Perl 6 - the future is here, just unevenly distributed

IRC log for #openstack-rally, 2014-12-29

| Channels | #openstack-rally index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:26 dmellado_ joined #openstack-rally
02:40 yaguang joined #openstack-rally
02:46 openstackgerrit Li Yingjun proposed stackforge/rally: Move files to common lib(Part 3)  https://review.openstack.org/143938
03:24 rvcehimanshu joined #openstack-rally
03:35 nkhare joined #openstack-rally
03:37 yaguang joined #openstack-rally
04:25 yaguang joined #openstack-rally
04:30 dmellado_ joined #openstack-rally
04:38 dmellado joined #openstack-rally
04:41 dmellado_ joined #openstack-rally
04:58 rvcehimanshu joined #openstack-rally
05:17 openstackgerrit li,chen proposed stackforge/rally: Use string.ascii_lowercase and string.ascii_uppercase only  https://review.openstack.org/144255
05:17 rdas joined #openstack-rally
05:56 andreykurilin joined #openstack-rally
05:56 rvasilets joined #openstack-rally
06:18 neeti joined #openstack-rally
07:02 rdas joined #openstack-rally
07:16 neeti joined #openstack-rally
07:20 yfried joined #openstack-rally
07:28 nmagnezi joined #openstack-rally
07:34 pcaruana joined #openstack-rally
08:00 andreykurilin1 joined #openstack-rally
08:01 gema joined #openstack-rally
08:47 rvcehimanshu left #openstack-rally
09:03 k4n0 joined #openstack-rally
09:12 amaretskiy joined #openstack-rally
09:12 nmagnezi boris-42, hello :)
09:13 boris-42 nmagnezi: hi there
09:13 nmagnezi boris-42, hi Boris, How are things?
09:15 boris-42 nmagnezi: good
09:15 boris-42 nmagnezi: you?
09:17 nmagnezi boris-42, great
09:17 nmagnezi boris-42, started to work on my patch and i can use some insights from you, in regards to implementation
09:17 nmagnezi boris-42, (for ipv6 tests)
09:17 nmagnezi boris-42, if that's ok :)_
09:23 andreykurilin joined #openstack-rally
09:29 tkammer joined #openstack-rally
09:32 nmagnezi boris-42, ?
09:32 boris-42 nmagnezi: ah sorry*
09:32 boris-42 nmagnezi: sure just ask me
09:33 nmagnezi boris-42, cool. so i was looking at rally/rally/benchmark/scenarios/neutron/utils.py
09:33 nmagnezi boris-42, the NeutronScenario Class
09:34 nmagnezi boris-42, if i want to implement IPv6 tests i would either need to add tests or modify the subnet creation tests
09:34 nmagnezi boris-42, plus params such as SUBNET_IP_VERSION
09:34 boris-42 nmagnezi: could you just try to implement some DRAFT
09:34 boris-42 nmagnezi: even without tests
09:35 boris-42 nmagnezi: and we will continue discussion in review.openstack.org
09:35 nmagnezi boris-42, np
09:35 boris-42 nmagnezi: change whatever you need
09:35 boris-42 nmagnezi: it's okay=)
09:35 nmagnezi boris-42, but that's a question that might influence my implementation
09:35 boris-42 nmagnezi: to change what you need=)
09:35 boris-42 nmagnezi: ok let's include amaretskiy
09:36 nmagnezi boris-42, should I just inherit from this class and make ipv6 tests? or just modify the existing one?
09:36 amaretskiy hi
09:36 boris-42 nmagnezi: amaretskiy is refactoring currently network part
09:36 nmagnezi amaretskiy, hi there
09:36 amaretskiy hi nmagnezi
09:37 nmagnezi amaretskiy, so context: i want to implement neutron ipv6 tests
09:37 amaretskiy nmagnezi pleas etake a look at our new rally.benchmark.wrappers.network.NeutronWrapper
09:37 nmagnezi amaretskiy, in neutron ipv6 is mostly reflected (configuration wise) in subnet creation
09:38 amaretskiy most routines from scenarios.neutron.utils should call wrapper
09:38 * nmagnezi looking
09:38 amaretskiy so current NeutronScenario should be refactored
09:39 amaretskiy NeutronWrapper is a place where we have all neutron routines that we can use anywhere (scenarios, context, etc..), witj unified interface with nova-natwork
09:40 amaretskiy *nova-network
09:40 nmagnezi amaretskiy, i do not know if nova-network has ipv6 support
09:41 amaretskiy this wrapper is actually in further development so merge conflicts are possible
09:41 amaretskiy that is no matter if there is an ipv6 in nova, but unified interface does matter
09:41 nmagnezi amaretskiy, but how does this wrapper related to https://github.com/stackforge/rally/blob/master/rally/benchmark/scenarios/neutron/utils.py#L25 exactly?
