Perl 6 - the future is here, just unevenly distributed

IRC log for #openstack-rally, 2015-01-21

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

All times shown according to UTC.

Time Nick Message
00:16 mwagner_lap joined #openstack-rally
00:34 dmorita joined #openstack-rally
00:37 openstackgerrit Andrey Kurilin proposed stackforge/rally: WIP: fix several py3 issues  https://review.openstack.org/148763
00:50 e0ne joined #openstack-rally
01:07 jlk boris-42: am I doing something wrong to get these maximum failure percentages so high?  http://rally.openstack.blueboxgrid.com/bbg-tardis-validate.html#/NeutronNetworks.create_and_delete_ports
02:45 jtaleric joined #openstack-rally
04:03 psd_ joined #openstack-rally
04:12 coolsvap joined #openstack-rally
04:35 rdas joined #openstack-rally
04:47 asrangne joined #openstack-rally
05:00 nkhare joined #openstack-rally
05:04 rdas joined #openstack-rally
05:11 nkhare joined #openstack-rally
05:35 tfreger joined #openstack-rally
06:14 yingjun joined #openstack-rally
06:16 neeti joined #openstack-rally
06:22 exploreshaifali joined #openstack-rally
06:51 prmtl joined #openstack-rally
06:58 yfried joined #openstack-rally
06:58 yfried_ joined #openstack-rally
07:17 yfried_ boris-42: good morning
07:17 yfried_ boris-42: how are you?
07:17 yfried_ boris-42: I pushed a pretty major refactor yesterday https://review.openstack.org/148637
07:17 yfried_ boris-42: I'd like your opinion if this is a road you approve of
07:29 nmagnezi joined #openstack-rally
07:31 plieb joined #openstack-rally
07:57 yfried_ is int possible to install to Rally instances on the same machine using the -v flag?
08:22 yingjun joined #openstack-rally
08:25 yingjun psd, hi
08:26 psd yingjun, Hi!
08:27 yingjun psd, about your patch https://review.openstack.org/#/c/148308/4/rally/cmd/envutils.py, i mean is it possible to add _() for the text
08:27 yingjun psd, like MSG_MISSING_ARG = _("Missing argument: --%(arg_name)s")
08:28 psd yingjun, I am new to rally, What does _() do?
08:28 psd I have seen it lots of times...
08:29 psd I don't think it would be a problem
08:32 exploreshaifali joined #openstack-rally
08:32 yingjun psd, see https://github.com/stackforge/rally/blob/master/rally/common/i18n.py#L25, this is not a big problem, it’s used to mark the text for translation.
08:35 psd yingjun,Thanks
08:35 psd i'll make the change :)
08:35 yingjun psd, thanks
08:38 psd yingjun, Should I also do it for the message passed in with_default_deployment ?
08:39 tkammer joined #openstack-rally
08:42 yingjun psd, that’s what i mean :)
08:45 arxcruz joined #openstack-rally
08:45 openstackgerrit Prabhjyot Singh Sodhi proposed stackforge/rally: Improving error message  https://review.openstack.org/148308
08:46 psd yingjun, :)
08:47 psd joined #openstack-rally
08:48 yingjun psd, thanks, looks good, lets wait for jenkins
08:57 oanufriev joined #openstack-rally
08:58 Aswad joined #openstack-rally
09:09 psd yingjun, yeah!
09:11 asrangne joined #openstack-rally
09:13 openstackgerrit Jacob Liberman proposed stackforge/rally: Replace ' with " in rally/benchmark  https://review.openstack.org/148851
09:14 Prabhjyot__ joined #openstack-rally
09:27 openstackgerrit Yair Fried proposed stackforge/rally: Refactor AtomicAction into AtomicMixin  https://review.openstack.org/148637
09:28 amaretskiy joined #openstack-rally
09:29 andreykurilin_ joined #openstack-rally
09:32 oanufriev joined #openstack-rally
09:35 yingjun joined #openstack-rally
09:35 openstackgerrit Wataru Takase proposed stackforge/rally: Add cold migrate scenario for Nova test  https://review.openstack.org/148307
09:35 yingjun joined #openstack-rally
09:57 yfried_ I'm trying to write a python script that's cloning and installing rally and then create a deployment and installs tempest
09:58 yfried_ andreykurilin: ^
09:58 yfried_ andreykurilin: how can I rally's python api instead of bash?
