| *** pvaneck has quit IRC | 00:52 | |
| *** beslemon_ has quit IRC | 01:16 | |
| *** banix has joined #openstack-rally | 03:00 | |
| banix | can someone check the log if you get a chance: http://logs.openstack.org/05/265105/34/check/gate-kuryr-dsvm-rally-nv/44cdbeb/console.html is this something we are doing incorrectly in Kuryr or there is a Rally issue; i see there was a bug that is related https://bugs.launchpad.net/kuryr/+bug/1556775 but that bug was fixed a while back | 03:04 |
|---|---|---|
| openstack | Launchpad bug 1556775 in Rally "gate-kuryr-dsvm-rally-nv is failing due to the recent rally change" [Critical,Fix released] - Assigned to Dmitry Ratushnyy (dratushnyy) | 03:04 |
| *** beslemon has joined #openstack-rally | 03:04 | |
| *** banix has quit IRC | 03:20 | |
| *** MaxPC has joined #openstack-rally | 04:03 | |
| *** MaxPC has quit IRC | 04:06 | |
| *** sshnaidm|afk has quit IRC | 04:13 | |
| *** Alex_Stef has joined #openstack-rally | 04:29 | |
| *** beslemon has quit IRC | 05:05 | |
| *** arnoldje has quit IRC | 05:08 | |
| *** rdas has joined #openstack-rally | 05:14 | |
| *** rcernin|off has joined #openstack-rally | 05:15 | |
| *** psd has joined #openstack-rally | 05:16 | |
| *** meteorfox has quit IRC | 05:24 | |
| *** anevenchannyy has quit IRC | 05:24 | |
| *** athinkingmeat has quit IRC | 05:24 | |
| *** andreykurilin__ has quit IRC | 05:25 | |
| *** serverascode has quit IRC | 05:25 | |
| *** kun_huang has quit IRC | 05:25 | |
| *** boris-42 has quit IRC | 05:25 | |
| *** zhiyan has quit IRC | 05:25 | |
| *** anevenchannyy has joined #openstack-rally | 05:26 | |
| *** rcernin|off has quit IRC | 05:26 | |
| *** meteorfox has joined #openstack-rally | 05:28 | |
| *** athinkingmeat has joined #openstack-rally | 05:32 | |
| *** andreykurilin__ has joined #openstack-rally | 05:32 | |
| *** zhiyan has joined #openstack-rally | 05:33 | |
| *** boris-42 has joined #openstack-rally | 05:34 | |
| *** serverascode has joined #openstack-rally | 05:34 | |
| *** kun_huang has joined #openstack-rally | 05:37 | |
| *** rcernin|off has joined #openstack-rally | 06:14 | |
| *** rcernin|off is now known as rcernin | 06:21 | |
| *** zackf has quit IRC | 06:25 | |
| *** Guest8385 has joined #openstack-rally | 06:29 | |
| *** abalutoiu has joined #openstack-rally | 06:38 | |
| *** akrzos has joined #openstack-rally | 06:49 | |
| *** akrzos has quit IRC | 06:52 | |
| *** akrzos has joined #openstack-rally | 06:52 | |
| *** akrzos has quit IRC | 06:54 | |
| *** pcaruana has joined #openstack-rally | 07:09 | |
| *** arxcruz has joined #openstack-rally | 07:20 | |
| *** Alex_Stef has quit IRC | 07:25 | |
| *** psd has quit IRC | 07:44 | |
| *** amaretskiy has joined #openstack-rally | 07:45 | |
| *** akrzos has joined #openstack-rally | 07:55 | |
| *** tesseract has joined #openstack-rally | 07:56 | |
| *** akrzos has quit IRC | 07:59 | |
| *** Alex_Stef has joined #openstack-rally | 08:00 | |
| *** ushkalim has joined #openstack-rally | 08:01 | |
| *** openstackgerrit has quit IRC | 08:03 | |
| *** openstackgerrit has joined #openstack-rally | 08:03 | |
| *** psuriset has quit IRC | 08:10 | |
| *** permalac has joined #openstack-rally | 08:21 | |
| *** psuriset has joined #openstack-rally | 08:27 | |
| *** psuriset has quit IRC | 08:35 | |
| *** zackf has joined #openstack-rally | 08:37 | |
| *** zackf has quit IRC | 08:42 | |
| *** Alex_Stef has quit IRC | 08:43 | |
| *** yingjun has joined #openstack-rally | 08:46 | |
| *** psuriset has joined #openstack-rally | 08:47 | |
| *** alexchadin has joined #openstack-rally | 08:52 | |
| *** rdas has quit IRC | 08:54 | |
| *** Guest8385 has quit IRC | 08:58 | |
| *** yingjun has quit IRC | 09:01 | |
| *** yingjun has joined #openstack-rally | 09:01 | |
| *** rdas has joined #openstack-rally | 09:02 | |
| *** e0ne has joined #openstack-rally | 09:04 | |
| *** yingjun has quit IRC | 09:06 | |
| *** yingjun has joined #openstack-rally | 09:06 | |
| *** Alex_Stef has joined #openstack-rally | 09:08 | |
| *** Guest8385 has joined #openstack-rally | 09:11 | |
| *** alexchadin has quit IRC | 09:20 | |
