17:14:22 <boris-42> #startmeeting rally
17:14:23 <openstack> Meeting started Tue Apr  8 17:14:22 2014 UTC and is due to finish in 60 minutes.  The chair is boris-42. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:14:24 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:14:24 <boris-42> hughsaunders ping
17:14:27 <openstack> The meeting name has been set to 'rally'
17:14:28 <boris-42> marcoemorais ping
17:14:34 <boris-42> aswadrangnekar  ping
17:14:38 <boris-42> eyerediskin ping
17:14:50 <boris-42> stannie tzabal  ping
17:15:20 <aswadrangnekar> boris-42 hello
17:15:24 <eyerediskin> sup
17:15:33 <boris-42> msdubov ping
17:15:41 <tzabal> here
17:15:53 <boris-42> #topic gates & functional tests
17:16:06 <boris-42> eyerediskin could you pls say what did you done?)
17:16:11 <boris-42> eyerediskin and what are current goals
17:16:13 <xwizard_> Hi there :)
17:16:26 <msdubov> boris-42 hi
17:16:46 <boris-42> xwizard_ hi =)
17:16:54 <boris-42> kun_huang hi
17:17:28 <kun_huang> boris-42 hi
17:17:39 <eyerediskin> currently i work on rally-gate-jobs blueprint
17:18:29 <eyerediskin> one job was added (rally-scenarios)
17:18:53 <boris-42> eyerediskin so could you pls elaborate what we are going to do on these gates
17:19:09 <boris-42> eyerediskin cause I am sure that not everybody knows=)
17:19:39 <eyerediskin> currently we have non-voting jobs which tests rally installation on ubuntu-12.04 and centos-6
17:19:50 <eyerediskin> this jobs is about to be voting soon
17:20:07 <tzabal> nice
17:20:23 <eyerediskin> also we have non-voting rally-scenarios job, which run all scenarios from samples directory
17:20:35 <boris-42> eyerediskin seems like it's voting?
17:20:50 <eyerediskin> no way
17:21:02 <kun_huang> eyerediskin it works now, all scenarios?
17:21:17 <eyerediskin> omg it is voting >_<
17:21:43 <marcoemorais> boris-42: pong
17:21:48 <eyerediskin> kun_huang: not all. currenly it skips large-ops scenarios
17:21:50 <boris-42> marcoemorais hey there meeting time
17:21:57 <boris-42> eyerediskin does it runs anything?
17:22:03 <boris-42> eyerediskin where are the results?)
17:22:09 <boris-42> eyerediskin or some logs?
17:22:32 <boris-42> eyerediskin cause actually my patch https://review.openstack.org/#/c/85732/ doesn't work properly..
17:22:46 <boris-42> eyerediskin and seems like check-rally-scenarios passed..
17:22:51 <eyerediskin> boris-42: im not sure it actually runs. i cant find any logs. im going to go deeper with this
17:23:20 <boris-42> eyerediskin did you ask somebody from infra?
17:24:15 <kun_huang> eyerediskin I think it could be cool that seeing scenarios output in log here http://logs.openstack.org/32/85732/4/check/check-rally-scenarios/62bb2ff/console.html
17:24:48 <msdubov> eyerediskin, boris-42 Btw how do we distinguish large-ops scenarios from other ones? Just manually set what should be skipped?
17:25:20 <boris-42> msdubov there is special "if" condition in copy paste lol
17:27:49 <boris-42> So out future steps are:
17:27:59 <boris-42> 1) add tests of CLI into rally-install jobs
17:28:12 <eyerediskin> boris-42: sorry some power outage was here
17:28:25 <boris-42> 2) fix scenario gate to run actually all benchmarks (except large-ops)
17:28:47 <boris-42> And we will think about how to test FakeCloud (so benchmarking rally with large-ops tests)
17:29:16 <boris-42> eyerediskin do you have anything to add ^
17:29:26 <boris-42> or does somebody has any questions?
17:29:44 <boris-42> btw marcoemorais ^ how is your patch? with preprocessing ?
17:29:56 <eyerediskin> tests for CLI are almost done
17:30:41 <eyerediskin> here is the patch: https://review.openstack.org/#/c/85738/ feel free to -1 %)
17:30:49 <marcoemorais> boris-42: patch is not working
17:31:06 <boris-42> marcoemorais what patch?
17:31:42 <marcoemorais> boris-42: preprocessing patch that we discussed on friday
17:31:57 <boris-42> marcoemorais could you share what you have
17:31:58 <boris-42> marcoemorais WIP
17:32:10 <boris-42> marcoemorais I would like to take a look probably I'll be able to help a bit with it
17:32:30 <marcoemorais> boris-42: ok
17:32:38 <boris-42> marcoemorais thanks
17:33:14 <boris-42> #topic rally info command
17:33:29 <boris-42> msdubov please could you share your ideas with us?
17:34:02 <msdubov> boris-42 yep
17:34:17 <msdubov> I've prepared a doc https://docs.google.com/document/d/1MUufsxGYGc7GTmaN_GogbpVR_sXds6VSI6K3R9ybEmY/edit?usp=sharing
17:34:40 <msdubov> One important point to decide is what should be the actual syntax
17:34:55 <msdubov> So pls take a look at this doc and tell what you think
17:35:03 <msdubov> I present there two possibilities for that
17:35:36 <boris-42> msdubov pls allow commeitng
17:35:40 <boris-42> msdubov for everybody
17:35:45 <boris-42> msdubov cause how can I comment?