09:42 amaretskiy NeutronScenario, for example, has method _create_network()
09:43 amaretskiy this method should call wrappers.network.NeutronWrapper.create_network()
09:43 amaretskiy an so on...
09:43 pcaruana joined #openstack-rally
09:44 amaretskiy we create networks not only in scenarios, but also in contexts
09:44 amaretskiy so we should use wrapper everywhere
09:45 amaretskiy nmagnezi take a look at https://github.com/stackforge/rally/blob/master/rally/benchmark/context/network.py#L64
09:45 amaretskiy this works well both with nova-network and neutron
09:46 amaretskiy so 1) use wrapper routines in any part of rally 2) unified interface solves problem of different networking implementations (minimal of specific code)
09:48 amaretskiy so, resume: 1) you should refactor existing code and use wrapper 2) improve the wrapper (ipv6 support)
09:48 exploreshaifali joined #openstack-rally
09:50 nmagnezi amaretskiy, ack, yet there are features supported with neutron that are not supported in nova-network
09:50 amaretskiy just make nova-network part silent
09:50 nmagnezi amaretskiy, for instance, ipv6 have 3 modes (stateful, stateless, slaac). yet nova-network only supports 1 of them
09:50 amaretskiy please submit the patch set - we will make a review :)
09:51 nmagnezi amaretskiy, ofcourse, just asking since it's my first (future) patch in rally
09:52 amaretskiy nmagnezi the idea is unified interface, so there are several choices: 1) silent 2) NotImplementedError 3) never called (methods are specific and called from neutron scenarios)
09:52 amaretskiy great :)
09:53 amaretskiy as a trick, you can put specific arguments into **kwargs
09:54 amaretskiy https://github.com/stackforge/rally/blob/master/rally/benchmark/wrappers/network.py#L94
10:01 andreykurilin joined #openstack-rally
10:03 tkammer joined #openstack-rally
10:41 oanufriev joined #openstack-rally
10:46 vkmc joined #openstack-rally
12:17 pboldin joined #openstack-rally
12:22 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the flavor context  https://review.openstack.org/141670
12:22 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the generic script-based benchmark for the VMs  https://review.openstack.org/141671
12:22 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the context benchmark_image  https://review.openstack.org/138466
12:22 openstackgerrit Pavel Boldin proposed stackforge/rally: Tests for rally gates  https://review.openstack.org/143469
12:32 pboldin1 joined #openstack-rally
12:42 openstackgerrit li,chen proposed stackforge/rally: Use string.ascii_lowercase and string.ascii_uppercase only  https://review.openstack.org/144255
12:43 pcaruana joined #openstack-rally
12:46 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the flavor context  https://review.openstack.org/141670
12:46 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the generic script-based benchmark for the VMs  https://review.openstack.org/141671
12:46 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the context benchmark_image  https://review.openstack.org/138466
12:46 openstackgerrit Pavel Boldin proposed stackforge/rally: Tests for rally gates  https://review.openstack.org/143469
12:58 yfried joined #openstack-rally
12:58 yfried joined #openstack-rally
13:02 tkammer joined #openstack-rally
13:03 openstackgerrit joined #openstack-rally
13:09 openstackgerrit Nikita Konovalov proposed stackforge/rally: [Sahara] make job executions delete syncronized  https://review.openstack.org/144297
13:11 jaypipes joined #openstack-rally
13:33 pboldin joined #openstack-rally
13:48 dspano joined #openstack-rally
13:49 pboldin joined #openstack-rally
13:50 openstackgerrit Nikita Konovalov proposed stackforge/rally: [Sahara] refactoring _launch_cluster  https://review.openstack.org/143089
13:59 tkammer joined #openstack-rally
14:34 openstackgerrit Nikita Konovalov proposed stackforge/rally: Neutron network context support  https://review.openstack.org/143671
14:48 openstackgerrit Nikita Konovalov proposed stackforge/rally: [Sahara] refactoring _launch_cluster  https://review.openstack.org/143089
15:05 yfried_ joined #openstack-rally
15:10 openstackgerrit Nikita Konovalov proposed stackforge/rally: Neutron network context support  https://review.openstack.org/143671
15:22 openstackgerrit Nikita Konovalov proposed stackforge/rally: [Sahara] refactoring _launch_cluster  https://review.openstack.org/143089
15:44 openstackgerrit Merged stackforge/rally: [Sahara] make job executions delete syncronized  https://review.openstack.org/144297
15:57 yfried_ joined #openstack-rally
16:03 boris-42 dspano: hey there are you around?