| strigazi | rvasilets__: are you there? | 09:24 |
| *** amaretskiy has left #openstack-rally | 09:24 | |
| *** yingjun has quit IRC | 09:25 | |
| *** yingjun has joined #openstack-rally | 09:25 | |
| strigazi | or amaretskiy: Are rally's packages up to date? https://github.com/openstack-packages/rally | 09:26 |
| *** yingjun has quit IRC | 09:30 | |
| *** tosky has joined #openstack-rally | 09:30 | |
| *** sshnaidm|afk has joined #openstack-rally | 09:32 | |
| *** amaretskiy has joined #openstack-rally | 09:35 | |
| *** adiantum has joined #openstack-rally | 09:45 | |
| *** mvk_ has quit IRC | 09:46 | |
| openstackgerrit | Merged openstack/rally: Updated from global requirements https://review.openstack.org/324866 | 09:50 |
| openstackgerrit | Andrey Kurilin proposed openstack/rally: Improve test_docstrigs https://review.openstack.org/291695 | 09:55 |
| *** yingjun has joined #openstack-rally | 10:00 | |
| *** mvk_ has joined #openstack-rally | 10:17 | |
| openstackgerrit | Illia Khudoshyn proposed openstack/rally: [CI]Prepare for gate job testing existing users context https://review.openstack.org/324577 | 10:19 |
| *** Guest8385 has quit IRC | 10:32 | |
| *** Guest8385 has joined #openstack-rally | 10:44 | |
| rvasilets__ | strigazi, hi | 10:55 |
| rvasilets__ | strigazi, They are updated due to the last release | 10:56 |
| rvasilets__ | strigazi, the last commit there should be https://github.com/openstack/rally/commit/d8fde48e865f7d5a0d8d66c4cc772e08dbda8dd6 | 10:57 |
| rvasilets__ | 18 april | 10:57 |
| openstackgerrit | Andrey Kurilin proposed openstack/rally: Modify file-exporter plugin and base TaskExporter https://review.openstack.org/301750 | 10:57 |
| openstackgerrit | Andrey Kurilin proposed openstack/rally: [docs] show references based on plugin bases https://review.openstack.org/301913 | 10:57 |
| openstackgerrit | Andrey Kurilin proposed openstack/rally: [Docs] Include cofig json schema in plugin refs https://review.openstack.org/327592 | 10:57 |
| *** adiantum has quit IRC | 11:03 | |
| *** adiantum has joined #openstack-rally | 11:04 | |
| openstackgerrit | Andrey Kurilin proposed openstack/rally: Improve test_docstrigs https://review.openstack.org/291695 | 11:19 |
| openstackgerrit | Illia Khudoshyn proposed openstack/rally: [CI]Prepare for gate job testing existing users context https://review.openstack.org/324577 | 11:30 |
| *** akrzos has joined #openstack-rally | 11:51 | |
| strigazi | rvasilets__: thanks | 11:52 |
| strigazi | rvasilets__: I have a question about the cleanup of resources: https://review.openstack.org/#/c/309546/7/rally/plugins/openstack/cleanup/resources.py | 11:54 |
| openstackgerrit | Andrey Kurilin proposed openstack/rally: Modify file-exporter plugin and base TaskExporter https://review.openstack.org/301750 | 12:04 |
| *** alexchadin has joined #openstack-rally | 12:14 | |
| *** sshnaidm|afk is now known as sshnaidm | 12:16 | |
| openstackgerrit | Illia Khudoshyn proposed openstack/rally: [CI]Prepare for gate job testing existing users context https://review.openstack.org/324577 | 12:18 |
| rvasilets__ | strigazi, Looks like that magnum baymodels could be deleted even if their heat stacks were deleted earlier. What would be going if we: 1)Create Magnum baymodel 2) Delete all heat resources. Will the baymodel would be steel working? | 12:20 |
| rvasilets__ | strigazi, I saw that magnum is dependent on heat. So I think that its not the good Idea to clean up heat resources first. But looks like it work=) | 12:23 |
| *** akrzos has quit IRC | 12:26 | |
| *** akrzos has joined #openstack-rally | 12:27 | |
| *** Guest8385 has quit IRC | 12:29 | |
| *** Anticimex has joined #openstack-rally | 12:29 | |
| *** abehl has joined #openstack-rally | 12:30 | |
| Anticimex | hi | 12:30 |
| Anticimex | are there rally devs who're aware of the concurrency issues when running jobs with parallelism? | 12:31 |
| *** akrzos has quit IRC | 12:31 | |
| rvasilets__ | Anticimex, hi, with what problems you faced with?) | 12:36 |
| Anticimex | trams can elaborate | 12:38 |