10:01 yfried_ rediskin: are you Sergey Skripnick?
10:01 rediskin yfried_, yes
10:01 yfried_ see my answer in bug
10:02 yfried_ also, do you happen to know how can I use rally's api in python instead of writing bash commands (ie "rally deployment-creat")
10:03 e0ne joined #openstack-rally
10:11 rediskin yfried_, thanks for reporting
10:12 openstackgerrit Andrey Kurilin proposed stackforge/rally: WIP: fix several py3 issues  https://review.openstack.org/148763
10:14 dmellado joined #openstack-rally
10:14 andreykurilin_ yfried: hi! you should look at https://github.com/stackforge/rally/blob/master/rally/api.py#L202
10:15 openstackgerrit Sergey Skripnick proposed stackforge/rally: Add new script for gate jobs  https://review.openstack.org/144633
10:25 openstackgerrit Arx Cruz proposed stackforge/rally: Add a validator for neutron networks  https://review.openstack.org/141813
10:36 andreykurilin_ joined #openstack-rally
10:40 openstackgerrit Anton Arefiev proposed stackforge/rally: Change cinder scenario: create and delete volume  https://review.openstack.org/146820
10:40 openstackgerrit Anton Arefiev proposed stackforge/rally: Add cinder scenario: create volume from volume(clone)  https://review.openstack.org/147449
10:43 openstackgerrit Arx Cruz proposed stackforge/rally: Add a validator for neutron networks  https://review.openstack.org/141813
10:46 openstackgerrit Anton Arefiev proposed stackforge/rally: Add cinder scenario: create volume from volume(clone)  https://review.openstack.org/147449
10:51 openstackgerrit Anastasia Kuznetsova proposed stackforge/rally: Add Mistral benchmark  https://review.openstack.org/144320
10:53 asrangne joined #openstack-rally
10:55 yfried_ arxcruz: you haven't answered my question
10:56 openstackgerrit Arx Cruz proposed stackforge/rally: Refactoring benchmark/validation.py  https://review.openstack.org/148877
10:56 yfried_ arxcruz: why can't you validate most of the stuff via the JSON schema
10:56 arxcruz yfried_, this patch was requested by boris, I don't know why he didn't want to use JSON schema :/
10:56 yfried_ andreykurilin: amaretskiy: looking for help with rally's python api
10:56 prmtl joined #openstack-rally
10:57 yfried_ arxcruz: that's a A answer... :)
10:57 arxcruz I'm new to rally, I'm not a specialist, if there's a better way to do let me know
10:57 arxcruz yfried_, :)
10:57 yfried_ arxcruz: same here
11:05 boris-42 joined #openstack-rally
11:15 yfried_ boris-42: good morning
11:15 yfried_ boris-42: I need some help with rally's api
11:20 boris-42 yfried_: api?
11:22 prmtl joined #openstack-rally
11:24 yfried_ boris-42: I'm working on this script http://paste.openstack.org/show/159660/
11:24 yfried_ boris-42: to install rally in our ci
11:25 yfried_ boris-42: so I wonder if L28-L33 could be replaced with something like L27
11:25 yfried_ boris-42: but I'm getting:
11:26 yfried_ boris-42: http://paste.openstack.org/show/159661/
11:32 andreykurilin yfried_: Does install_rally.sh script finished successfully?
11:33 obutenko joined #openstack-rally
11:33 yfried_ andreykurilin: yeah.
11:33 yfried_ andreykurilin: the subprocess calls work
11:34 yfried_ andreykurilin: the problem is with trying to use the python api
11:35 boris-42 yfried_: don't use rally.cmd.commands
11:35 boris-42 yfried_: as python lib interface
11:35 boris-42 yfried_: you should use rally.api
11:35 andreykurilin yfried_: can you check that database is exist
11:35 boris-42 yfried_: it will be simpler
11:35 andreykurilin ?
11:36 yfried_ boris-42: andreykurilin I tried that first. got the same result
11:36 yfried_ 1sec
11:37 msdubov_ joined #openstack-rally
11:39 boris-42 arxcruz: https://review.openstack.org/#/c/148877/ put this patch first
11:40 boris-42 arxcruz: or indpendent from that first one
11:41 yfried_ andreykurilin: boris-42: http://paste.openstack.org/show/159667/ http://paste.openstack.org/show/159668/
11:42 yfried_ andreykurilin: boris-42: using the cli I was able to create a deployment and install tempest
11:42 yfried_ so the db is fine
11:42 andreykurilin hm
11:43 arxcruz boris-42, any chance you know how to remove the dependence ? :D
11:45 andreykurilin yfried_: db is fine and table rally.deployments exists?)