16:07 openstackgerrit Dave Spano proposed stackforge/rally: Move CinderVolumes tasks from rally.yaml to rally-neutron.yaml  https://review.openstack.org/144183
16:08 dspano boris-42: I am now. :)
16:11 openstackgerrit Anastasia Kuznetsova proposed stackforge/rally: Add Mistral benchmark  https://review.openstack.org/144320
16:12 boris-42 dspano: hey hey
16:12 boris-42 dspano: there is small nit in your patch
16:12 boris-42 dspano: one extra space lol=)
16:12 boris-42 dspano: that breaks everything
16:12 boris-42 dspano: could you update it?)
16:13 msdubov joined #openstack-rally
16:13 dspano boris-42: Yeah, I saw your email before I got on. I amended it a few minutes ago.
16:15 boris-42 dspano: thanks=)
16:19 openstackgerrit Anastasia Kuznetsova proposed stackforge/rally: Add Mistral benchmark  https://review.openstack.org/144320
16:25 openstackgerrit Anastasia Kuznetsova proposed stackforge/rally: Add Mistral benchmark  https://review.openstack.org/144320
16:37 msdubov joined #openstack-rally
16:37 openstackgerrit Anastasia Kuznetsova proposed stackforge/rally: Add Mistral benchmark  https://review.openstack.org/144320
17:02 openstackgerrit Anastasia Kuznetsova proposed stackforge/rally: Add Mistral benchmark  https://review.openstack.org/144320
17:26 openstackgerrit Roman Vasilets proposed stackforge/rally: Add murano base for murano benchmarks  https://review.openstack.org/137650
17:26 jlk boris-42: I'm working today though!
17:29 openstackgerrit Roman Vasilets proposed stackforge/rally: Add murano base for murano benchmarks  https://review.openstack.org/137650
17:31 tkammer joined #openstack-rally
17:38 exploreshaifali joined #openstack-rally
17:38 shaifali_ joined #openstack-rally
17:39 exploreshaifali joined #openstack-rally
17:52 openstackgerrit Roman Vasilets proposed stackforge/rally: Add murano benchmarks  https://review.openstack.org/137661
17:53 openstackgerrit Roman Vasilets proposed stackforge/rally: Add murano base for murano benchmarks  https://review.openstack.org/137650
18:09 boris-42 jlk: I hope you enjoy holidays?)
18:10 jlk boris-42: well enough. Did some hacking, did a lot of pleasure reading, did way too much eating.
18:11 boris-42 jlk:  )
18:11 openstackgerrit Alexander Maretskiy proposed stackforge/rally: (WIP) Improvements for scenarion VMTasks.boot_runcommand_delete  https://review.openstack.org/144337
18:14 jlk boris-42: I took your advice and looked at removing the entire block of code that tries to figure out end points and ports from config, and things work a whole lot better. auth_url is the only thing keystone client code needs, it figures the rest out from the service catalog. Forcing endpoint is bad, it blocks the discovery. So I'm going to do a PR that removes that hunk of code and any tests around it.