| trams | well. rally seem to confuse which projects each thread belongs too. Thi results in errors but also that rally fails to clean up | 12:45 |
| trams | An example would be NeutronNetworks.create_and_list_subnets | 12:46 |
| andreykurilin | trams Anticimex: Are you running several rally processes in parallel? | 12:48 |
| trams | nope, only one process | 12:48 |
| trams | but tasks have concurrency ofc. | 12:49 |
| trams | platform el7, python 2.7.5 | 12:49 |
| rvasilets__ | is there any traces or logs? | 12:50 |
| trams | sure. :) | 12:50 |
| andreykurilin | trams: what rally version is used? | 12:51 |
| trams | https://paste.fedoraproject.org/376614/76680146/ | 12:51 |
| trams | rally --version | 12:51 |
| trams | 0.4.1~dev20 | 12:51 |
| trams | hm. that was a shitty paste | 12:52 |
| rvasilets__ | ) | 12:53 |
| andreykurilin | trams: sorry, I don't understand you issue. Could you share more details? | 12:53 |
| trams | ofc. the stack trace is because neutron isnt happy | 12:53 |
| rvasilets__ | Full log of the task for example | 12:54 |
| trams | hm, i'll try to explain. i believe the issue is that when rally uses multiple tenants and users, created resources are only vissible to those specific users, but rally seem to get confused and use the wrong tennant to delete/list the network after it is created | 12:55 |
| trams | the problem seem to be generall, not for a specific task, but i guess it is easier with a somewhat simple one, like create_and_delete_subnets | 12:56 |
| andreykurilin | trams: could you share code of _run_scenario_once method from rally/task/runner.py ? | 12:57 |
| trams | andreykurilin: sure. | 12:57 |
| trams | andreykurilin: https://paste.fedoraproject.org/376616/77134146/ | 12:59 |
| andreykurilin | trams: everything looks good | 12:59 |
| *** akrzos has joined #openstack-rally | 13:00 | |
| andreykurilin | trams: Does it reproduced only with neutron scenarios? | 13:00 |
| *** dpaterson has joined #openstack-rally | 13:01 | |
| trams | andreykurilin: no, it seems to work with others as well. | 13:02 |
| trams | andreykurilin: reducing concurrency to 1 always removes the issue | 13:02 |
| andreykurilin | trams: it is strange | 13:02 |
| andreykurilin | too strange | 13:02 |
| trams | andreykurilin: yes. at first i was convinced that the problem was not related at all to rally but instead to the deployment, but after disabling ssl everywhere and lookuing in pcaps very carefully i changed opinion | 13:03 |
| *** absubram has quit IRC | 13:05 | |
| andreykurilin | trams: 1) we launch most of scenarios in our gates for each new change. As you can see http://logs.openstack.org/95/291695/3/check/gate-rally-dsvm-neutron-rally/f1efb3a/rally-plot/results.html.gz there are few scenarios which have concurrency==1(they take took much time); 2) each iteration of task initialize new class of scenario and launch specified method. This method initialize openstack clients one time, so I cannot imagine situation wh | 13:07 |
| *** arnoldje has joined #openstack-rally | 13:08 | |
| openstackgerrit | Andrey Kurilin proposed openstack/rally: [Verify] Add tempest configs and move to new section https://review.openstack.org/315171 | 13:19 |
| *** yingjun has quit IRC | 13:20 | |
| *** yingjun has joined #openstack-rally | 13:20 | |
| andreykurilin | trams: I have no ideas about the core of issue. I can only suggest you to do something like - https://review.openstack.org/#/c/288522/ to ensure that everything is write at rally side:) | 13:21 |
| *** yingjun has quit IRC | 13:25 | |
| ciprian-barbu | ylobankov: ping | 13:36 |
| *** yingjun has joined #openstack-rally | 13:38 | |
| *** arnoldje has quit IRC | 13:39 | |
| *** banix has joined #openstack-rally | 13:39 | |
| *** yingjun has quit IRC | 13:39 | |
| *** yingjun has joined #openstack-rally | 13:39 | |
| *** MaxPC has joined #openstack-rally | 13:40 | |
| trams | andreykurilin: nice, thanks! | 13:40 |
| ylobankov | ciprian-barbu: pong | 13:41 |
| banix | hi. I get this error in our Rally gate job: “ERROR (EndpointNotFound): publicURL endpoint for compute service in RegionOne region not found” | 13:42 |