17:35:55 <msdubov> boris-42 done
17:39:41 <boris-42> msdubov nice doc
17:40:04 <boris-42> hughsaunders marcoemorais eyerediskin thoughths ?)
17:40:09 <boris-42> tzabal ^
17:41:27 <boris-42> anybody?)
17:42:02 <msdubov> boris-42 hughsaunders tzabal eyerediskin  What would you prefer? "Sructured" syntax or the "query-based" one?
17:42:02 <marcoemorais> boris-42: seems like this is a souped-up version of pydoc?
17:42:09 <tzabal> boris-42 for the work of msdubov, i like better the query based mode, but also structured is good too
17:42:39 <marcoemorais> msdubov: user would use rally info XXX to copy-paste stuff out of the docstrings?
17:42:53 <boris-42> marcoemorais actually to understand
17:42:58 <boris-42> marcoemorais what it has in rally
17:43:12 <msdubov> marcoemorais, e.g. you see all those sample configs in /doc
17:43:20 <boris-42> marcoemorais "rally info context" returns short explanation of what the hell is it
17:43:29 <boris-42> how to use it and what context you have
17:43:37 <msdubov> marcoemorais, You can then just call "rally info ExistingCloud" to understand what that means
17:44:05 <boris-42> marcoemorais such kind of smart explorer that allows you to avoid diving into the code
17:44:55 <marcoemorais> ok I get it
17:45:42 <boris-42> marcoemorais btw I think that query one is good but we should exited "search" stuff
17:45:47 <boris-42> msdubov ^
17:46:15 <boris-42> okay I think everybody agree that it is missing stuff?)
17:46:19 <msdubov> boris-42 What do you mean by "exited"?
17:46:27 <boris-42> extend*
17:46:45 <msdubov> boris-42 So that it also will handle missspelling or smth else?
17:47:11 <boris-42> msdubov to allow to find by part of name info that you are looking for
17:47:21 <msdubov> boris-42 Agree
17:47:22 <boris-42> msdubov "nova benchmarks"
17:47:26 <boris-42> msdubov like that=)
17:47:36 <msdubov> boris-42 But this will make the implementation a bit more involved, as I wrote in the doc
17:47:40 <marcoemorais> msdubov: docstrings are usually for developers so that is where I would hesitate… is there going to be a way to suppress certain parts of the docstrings from being shown to users?
17:47:43 <msdubov> boris-42 But I also think it should be there
17:48:08 <msdubov> marcoemorais, I'm actually not sure we will have very complex docstrings...
17:48:11 <boris-42> msdubov yep I think it's not so complicated to implement
17:48:18 <boris-42> marcoemorais ^
17:48:25 <msdubov> marcoemorais, But in any case there is always this "first line" of docstrings
17:48:31 <msdubov> marcoemorais, Which is a short summary
17:48:56 <boris-42> marcoemorais first line could be used in table when we have multiple results
17:49:05 <boris-42> msdubov e.g. "nova"
17:49:08 <boris-42> marcoemorais ^
17:49:31 <marcoemorais> msdubov boris-42: let us look at an example https://github.com/stackforge/rally/blob/master/rally/benchmark/scenarios/nova/servers.py#L44
17:49:55 <marcoemorais> msdubov boris-42  $ rally info NovaServers
17:49:55 <marcoemorais> emits: Tests booting an image and then listing servers.
17:50:12 <boris-42> marcoemorais agree not perfect title..
17:50:13 <marcoemorais> correction: rally info NovaServers.boot_and_delete
17:50:33 <msdubov> boot_and_list_server?
17:50:43 <marcoemorais> msdubov: yes
17:50:54 <boris-42> marcoemorais in case of rally info NovaServers
17:51:05 <boris-42> it should show dosctrings of NovaServers
17:51:17 <boris-42> + probably list of all benchmarks
17:51:21 <boris-42> in NovaServers
17:51:51 <tzabal> so on this example a new docstring for the class?
17:51:57 <msdubov> tzabal, Yes
17:52:11 <tzabal> msdubov ok
17:52:18 <msdubov> tzabal, The work on this blueprint will actually involve rewriting docstrings/adding new ones
17:52:38 <msdubov> tzabal, But not only docstring
17:52:48 <tzabal> msdubov this is very good
17:52:53 <msdubov> tzabal, We will have to retrieve a list of benchmark scenarios
17:53:03 <msdubov> tzabal, which are marked by @base.scenario()
17:53:09 <boris-42> msdubov there is already method for that
17:53:29 <msdubov> boris-42, great
17:53:32 <marcoemorais> msdubov boris-42: one thing that I think might be useful is if we emitted a prepopulated config on standard out.. let me exaplain
17:53:32 <marcoemorais> $ rally info NovaServers. boot_and_delete_server  —template-config
17:53:32 <marcoemorais> { "NovaServers.boot_and_delete_server": [ … }
17:53:46 <boris-42> msdubov yep that one imho is great
17:53:50 <boris-42> marcoemorais **
17:54:01 <boris-42> marcoemorais btw it can be done via doc/sampels
17:54:06 <msdubov> boris-42, marcoemorais Agree, will add to the doc
17:54:22 <boris-42> marcoemorais so just show from sample stuff
17:55:20 <marcoemorais> boris-42: typically I have to edit the scenario config and then keep those edited scenarios somewhere else; if we could emit the scenario config as templates to be edited by the user (like a Vagrantfile) that would be good
17:56:38 <boris-42> marcoemorais you may propose some syntax for that?
17:58:12 <boris-42> seems like that time is up
17:58:21 <boris-42> we should move to rally chat
17:58:33 <boris-42> #endmeeting