11:45 boris-42 arxcruz: yep I know
11:45 boris-42 arxcruz: git checkout master
11:46 arxcruz boris-42, of course you know ;) can you share your knowledge ?
11:46 yfried_ andreykurilin: boris-42: http://paste.openstack.org/show/159671/
11:46 yfried_ yes
11:46 boris-42 arxcruz: git checkout -b some_other_branch
11:46 boris-42 arxcruz: git cherry-pick <HASH_OF_SECOND_PATCH>
11:46 boris-42 arxcruz: git review -R
11:46 boris-42 arxcruz: will remove dep
11:46 arxcruz boris-42, thanks
11:46 * arxcruz need to learn more about git stuff
11:48 openstackgerrit Arx Cruz proposed stackforge/rally: Refactoring benchmark/validation.py  https://review.openstack.org/148877
12:02 arxcruz boris-42, done :)
12:03 akuznetsova Hi, guys, I have a question. I run tox -e py27 and one of tests (test_schema_is_valid) is failed with error:  These scenarios don't have samples: (list of scenarios). What do I need to add?
12:03 andreykurilin yfried_: I know how to fix your script:)
12:04 msdubov_ akuznetsova: We have a directory for samples: https://github.com/stackforge/rally/tree/master/samples/tasks/scenarios
12:04 boris-42 arxcruz: you need to add samples in samples/
12:04 boris-42 akuznetsova: ^
12:04 boris-42 arxcruz: sorry*
12:05 andreykurilin yfried_: if you add these two lines http://paste.openstack.org/show/159685/ , your script should work
12:06 akuznetsova msdubov_, thanks
12:07 openstackgerrit Merged stackforge/rally: Add cold migrate scenario for Nova test  https://review.openstack.org/148307
12:08 akuznetsova msdubov_, oh, you changed path to samples from rally/doc/samples to rally/samples
12:09 yfried_ andreykurilin: tnx. is it documented somewhere?
12:09 andreykurilin yfried_: no:(
12:10 yfried_ andreykurilin: well too bad
12:14 andreykurilin msdubov_: Are you here?)
12:15 yfried_ andreykurilin: how can I access rally-manage? its commands aren't in rally.api
12:17 andreykurilin yfried_, boris-42: looks like rally.api doesn't cover all rally stuff:(
12:18 openstackgerrit Sergey Skripnick proposed stackforge/rally: Allow to change db url via environment variable  https://review.openstack.org/148907
12:19 andreykurilin yfried_: so you can do it by two ways: using rally.cmd or move some logic from rally-manage to rally.api :)
12:21 openstackgerrit Merged stackforge/rally: Add list_snapshots scenario for Cinder test  https://review.openstack.org/146580
12:24 psd boris-42, Hey!
12:27 openstackgerrit Anton Arefiev proposed stackforge/rally: Add image arg for creating volume from image  https://review.openstack.org/146820
12:27 openstackgerrit Anton Arefiev proposed stackforge/rally: Add cinder scenario: create volume from volume(clone)  https://review.openstack.org/147449
12:28 boris-42 psd: hey hey
12:28 boris-42 andreykurilin: yfried_ it should cover
12:28 boris-42 andreykurilin: yfried_ actually I even have in RoadMap point about that
12:28 boris-42 yfried_: andreykurilin msdubov_ is going to start working on api
12:29 boris-42 soio
12:29 andreykurilin boris-42: our TODO list is too big:)
12:29 boris-42 soo
12:29 boris-42 andreykurilin: lol
12:29 boris-42 andreykurilin: it was not bigger then 1 year ago
12:29 msdubov_ andreykurilin: Hi, was away for a lunch
12:30 nkhare joined #openstack-rally
12:30 andreykurilin msdubov_: hi. we missed some interesting(and important) information from docs:)
12:31 msdubov_ andreykurilin:?