| banix | i see there was a Rally bug that causes the same error: https://bugs.launchpad.net/kuryr/+bug/1556775 but that bug was fixed a while back | 13:42 |
| openstack | Launchpad bug 1556775 in Rally "gate-kuryr-dsvm-rally-nv is failing due to the recent rally change" [Critical,Fix released] - Assigned to Dmitry Ratushnyy (dratushnyy) | 13:42 |
| banix | Can someone have a look at the log here http://logs.openstack.org/05/265105/34/check/gate-kuryr-dsvm-rally-nv/44cdbeb/console.html and see if this is something we are doing incorrectly in Kuryr or there is a Rally issue | 13:43 |
| andreykurilin | (facepalm) | 13:43 |
| ylobankov | ciprian-barbu: I am reviewing you patch right now | 13:43 |
| andreykurilin | banix: hi | 13:43 |
| banix | andreykurilin: hi | 13:43 |
| *** yingjun has quit IRC | 13:43 | |
| andreykurilin | banix: I know the reason of the issue:( | 13:44 |
| andreykurilin | banix: I think it relates to https://review.openstack.org/#/c/311526/ | 13:44 |
| *** zackf has joined #openstack-rally | 13:45 | |
| *** zackf has quit IRC | 13:45 | |
| *** zackf has joined #openstack-rally | 13:45 | |
| banix | andreykurilin: looks like it :) anything we can do about it? I guess we hit the rror because we do not have Nova service running. right? | 13:46 |
| andreykurilin | banix: yes:( | 13:46 |
| andreykurilin | banix: could you post a small change to rally repo? | 13:47 |
| banix | shoudln’t the nova boot statement go inside the if block right above it? | 13:47 |
| andreykurilin | yes | 13:47 |
| andreykurilin | it should | 13:47 |
| banix | sure, will do | 13:47 |
| andreykurilin | like we have at L58 | 13:47 |
| andreykurilin | banix: thanks | 13:48 |
| banix | thanks for pointing me to the source of the problem | 13:48 |
| andreykurilin | thanks for pinging us:) | 13:48 |
| ylobankov | ciprian-barbu: once Jenkins gives a +1 to you, I will +2 it. | 13:50 |
| *** alexchadin has quit IRC | 13:52 | |
| andreykurilin | ylobankov: It haf +1 from jenkins in previous patch-set. The difference between patch-sets are in commit message. If we talk about https://review.openstack.org/#/c/315171/ :) | 13:53 |
| andreykurilin | ciprian-barbu: Please, use labels "Closes-Bug: #7777", "Related-Bug: #777" and others in commit messages in futures;) It helps to write release notes based on commit messages. Also, jenkins detects such labels and attachs fix to bug report | 13:55 |
| openstack | bug 7777 in Ubuntu "iso-codes update request" [Medium,Fix released] https://launchpad.net/bugs/7777 | 13:55 |
| openstack | bug 777 in helix-player (Ubuntu) "installscript fails to create a symbolic link for plugin registration (mozilla-firefox)" [Medium,Invalid] https://launchpad.net/bugs/777 | 13:55 |
| trams | andreykurilin: did another test with create_and_list_subnet. setting concurrency to 1 always work, where higher concurrency gives errors. | 13:56 |
| trams | andreykurilin: and also the requirement to clean up after rally manually | 13:56 |
| andreykurilin | trams: We had problems only with two neutron scenarios, but it was fixed a month ago and your code includes the fix(if you posted code which is actually used). Btw, what is the issue on cleanup steps? any traces/logs? | 13:58 |
| trams | andreykurilin: if you wish, i can run two rallys in parallell with concurrency 1. that should create a concurrency situation but isolate if rally have the issue or if it is neutron | 13:58 |
| trams | andreykurilin: i'm pretty sure the clean up errors are follow up errors rather than a seperate issue | 14:00 |
| *** alexchadin has joined #openstack-rally | 14:01 | |
| trams | interesting. doing a concurrency of 10, and number of tennants 10 also yields 0 errors | 14:01 |
| trams | will do the test some more, to see if this is always true | 14:02 |
| openstackgerrit | Illia Khudoshyn proposed openstack/rally: [CI]Prepare for gate job testing existing users context https://review.openstack.org/324577 | 14:02 |
| andreykurilin | O_o | 14:04 |
| trams | no it can still fail, but fails a lot less | 14:06 |
| openstackgerrit | Mohammad Banikazemi proposed openstack/rally: Use Nova only when Nova is available https://review.openstack.org/327691 | 14:07 |