18:16 amaretskiy joined #openstack-rally
18:22 yfried_ joined #openstack-rally
18:30 NikitaKonovalov joined #openstack-rally
18:31 pboldin joined #openstack-rally
18:39 andreykurilin joined #openstack-rally
18:50 boris-42 jlk: actually I think that that is the right direction+)
18:50 boris-42 jlk: btw what about your another patch?
18:51 boris-42 jlk:  https://review.openstack.org/#/c/140150/
18:51 jlk It may obviate that patch
18:51 jlk wait no
18:51 jlk sorry.
18:51 jlk Yeah, the RPS, you want more test cases around it yah? That's something I can work on
18:52 jlk I'm more concerned with the ports thing though as it means we have to fork for now
18:53 boris-42 jlk: so actually those tests are super easy
18:54 boris-42 jlk: I mean you should change some .yaml
18:54 boris-42 jlk: and a bit of unittests=)
18:54 jlk nod
18:55 boris-42 jlk: but I agree that removing that shit from osclients is quite important stuff
18:55 boris-42 =)
18:56 boris-42 jlk: heh spend whole day to setup new laptop
18:56 boris-42 jlk: still in the middle of process=)
19:11 pboldin joined #openstack-rally
19:15 jlk haha, so many tests around this keystone stuffs.
19:17 boris-42 jlk: =))
19:29 nmagnezi joined #openstack-rally
20:47 eye_ joined #openstack-rally
21:10 openstackgerrit Oleh Anufriiev proposed stackforge/rally: WIP Report generation improvement  https://review.openstack.org/143866
21:16 dspano joined #openstack-rally
21:21 openstackgerrit Jesse Keating proposed stackforge/rally: Don't hard set keystone endpoint  https://review.openstack.org/144366
21:22 jlk boris-42: https://review.openstack.org/144366  :)
21:23 andreykurilin joined #openstack-rally
21:28 boris-42 andreykurilin1: ^
21:28 jlk oops, that's going to fail
21:28 boris-42 jlk: I'll need to test with FUEL team this stuff
21:29 jlk found some more admin_port references
21:29 boris-42 jlk: otherwise they can fired me=)
21:29 jlk haha
21:29 jlk yup]
21:30 boris-42 jlk: but I definitely would like to remove that crap=)
21:33 jlk oh interesting
21:33 jlk looks like existing deployments would break with this, because the database has admin_port in it
21:34 jlk so maybe not change the signature of the endpoint object, but just don't do anything with admin_port
21:37 boris-42 jlk: I think we can get rid of it
21:37 boris-42 jlk: and change everything
21:37 boris-42 jlk: it will be a bit non backward compatible change
21:37 jlk yeah
21:37 boris-42 jlk: but personally I don't care=)
21:38 jlk I don't care... much
21:38 jlk I can always re-do my deployments
21:38 boris-42 jlk:  I really would like to make normal code
21:38 jlk okay
21:38 jlk so I'll leave the PR as is, and just fiddle my deployment
21:38 boris-42 jlk: I think everybody is redeploying stuff
21:40 boris-42 jlk: but I have to change this stuff soon=)
21:40 boris-42 jlk: there are going to be few more big changes in framework
21:40 boris-42 and I will start efforts related to stabilization
21:40 jlk nod
21:41 jlk alright, with a re-dene existing deployment, works fine.
21:42 jlk We could do a database migration for this, that drops the admin_port key.....  :D
21:42 boris-42 jlk: lol
21:43 boris-42 jlk: database migration
21:43 boris-42 jlk: lol
21:43 boris-42 =)
21:43 jlk it's the OpenStack Way!
21:43 boris-42 jlk: fuck OpenStack Way
21:43 boris-42 jlk: OpenStack Way made me make Rally
21:43 boris-42 =)
21:44 boris-42 jlk: change it here https://github.com/stackforge/rally/blob/master/rally/objects/endpoint.py#L34
21:44 jlk and there was much rejoicing.