12:32 andreykurilin msdubov_: rally.api required http://paste.openstack.org/show/159685/ to work from external scripts
12:33 openstackgerrit Anton Arefiev proposed stackforge/rally: Add image arg for creating volume from image  https://review.openstack.org/146820
12:33 openstackgerrit Anton Arefiev proposed stackforge/rally: Add cinder scenario: create volume from volume(clone)  https://review.openstack.org/147449
12:35 msdubov_ andreykurilin: Thanks for pointing out, will add to the wishlist
12:36 andreykurilin yfried_: our docs guru will fix this missed information about rally.api :) ^
12:37 yfried_ andreykurilin: about manage or about oslo?
12:37 andreykurilin yfried_: about oslo
12:38 yfried_ andreykurilin: tnx
12:39 openstackgerrit Yair Fried proposed stackforge/rally: Adds tempest install to rally.api  https://review.openstack.org/148912
12:44 psd boris-42, how are you doing ?
12:46 RaginBajin joined #openstack-rally
12:47 RaginBajin joined #openstack-rally
12:48 psd boris-42, got some task for me?
12:51 yfried_ boris-42: andreykurilin: weird output using rally verify:
12:51 yfried_ http://paste.openstack.org/show/159695/
12:51 yfried_ boris-42: andreykurilin: it works fine without "set" or "regex"
12:52 boris-42 psd: hey there
12:52 boris-42 psd: let me first re-review your patch
13:00 psd boris-42, Yeah! sure :)
13:11 cdent joined #openstack-rally
13:15 pboros joined #openstack-rally
13:16 openstackgerrit Sergey Skripnick proposed stackforge/rally: Add new script for gate jobs  https://review.openstack.org/144633
13:17 openstackgerrit Sergey Skripnick proposed stackforge/rally: Add new script for gate jobs  https://review.openstack.org/144633
13:27 yfried_ msdubov_: re https://review.openstack.org/#/c/148912/
13:27 yfried_ msdubov_: since we have unittest for Tempest.install() isn't that enough?
13:31 msdubov_ yfried_: Nope, we want to have 100% coverage. Especially for such important methods as in the API. So basically the unit test just has to check that the verifier gets called as expected
13:32 msdubov_ psd Just reviewed https://review.openstack.org/#/c/148308/ could you please fix?
13:32 psd msdubov_, Sure :)
13:32 yfried_ msdubov_: I have no experience with unitest, but I'll tyr
13:32 yfried_ try
13:32 boris-42 yfried_: lol? I thought you have?)
13:33 openstackgerrit Arx Cruz proposed stackforge/rally: Refactoring benchmark/validation.py  https://review.openstack.org/148877
13:33 yfried_ boris-42: why would you think that?
13:33 msdubov_ yfried_: As you've just moved the install code from cmd/manage.py you can also take the code from its unit tests: https://github.com/stackforge/rally/blob/master/tests/unit/cmd/test_manage.py#L57-L61
13:33 boris-42 yfried_: https://review.openstack.org/#/c/148187/
13:34 boris-42 yfried_: you have touched here /tests/unit +)
13:34 msdubov_ yfried_: And you should change these lines in test_manage.py, to just check that you call your new api.install_tempest method
13:36 yfried_ msdubov_: tnx
13:38 yfried_ boris-42: I am an expert in "The Science of Deduction"
13:38 yfried_ boris-42: :)
13:39 boris-42 yfried_: ahah
13:43 mwagner_lap joined #openstack-rally
13:48 openstackgerrit Anton Arefiev proposed stackforge/rally: Add image arg for creating volume from image  https://review.openstack.org/146820
13:48 openstackgerrit Anton Arefiev proposed stackforge/rally: Add cinder scenario: create volume from volume(clone)  https://review.openstack.org/147449
13:48 openstackgerrit Sergey Skripnick proposed stackforge/rally: Allow changing db url in unittests via env var  https://review.openstack.org/148907
13:49 Miouge joined #openstack-rally
13:51 openstackgerrit Prabhjyot Singh Sodhi proposed stackforge/rally: Improving error message  https://review.openstack.org/148308
14:08 exploreshaifali joined #openstack-rally
14:24 boris-42 yfried_: https://review.openstack.org/#/c/147571/
14:24 dpaterson joined #openstack-rally
14:35 e0ne joined #openstack-rally
14:37 openstackgerrit Arx Cruz proposed stackforge/rally: Refactoring benchmark/validation.py  https://review.openstack.org/148877
14:40 yfried_ boris-42: what do you mean "return hidden"
14:40 yfried_ boris-42: remove?