| *** abehl has quit IRC | 14:09 | |
| *** asdfasdfasdf has joined #openstack-rally | 14:16 | |
| *** vkhalin has joined #openstack-rally | 14:16 | |
| *** asdfasdfasdf has quit IRC | 14:16 | |
| vkhalin | test | 14:17 |
| andreykurilin | vkhalin: hi | 14:17 |
| *** akrzos has quit IRC | 14:17 | |
| vkhalin | Houston I have a problem | 14:17 |
| ciprian-barbu | ylobankov: I can amend the commit message, would you like me to do that ? | 14:18 |
| ylobankov | ciprian-barbu: No, andreykurilin added the link to the bug :) | 14:18 |
| ciprian-barbu | ylobankov: great, thank you and sorry about that | 14:19 |
| ylobankov | ciprian-barbu: could you please assigne https://bugs.launchpad.net/rally/+bug/1580695 to yourself? | 14:20 |
| openstack | Launchpad bug 1580695 in Rally "flavor_ref and flavor_ref_alt have hardcoded 64 and 128 ram size" [Undecided,In progress] | 14:20 |
| *** abehl has joined #openstack-rally | 14:22 | |
| vkhalin | according to this man - https://rally.readthedocs.io/en/0.0.2/plugins.html#id8 I tried to add my custom scenario to Rally, placed filename.py to directory ~/.rally/plugins and started custom scenario via "rally task start ..." and got message that Task config is invalid: `There are no benchmark scenarios with names: `VMCustomScenario.boot_runcommand_delete`.` | 14:22 |
| vkhalin | where I've failed? | 14:22 |
| *** abehl has quit IRC | 14:22 | |
| *** anshul has joined #openstack-rally | 14:22 | |
| *** anshul is now known as Guest76774 | 14:23 | |
| *** arnoldje has joined #openstack-rally | 14:24 | |
| andreykurilin | vkhalin: could you share code of scenario? | 14:26 |
| openstackgerrit | Andrey Kurilin proposed openstack/rally: WIP: some work around deployment configs https://review.openstack.org/327715 | 14:30 |
| *** rook-remote has joined #openstack-rally | 14:31 | |
| *** alexchadin has quit IRC | 14:32 | |
| *** akrzos has joined #openstack-rally | 14:40 | |
| *** pcaruana has quit IRC | 14:41 | |
| *** banix has quit IRC | 14:45 | |
| *** banix has joined #openstack-rally | 14:49 | |
| *** vkhalin has quit IRC | 14:55 | |
| *** akrzos has quit IRC | 14:58 | |
| *** KevinE has joined #openstack-rally | 15:07 | |
| KevinE | can anyone help me with a strange issue I'm getting in rally? | 15:13 |
| *** abalutoiu has quit IRC | 15:26 | |
| *** rdas has quit IRC | 15:26 | |
| *** tesseract has quit IRC | 15:30 | |
| *** rcernin has quit IRC | 15:32 | |
| *** MaxPC has quit IRC | 15:33 | |
| *** rook-remote has quit IRC | 15:34 | |
| *** MaxPC has joined #openstack-rally | 15:44 | |
| *** KevinE_ has joined #openstack-rally | 15:45 | |
| *** KevinE has quit IRC | 15:46 | |
| *** banix has quit IRC | 15:46 | |
| *** banix has joined #openstack-rally | 15:47 | |
| *** beslemon has joined #openstack-rally | 15:57 | |
| *** beslemon has quit IRC | 15:57 | |
| *** beslemon has joined #openstack-rally | 15:58 | |
| *** arxcruz has quit IRC | 16:01 | |
| *** Guest76774 has quit IRC | 16:02 | |
| *** pvaneck has joined #openstack-rally | 16:22 | |
| *** leoswaldo has joined #openstack-rally | 16:27 | |
| *** MaxPC has quit IRC | 16:27 | |
| *** amaretskiy has quit IRC | 16:31 | |
| openstackgerrit | Illia Khudoshyn proposed openstack/rally: [CI]Prepare for gate job testing existing users context https://review.openstack.org/324577 | 16:32 |
| andreykurilin | KevinE_: what happened? | 16:32 |
| *** adiantum has quit IRC | 16:34 | |
| *** klindgren has quit IRC | 16:34 | |
| *** leoswaldo has left #openstack-rally | 16:35 | |
| KevinE_ | andreykurilin: I'm just really bad at python lol | 16:37 |
| andreykurilin | lol | 16:37 |
| *** ushkalim has quit IRC | 16:37 | |
| KevinE_ | andreykurilin: can you check this out for me? http://pastebin.com/XzjHZi7E | 16:39 |
| KevinE_ | So I copied from a scenarios file first. There is a method that I pulled (that I need) that used to tie to an older version of utils.generate_random_name. Now that method is in a class, RandomNameGeneratorMixin | 16:40 |
| *** akrzos has joined #openstack-rally | 16:40 | |