21:44 boris-42 jlk: hehe=)
21:44 boris-42 jlk: so no migration at this moment
21:44 jlk I think I may speak about Rally at the next operators meet up, unless somebody from your org would be doing the same. Operators Meetup is in March, in eastern USA
21:45 boris-42 jlk: so I am not sure that I'll get visa till March
21:45 boris-42 jlk: so that will be actually nice
21:45 boris-42 =)
21:46 jlk awesome. I was already planning on speaking about it here in Seattle for one of our next openstack meetups (much smaller group), but speaking at the ops meetup might get more users and contributors
21:46 boris-42 jlk: more feedback=)
21:47 openstackgerrit Merged stackforge/rally: Move CinderVolumes tasks from rally.yaml to rally-neutron.yaml  https://review.openstack.org/144183
22:00 boris-42 eye_: ooOO docker-push http://185.8.56.87/logs/DUXRFgVQMIpGAfwe/docker-push/ : SUCCESS in 7m 54s
22:01 eye_ boris-42, wow %)
22:02 boris-42 eye_: LOLLL https://registry.hub.docker.com/u/frobware/openstack-rally/
22:03 boris-42 eye_: btw https://registry.hub.docker.com/u/rallyforge/rally/tags/manage/ it should be here?
22:04 eye_ boris-42, not yet. seems i forgot to add actually "docker push" command to that script
22:04 boris-42 eye_: ahah
22:04 boris-42 eye_: btw rename it
22:04 boris-42 eye_: from rally-current to rally-master
22:05 eye_ boris-42, AFAIR it is default name for new build
22:07 boris-42 eye_: btw could you put some description?
22:07 boris-42 eye_: to repo
22:07 openstackgerrit Jesse Keating proposed stackforge/rally: Don't hard set keystone endpoint  https://review.openstack.org/144366
22:07 boris-42 eye_: e.g. "OpenStack enterprise grade, production ready testing Tool"
22:07 eye_ -_-
22:07 boris-42 we need more marketing bullshit
22:07 boris-42 =)
22:28 jlk um... pep8 is failing for everybody right, with the single quotes thing?
22:29 andreykurilin joined #openstack-rally
22:33 dspano boris-42: Just wondering. It seems that we have environments that don't support ephemeral user-creation. Is there currently a way to statically set the created user?
22:34 boris-42 dspano: nope
22:34 boris-42 dspano:  I am working on that task already 8 months=)
22:34 boris-42 dspano: and it's essential priority for rally
22:34 boris-42 =)
22:35 dspano boris-42: That might be something I'd be interested in helping with.
22:36 boris-42 dspano: so basically tasks are too huge
22:37 boris-42 dspano: usually I am fighting to make patches less then 3k lines of changes=)
22:37 openstackgerrit Oleh Anufriiev proposed stackforge/rally: WIP Report generation improvement  https://review.openstack.org/143866
22:37 boris-42 dspano: and actually I am quite close to full implementation
22:37 dspano boris-42: Cool.
22:39 boris-42 dspano: but we have billions other tasks
22:40 boris-42 dspano: if you wish to help =)
23:06 dspano boris-42: Oh, I see it's support-existing-users under the roadmap.
23:08 dspano boris-42: If you're almost done with this. I don't want to muddy the waters.
23:19 boris-42 dspano: it's quite hard task in any case=)
23:20 boris-42 dspano: not sure that you can work 200% of you working time on it?)
23:21 boris-42 dspano: btw are you going to work on that task?
23:21 boris-42 dspano: related to @base.runner decorator?
23:31 dspano boris-42: Adding the context method to rally/benchmark/runners/base.py?
23:31 boris-42 dspano: no not context method
23:31 boris-42 dspano: it will look like it
23:33 boris-42 dspano: just to add extra fileds to class
23:33 dspano boris-42: I'm a dummy. Now I know what's going on.
23:34 boris-42 dspano: so yep that way prevent dummy mistakes
23:34 boris-42 dspano: for people who is writing plugins
23:34 boris-42 dspano: like you can't call method (decorator) with wrong arguments
23:34 boris-42 self checking stuff*
23:38 boris-42 dspano: so wanna work on it?
23:42 dspano boris-42: Yeah. If I run into trouble, I'll cry.
23:43 boris-42 dspano: lol=)
23:43 boris-42 dspano: just cry in right place=) here=)
23:45 dspano boris-42: I will. Lol.
23:47 dspano boris-42: Time for dinner. Have a good night.

| Channels | #openstack-rally index | Today | | Search | Google Search | Plain-Text | summary