14:41 yfried_ boris-42: I thought hidden means you can't config this context in yaml/json
14:56 e0ne joined #openstack-rally
14:57 openstackgerrit Nikita Konovalov proposed stackforge/rally: [Sahara] Add Cloudera plugin support  https://review.openstack.org/146400
15:01 nkhare joined #openstack-rally
15:06 openstackgerrit Roman Vasilets proposed stackforge/rally: Add murano benchmarks  https://review.openstack.org/137661
15:06 openstackgerrit Roman Vasilets proposed stackforge/rally: Add murano base for murano benchmarks  https://review.openstack.org/137650
15:10 Miouge joined #openstack-rally
15:22 boris-42 yfried_: yep andreykurilin explained me
15:22 boris-42 yfried_: why it is required=) I just forgot that you are adding user facing argument
15:22 boris-42 yfried_: sry about that mergin it
15:47 msdubov_ joined #openstack-rally
15:58 openstackgerrit Anastasia Kuznetsova proposed stackforge/rally: Add Mistral benchmark  https://review.openstack.org/144320
16:01 e0ne joined #openstack-rally
16:02 jaypipes joined #openstack-rally
16:03 openstackgerrit Merged stackforge/rally: Adds Tempest source repo support for "Tempest" context  https://review.openstack.org/147571
16:13 prmtl joined #openstack-rally
16:27 openstackgerrit Andrey Kurilin proposed stackforge/rally: WIP: fix several py3 issues  https://review.openstack.org/148763
16:38 openstackgerrit Merged stackforge/rally: [Sahara] Add Cloudera plugin support  https://review.openstack.org/146400
16:59 meteorfox joined #openstack-rally
17:11 obutenko_ joined #openstack-rally
17:51 baker joined #openstack-rally
18:04 dpaterson joined #openstack-rally
18:13 openstackgerrit Jesse Keating proposed stackforge/rally: Update location to samples  https://review.openstack.org/148726
18:24 andreykurilin_ joined #openstack-rally
18:24 psd|afk joined #openstack-rally
18:26 RaginBajin joined #openstack-rally
18:27 openstackgerrit Prabhjyot Singh Sodhi proposed stackforge/rally: Improving error message  https://review.openstack.org/148308
18:28 RaginBajin joined #openstack-rally
18:29 boris-42 joined #openstack-rally
18:30 e0ne joined #openstack-rally
18:31 openstackgerrit Merged stackforge/rally: Refactoring benchmark/validation.py  https://review.openstack.org/148877
18:53 yfried_ joined #openstack-rally
19:24 openstackgerrit Prabhjyot Singh Sodhi proposed stackforge/rally: Improving error message  https://review.openstack.org/148308
19:25 exploreshaifali joined #openstack-rally
19:29 boris-42 jlk: ping
19:29 boris-42 jaypipes: you are using old deprecreated sla checker
19:29 boris-42 jlk: ^
19:29 boris-42 jaypipes: sorry
19:29 jaypipes np :)
19:29 jlk boris-42: oooh!
19:29 boris-42 jlk: https://github.com/stackforge/rally/blob/master/rally-jobs/rally.yaml#L154-L155 you should try this one
19:29 boris-42 jlk: it has min/max arguments
19:29 boris-42 jlk: sorry for long reply, yestarday I just died=)
19:30 openstackgerrit Jacob Liberman proposed stackforge/rally: Replace ' with " in rally/benchmark  https://review.openstack.org/148851
19:30 jlk boris-42: I was following http://rally.readthedocs.org/en/latest/tutorial/step_3_adding_success_criteria_for_benchmarks.html
19:30 boris-42 jlk: =(
19:30 boris-42 jlk: we need to fix that
19:30 boris-42 jlk: we made some mistakes in read the docs (cause refacotirng was really huge)
19:31 jlk yup, no biggie. Glad to have a way out
19:31 boris-42 jlk: so you can make a good thing and fix that place=) when you try this new one=)
19:31 plieb joined #openstack-rally
19:31 jlk sure, I already have one PR open related to the docs change :)
19:33 boris-42 jlk: haha=)
19:34 boris-42 jlk: heh we will find the way to make docs samples testable as well=)
20:15 nmagnezi joined #openstack-rally
20:15 nmagnezi joined #openstack-rally
20:22 plieb boris-42: ping
20:26 baker joined #openstack-rally
20:26 openstackgerrit Jesse Keating proposed stackforge/rally: Replace deprecated max_failure_precent sla  https://review.openstack.org/149039
20:26 boris-42 plieb: hey
20:40 e0ne joined #openstack-rally
20:42 psd joined #openstack-rally
20:50 plieb boris-42: i had a question about the gerrit workflow
20:50 plieb you asked me to update a few files in the patch
20:50 plieb and i was also asked to rebase because the merge failed
20:50 plieb after i already updated the files in the patch
20:50 plieb what is best way to handle that in gerrit?