| KevinE_ | andreykurilin: That second class/file is listed below the ~~. I can't figure out how to edit that little method to return a correctly randomly generated name | 16:41 |
| andreykurilin | KevinE_: what is a correctly randomly generated name for you?) | 16:41 |
| KevinE_ | andreykurilin: I'm not sure. What I see is that the second class is up-to-date and should produce a good result. The only params I see is the format and length. In the old class it has prefix which I guess was before someone decided to change it to a format | 16:42 |
| KevinE_ | andreykurilin: I think all that needs to be done is correctly pull that resource_name_format, and somehow return from the correct method | 16:43 |
| andreykurilin | Kevin_: what is wrong now?) do you need to specify custom prefix? | 16:43 |
| KevinE_ | andreykurilin: So I'm getting error "AttributeError: 'module' object has no attribute 'generate_random_name'" because I think the way it calls that method is wrong | 16:44 |
| KevinE_ | it's not imported or something | 16:44 |
| andreykurilin | KevinE_: it looks like you are trying to call custom plugin. Right? | 16:44 |
| andreykurilin | *to execute custom scenario or custom context | 16:44 |
| *** beslemon has quit IRC | 16:45 | |
| KevinE_ | andreykurilin: the only thing I'm changing in these files is that one method | 16:45 |
| KevinE_ | oh wait yes lol I'm executing custom scenarios | 16:46 |
| *** klindgren has joined #openstack-rally | 16:46 | |
| KevinE_ | andreykurilin: I have an old internal repo I'm trying to fix to merge upstream | 16:46 |
| andreykurilin | KevinE_: does it really important to you to specify the length of name? | 16:46 |
| andreykurilin | KevinE_: let me describe the reason of changes in generate_random_name method | 16:47 |
| KevinE_ | andreykurilin: Not really sure, that var used to be 10 and I changed it to 25 when I counted the chars in the format, that's literally all I know about how these segments of code work | 16:47 |
| *** e0ne has quit IRC | 16:48 | |
| andreykurilin | KevinE_: previously, different scenarios and contexts were able to generate random names whatever the want, but it was not good idea, since there was not ability to identify resource. Now, each resource created by rally has name which include word rally and a part of task id. in the nearest future, you will be able to disable automatic cleanup after the task, check failed resources and launch cleanup for particular task manually. That is wh | 16:50 |
| andreykurilin | KevinE_: In your case, you need to change your code and call self.generate_random_name from scenario instead of utils.generate_random_name . It means that you need to drop ability to setup the length of name | 16:52 |
| KevinE_ | andreykurilin: Totally fine with me, I just want compatibility :) | 16:53 |
| KevinE_ | andreykurilin: so I can delete the length, and the prefix since that's not even a thing anymroe, what do I need to pass now? | 16:54 |
| *** diogogmt has joined #openstack-rally | 16:57 | |
| *** akrzos has quit IRC | 16:58 | |
| andreykurilin | KevinE_: you should pass nothing to latest generate_random_name method | 16:59 |
| *** mvk_ has quit IRC | 16:59 | |
| *** akrzos has joined #openstack-rally | 16:59 | |
| andreykurilin | KevinE_: What does you mean when said "want compatibility"? Do you want to port your scenario to latest rally or to make your scenarios compatible for both old release and new one? | 16:59 |
| KevinE_ | andreykurilin: port scenario to latest rally currently | 17:00 |
| andreykurilin | so you need to just call self.generate_random_name() | 17:00 |
| *** akrzos has quit IRC | 17:03 | |
| KevinE_ | @classmethod~~ def _generate_random_name(self):~~ return self.generate_random_name() | 17:03 |
| KevinE_ | andreykurilin: that's my method correct? :) | 17:04 |
| dpaterson | boris-42: any plans for a rally mid-cycle sprint? Trying to budget travel for second half of 2016 | 17:04 |
| KevinE_ | andreykurilin: http://pastebin.com/MGEzWarm There i have listed my new method with what you said. Now I'm getting error that _generate_random_name() takes 1 argument but 2 given, from the class that I have pasted there. How should I fix that part now? | 17:06 |
