplementation will not take a lot of time
17:19:05 <boris-42> amaretskiy: yep they are quite close to each otehr
17:19:10 <boris-42> amaretskiy: just mako sux=)
17:19:11 <amaretskiy> :)
17:19:24 <boris-42> amaretskiy: so about tooltip
17:19:30 <boris-42> amaretskiy: wanna discuss?
17:19:48 <amaretskiy> i'm going to show hints with pretty js tooltips
17:20:06 <boris-42> amaretskiy: yep so we can just put basic infomration
17:20:11 <amaretskiy> but do not move charts/tables data into tooltips
17:20:13 <boris-42> what means iteration what is measured
17:20:19 <boris-42> amaretskiy: nonon
17:20:25 <boris-42> amaretskiy: just put extra info
17:20:29 <boris-42> amaretskiy: about what is on graphs
17:20:46 <boris-42> amaretskiy: cause not all users can understand when they see them first time
17:21:11 <amaretskiy> important note: tooltips implementation for charts is limited
17:21:35 <amaretskiy> i mean tooltips that appear when we put cursor over the chart
17:21:51 <amaretskiy> for other html objects - no problem
17:21:51 <boris-42> amaretskiy: no
17:21:55 <boris-42> amaretskiy: don't do that
17:22:04 <boris-42> amaretskiy: just put on top of charts on right
17:22:10 <boris-42> amaretskiy: some div with [?]
17:22:25 <boris-42> and if user moves over it it shows tooltip with explanation
17:22:26 <amaretskiy> ok, that is absolutely not a problem :)
17:22:35 <amaretskiy> tooltips will b every nice :)
17:22:37 <boris-42> amaretskiy: that will be great improvment (even if it simple)
17:22:55 <amaretskiy> okay :)
17:23:11 <boris-42> okay let's move on other things
17:23:22 <boris-42> #topic documentation refactoring
17:23:28 <boris-42> msdubov_: your turn
17:23:40 <msdubov_> boris-42 thanks
17:23:41 <boris-42> what we did, what are our plans
17:24:24 <msdubov_> So we have finally completed it: http://rally.readthedocs.org/en/latest/
17:24:45 <msdubov_> One of the greatest things about the new docs is this "step-by-step" Rally tutorial: http://rally.readthedocs.org/en/latest/tutorial.html
17:25:02 <msdubov_> It has to be extended a bit and that's the major plan for the future development of these docs
17:25:12 <msdubov_> e.g. stuff about different scenario runners is missing etc.
17:25:57 <msdubov_> eom
17:26:57 <boris-42> ьmsdubov_ sounds good
17:27:20 <boris-42> msdubov_: btw it will be nice to put extra info about .rally directory
17:27:41 <msdubov_> Yep I believe there will be lots of small changes as soon as we receive more feedback on these dosc
17:27:41 <boris-42> msdubov_: at least in tutorial about using multple clouds
17:28:04 <msdubov_> I also hope we will be able to get the RTD Sphinx theme in global requirementss
17:28:11 <boris-42> msdubov_: not sure lol
17:28:15 <msdubov_> To be able to review it
17:28:23 <boris-42> msdubov_: but redixin can make for us job
17:28:30 <boris-42> msdubov_: on Mirnatis Rally CI
17:28:44 <msdubov_> boris-42: redixin That would be nice!
17:29:01 <boris-42> okay
17:29:05 <boris-42> let's move
17:29:16 <boris-42> #topic Murano base
17:29:21 <boris-42> rvasilets: ^ hey
17:29:28 <boris-42> rvasilets: could you please provide update
17:29:38 <boris-42> I hope to get that patches ready and merge them=)
17:31:05 <rvasilets> Hi. For a long time I am working on adding Murano to Rally. There are a lot of issues. The most serious issue was that we don't wan't to download a big image from internet
17:31:11 <rvasilets> wicj murano app need
17:31:48 <boris-42> rvasilets: so did you find the way to avoid that issue and make something simple?
17:31:58 <rvasilets> Now I need to fix only unit tests for that patch and rebase it on the changes that was meged last wekk
17:32:00 <boris-42> rvasilets: like puting fake stuff
17:32:07 <rvasilets> Yes
17:32:11 <rvasilets> I found
17:32:19 <boris-42> rvasilets: that sounds great
17:32:55 <rvasilets> Now I have add 9Kb folder to Rally and creates everything temporary
17:33:35 <rvasilets> The key idea here is that this folder contains murano app that
17:33:40 <rvasilets> don't need image
17:33:53 <boris-42> кmsdubov_ щлфн)
17:33:57 <rvasilets> eom
17:34:00 <boris-42> rvasilets: yep 9kb directory is not issue
17:34:05 <boris-42> rvasilets: great work on this
17:34:22 <msdubov_> boris-42 wat? :)
17:34:28 <boris-42> msdubov_: ??
17:34:45 <boris-42> msdubov_: we need a 9kb directory in rally-jobs/extra
17:34:46 <msdubov_> boris-42: okay seems like you occasionally mentioned me
17:34:57 <boris-42> msdubov_: oh sory
17:35:12 <boris-42> #topic Making gate-rally-dsvm-cli publish ouptut to files
17:35:19 <boris-42> oanufriev: ^
17:35:21 <oanufriev> so
17:35:34 <oanufriev> there is a patch on review https://review.openstack.org/#/c/147098/
17:35:59 <boris-42> oanufriev: could you explain for everybody that idea and why we need it
17:36:32 <oanufriev> there implemented writting  and collecting reports of rally commands during functional tests
17:36:54 <oanufriev> then this reports become available in gerrit
17:38:05 <oanufriev> under gate-rally-dsvm-cli reference
17:38:48 <boris-42> oanufriev: btw I would like to see some manuals  )
17:39:23 <boris-42> openstack: probably in new readthedocs we should put explanation + in rally/tests/functional/README.rst
17:39:36 <oanufriev> so, functional tests become more informative. By the way, you can get this reports running functional tests local
17:39:41 <boris-42> oanufriev: otherwise newbies won't now about it
17:40:22 <boris-42> oanufriev: deff agree on that
17:40:43 <oanufriev> you should decorate your test method with test.functional.utils.gen_publish_path
17:40:48 <boris-42> oanufriev: without this feature we need to run locally in any case when there are some changes that requires documents
17:41:41 <oanufriev> and declare 'report_path' argument
17:42:36 <oanufriev> instance of ReportPath  will be passed to it
17:43:10 <boris-42> oanufriev I would prefer to see it documented
17:43:33 <oanufriev> ok. WBD
17:43:35 <boris-42> openstack: could you put some documentation on readme.rst in tests/functional/* ?
17:43:40 <boris-42> oanufriev: great
17:43:44 <oanufriev> eof
17:43:49 <boris-42> oanufriev: I will try to review your patch
17:43:55 <oanufriev> ok
17:44:02 <boris-42> msdubov_: amaretskiy rvasilets ^ please do the same guys
17:44:07 <boris-42> It's really cool stuf
17:44:53 <boris-42> #topic open disucssion
17:45:04 <boris-42> maybe somebody would like to discuss anything?
17:45:07 <rvasilets> no question
17:45:35 <amaretskiy> no questions
17:45:52 <msdubov_> msdubov_ no from me
17:45:53 <boris-42> msdubov_: oanufriev ?
17:46:10 <oanufriev> nop
17:46:14 <boris-42> okay seems like we can end meething and go home=)
17:46:22 <boris-42> Thank you guys and see you tomorrow=)
17:46:34 <boris-42> #endmeeting