20:51 boris-42 plieb: so
20:51 boris-42 you didn't rebase it
20:51 boris-42 plieb: and I have one question
20:51 boris-42 plieb: your patch is in master branch locally?
20:51 plieb no a topic branch locally
20:51 boris-42 plieb: so not a master?
20:52 plieb right
20:52 boris-42 plieb: okay then you need to do next (let's assume you are on your branch)
20:52 boris-42 git checkout master
20:52 boris-42 git remote update
20:52 boris-42 git pull
20:52 boris-42 git checkout -
20:52 boris-42 git rebase master
20:52 boris-42 (and fix merge conflicts)
20:52 boris-42 after that
20:53 boris-42 git add <all_files_that_had_merge_conflicts>
20:53 boris-42 git rebase --continue
20:53 boris-42 and then
20:53 boris-42 you can can run
20:53 boris-42 git review -R
20:53 boris-42 plieb: ^
20:53 plieb ah okay
20:54 plieb i thought git review -R would rebase
20:54 plieb thanks!
20:54 plieb should i do anything to clean up the patch set i already sent up for e=review?
21:03 ilbot3 joined #openstack-rally
21:03 Topic for #openstack-rally is now ☁ Rally RoadMap: http://goo.gl/JZkmwY ☁ Rally IRC chat logs http://irclog.perlgeek.de/openstack-rally ☁ Key persons to ask:  boris-42, msdubov, rediskin, andreykurilin, amaretskiy  ☁ To publish changes to Rally:  https://wiki.openstack.org/wiki/Rally/Develop#How_to_contribute   ☁ Rally weekly updates:  https://wiki.openstack.org/wiki/Rally/Updates
21:10 openstackgerrit Jacob Liberman proposed stackforge/rally: Replace ' with " in rally/benchmark  https://review.openstack.org/148851
21:10 plieb boris-42: thanks I updated the conflicting files and resubmitted
21:11 boris-42 plieb: great
21:12 boris-42 plieb: thank you for working on this
21:12 plieb my pleasure
21:16 boris-42 plieb: when you finish with this task, we have a lot of anothers
21:16 plieb 8)
21:16 boris-42 that are not sooo big and much more interesting=)
21:29 jlk wow, got the vm boot run command to work.
21:37 boris-42 jlk: we merged recently patch
21:37 boris-42 jlk: that allows it to run against neutron
21:37 boris-42 jlk: + add functional testing for that
21:37 boris-42 so now it should work lol=)
21:38 jlk right
21:38 jlk I had a problem with my cloud that the test was pointing out
21:38 jlk fixed the problem, test ran fine!
21:38 jlk slowly building up my release validation test suite
21:41 boris-42 jlk: so one of my goals is to make cloud valiation task template
21:41 boris-42 jlk: and put it in rally codebase
21:43 boris-42 jlk: btw soon we will public code for building thrid party testing for Rally
21:43 boris-42 jlk: like we have Mirnatis Rally CI
21:44 boris-42 jlk: so if you would like/have few servers yo can join this from bluebox side=)
21:46 msdubov_ joined #openstack-rally
21:48 RaginBajin joined #openstack-rally
21:54 baker joined #openstack-rally
22:11 jlk haha, extra servers are hard to come by
22:11 jlk but maybe
22:13 jlk boris-42: is there anything in the future for doing boot multiple servers, have them talk to each other?
22:26 mwagner_lap joined #openstack-rally
22:27 jlk boris-42: got a question on the  boot-runcommand-delete-with-disk benchmark. It seems to be making my instance attempt to boot from a volume that is... not bootable.