| Anticimex | andreykurilin: the way i understood our issue from a month back or so, was that high concurrency number on single tenant (iirc) basically meant concurrency-code messed up the pointers between its objects. so object A gets created at t_1. B gets created at t_2. A gets deleted at t_3, however, A has been replaced with something like C, which hasn't been created yet, and the error ends up being tha | 17:10 |
| Anticimex | t e.g. neutron can't delete ... | 17:10 |
| Anticimex | ... something that doesn't exist | 17:11 |
| Anticimex | i guess a trace of each thread with high precision timestamps could help sort out the global ordering of events | 17:11 |
| Anticimex | or similar | 17:11 |
| openstackgerrit | Merged openstack/rally: [Verify] Add tempest configs and move to new section https://review.openstack.org/315171 | 17:16 |
| *** permalac has quit IRC | 17:31 | |
| *** akrzos has joined #openstack-rally | 17:42 | |
| KevinE_ | hmm anyone else have a free second to help me trace a bit of python? | 17:50 |
| *** akrzos has quit IRC | 17:54 | |
| *** MaxPC has joined #openstack-rally | 17:57 | |
| *** serverascode has quit IRC | 17:58 | |
| *** serverascode has joined #openstack-rally | 18:00 | |
| *** rvasilets__ has quit IRC | 18:00 | |
| *** rvasilets__ has joined #openstack-rally | 18:00 | |
| *** abalutoiu has joined #openstack-rally | 18:10 | |
| *** arnoldje has quit IRC | 18:10 | |
| *** mgagne has quit IRC | 18:16 | |
| *** mgagne has joined #openstack-rally | 18:17 | |
| *** abalutoiu has quit IRC | 18:24 | |
| *** abalutoiu has joined #openstack-rally | 18:25 | |
| *** arnoldje has joined #openstack-rally | 18:29 | |
| njohnston | Hi! I have a quick question. I am working with an engineer to get him up to speed on Rally on a new development. He is trying out running a few tests from the samples directory, like samples/tasks/scenarios/nova/list-hypervisors.yaml - but he keeps getting the error "Completed: Exit context: `users`\nTask config is invalid: `Unable to setup context 'users': 'Failed to create the requested number of tenants.'`" | 18:34 |
| njohnston | This is against an Icehouse environment with Mitaka Rally; When I run Rally with debug logging I see: "2016-06-08 18:59:24.692 11197 ERROR rally.common.broker EndpointNotFound: admin endpoint for identity service in XXXX region not found" - but $OS_AUTH_URL is set in the rally deployment... http://paste.openstack.org/show/509002/ for the full log | 18:34 |
| *** banix has quit IRC | 18:38 | |
| *** banix has joined #openstack-rally | 18:43 | |
| *** smurke has quit IRC | 18:55 | |
| *** smurke has joined #openstack-rally | 18:57 | |
| *** e0ne has joined #openstack-rally | 18:59 | |
| *** e0ne has quit IRC | 19:02 | |
| KevinE_ | andreykurilin: you still around? :) | 19:06 |
| *** e0ne has joined #openstack-rally | 19:07 | |
| *** pvaneck has quit IRC | 19:10 | |
| *** harlowja has quit IRC | 19:12 | |
| *** pvaneck has joined #openstack-rally | 19:16 | |
| *** itlinux has joined #openstack-rally | 19:18 | |
| *** MaxPC has quit IRC | 19:36 | |
| *** MaxPC has joined #openstack-rally | 19:38 | |
| *** harlowja has joined #openstack-rally | 19:49 | |
| *** e0ne has quit IRC | 19:49 | |
| *** arnoldje has quit IRC | 19:55 | |
| *** MaxPC has quit IRC | 20:00 | |
| *** MaxPC has joined #openstack-rally | 20:02 | |
| *** arnoldje has joined #openstack-rally | 20:12 | |
| openstackgerrit | Kevin Esensoy proposed openstack/rally: [DO NOT MERGE] Add scenarios to VM https://review.openstack.org/326701 | 20:33 |
| *** rcernin has joined #openstack-rally | 20:36 | |
| *** MaxPC has quit IRC | 20:39 | |
| KevinE_ | Fault: {u'message': u'No valid host was found. There are not enough hosts available.', u'code': 500, u'created': u'2016-06-09T20:39:58Z'} | 20:40 |
| KevinE_ | anyone know what this error means? | 20:40 |
| openstackgerrit | Merged openstack/rally: Use Nova only when Nova is available https://review.openstack.org/327691 | 20:42 |
| *** akrzos has joined #openstack-rally | 20:43 | |
| *** banix has quit IRC | 20:51 | |
| *** rcernin has quit IRC | 21:15 | |
| *** dpaterson has quit IRC | 21:42 | |
| *** njohnston is now known as njohnston|afk | 21:49 | |
| openstackgerrit | Kevin Esensoy proposed openstack/rally: [DO NOT MERGE] Add scenarios to VM https://review.openstack.org/326701 | 21:50 |
| *** KevinE_ has quit IRC | 22:11 | |
| *** arnoldje has quit IRC | 22:13 | |
| andreykurilin__ | Anticimex: let me share some links which should describe how rally runs iterations in parallel | 22:15 |
| *** banix has joined #openstack-rally | 22:19 | |
| *** banix has quit IRC | 22:22 | |
| andreykurilin__ | Anticimex: https://github.com/openstack/rally/blob/master/rally/task/engine.py#L340-L348 this is a first part - obtain runner class(constant in your case) and execute run method on it. | 22:26 |
| andreykurilin__ | https://github.com/openstack/rally/blob/master/rally/plugins/common/runners/constant.py#L118-L119 is a constant runner. there is no run method, so it will use run method from parent class | 22:26 |
| andreykurilin__ | https://github.com/openstack/rally/blob/master/rally/task/runner.py#L154-L162 this method obtains class object of scenario and execute _run_scenario method of constant runner | 22:26 |
| andreykurilin__ | https://github.com/openstack/rally/blob/master/rally/plugins/common/runners/constant.py#L159 actually it is only about creating processes and threads. each iteration executes _run_scenario_once_with_unpack_args -> runner._run_scenario_once method | 22:26 |
| andreykurilin__ | https://github.com/openstack/rally/blob/master/rally/task/runner.py#L53 | 22:26 |
| andreykurilin__ | https://github.com/openstack/rally/blob/master/rally/task/runner.py#L62 as you can see, class of scenario is initialized on the last step. in each iteration we have separate instance of scenario class with own properties | 22:26 |
| Anticimex | cool, thanks | 22:27 |
| andreykurilin__ | Each iteration creates clients once and cache them. since instance of scenario class is not shared across iterations, cache is not shared too. | 22:27 |
| andreykurilin__ | Also cache means that for operations create and delete in one iteration, one client will be used | 22:28 |
| andreykurilin__ | so I can't imagine case when at create step we have user1 and at delete step we have user2 | 22:28 |
| andreykurilin__ | Anticimex: one more link:) https://github.com/openstack/rally/blob/master/rally/task/runner.py#L56-L57 input data is not shared too:) | 22:29 |
| trams | it would be too easy | 22:30 |
| trams | if it was only users | 22:30 |
| trams | and also mean that all concurrency would work if used just 1 tennant for all tests | 22:30 |
| trams | which doesnt avail the issue (but removing concurrency does) | 22:30 |
| trams | i will try to circle the issue more closely tomorrow | 22:31 |
| trams | and aslo write down what and what doesnt work and why for easy tracking | 22:31 |
| andreykurilin__ | "aslo write down what and what doesnt work and why for easy tracking" it can help | 22:31 |
| andreykurilin__ | I think you need to insert a lot of `LOG.debug` statements around the code | 22:32 |
| andreykurilin__ | but your issue sounds very strange | 22:32 |
| andreykurilin__ | https://review.openstack.org/#/c/324577/ <- an example of patch where you can see a lot of rally jobs which launch a lot of different scenarios. most of them uses constant runner with concurrency>1 | 22:34 |
| andreykurilin__ | so I'll waiting for more data tomorrow :) | 22:34 |
| *** abalutoiu has quit IRC | 22:35 | |
| Anticimex | andreykurilin__: thanks for the pointers :) | 22:36 |
| andreykurilin__ | Anticimex: you are welcome:) | 22:37 |
| *** MaxPC has joined #openstack-rally | 22:44 | |
| *** zackf has quit IRC | 22:58 | |
| *** davideagnello has joined #openstack-rally | 23:03 | |
| *** zackf has joined #openstack-rally | 23:04 | |
| *** davideagnello has quit IRC | 23:04 | |
| *** zackf has quit IRC | 23:08 | |
| *** MaxPC has quit IRC | 23:22 | |
| *** tosky has quit IRC | 23:43 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!