22:34 boris-42 jlk: so on first question answer yes
22:34 boris-42 jlk: we are building framework on VMTask base
22:34 boris-42 jlk: that will be a base for porting existing benchmarks in cloud
22:34 jlk alright
22:35 boris-42 jlk: like iperf, hpcc and so on
22:35 boris-42 jlk: https://docs.google.com/a/mirantis.com/document/d/1FQicnj6aIdbfYts9ofYdz0dUEf9zyssWqDjQR40h268/edit
22:35 boris-42 jlk: you can take a part in discussion here
22:35 boris-42 jlk: second thing
22:36 boris-42 jlk: it shouldn't actually boot VM from volume https://github.com/stackforge/rally/blob/master/rally/benchmark/scenarios/vm/vmtasks.py#L77-L87
22:36 boris-42 jlk: according to this code
22:37 boris-42 jlk: it should attach it..
22:40 jlk what does it attach it as though?
22:40 boris-42 jlk: ?)
22:40 jlk Attached To
22:40 jlk rally_volume_knkmrvuwzo on /dev/vda
22:41 jlk and for some reason the server is attempting to boot from /dev/vda
22:41 boris-42 jlk: oh
22:41 boris-42 jlk: maybe we should just boot VM
22:41 boris-42 jlk: and then attach volume to it?
22:41 jlk boot, and then attach?
22:41 jlk yeah
22:41 boris-42 jlk: using volumeattach?
22:41 boris-42 jlk: could you try to hack it ?
22:41 jlk I think the attach happens during boot or something
22:42 jlk or maybe name the attach something not vda
22:42 boris-42 jlk: btw
22:42 boris-42 jlk:  this works https://github.com/stackforge/rally/blob/master/rally-jobs/rally.yaml#L1015-L1037
22:43 jlk what is your normal instance storage?
22:43 jlk we're doing just local storage on the hypervisors
22:49 boris-42 jlk: not sure
22:49 boris-42 jlk: it's devstack
22:50 jlk lol, worksindevstack
22:50 boris-42 jlk: yep
22:50 boris-42 =)
22:50 boris-42 jlk: that is why we are bulding Mirantis CI
22:50 boris-42 jlk: so we will be able to say works in DevStack and MOS=)
22:50 jlk attaching it to vdb instead of vda totes works
22:51 jlk at least the system is booting
22:51 boris-42 jlk: intersting why in case of devstack we have different behaviour
22:51 jlk I wonder if the disk names in devstack are different
22:51 jlk or the timing is different
22:51 jlk in that the boot image gets the first disk before the volume
22:51 boris-42 jlk: maybe we should add abbility to speicfy where to attach volume
22:52 boris-42 ?
22:52 jlk honestly, I'm not sure.
22:52 jlk atually
22:52 boris-42 jlk: I mean there can be default value "vd*"
22:52 jlk actually
22:53 boris-42 but if somebody need=)
22:53 jlk why not just do something like "rally"
22:53 jlk then it'll get mounted to /dev/rally
22:53 jlk and be really safe from stomping on something else
22:53 boris-42 jlk: yep seems a good idea
22:53 boris-42 jlk: less conf options => simpler to use
22:53 boris-42 =)
22:54 jlk bingo
22:57 jlk yeah interesting. THe existing boot from volume stuff is usin vda as well
22:57 jlk maybe just pass a bootindex of not 0
23:10 leeantho joined #openstack-rally
23:14 openstackgerrit Jesse Keating proposed stackforge/rally: Avoid mapping non-boot volume to boot device  https://review.openstack.org/149076
23:17 boris-42 jlk: =)
23:18 jlk looks like I got some of the json wrong in my SLA patch :/
23:20 boris-42 jlk: yep
23:20 boris-42 jlk: it is wrong
23:20 boris-42 jlk: , at the end of line
23:20 boris-42 jlk: btw you won't belive
23:20 boris-42 jlk: but all browsers except IE 6 and 7 will parse such json
23:21 boris-42 jlk: and IE6,7 will break=)
23:21 boris-42 jlk: I mean they will crash*
23:21 jlk hrm, which line gets the comma?
23:21 jlk the comman at the end of the }  ?
23:25 openstackgerrit Jesse Keating proposed stackforge/rally: Replace deprecated max_failure_precent sla  https://review.openstack.org/149039
23:32 boris-42 jlk: i left reply =)
23:32 boris-42 jlk: so there shouldn't be such situations ",}"

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