*** dave-mccowan has joined #openstack-rally | 00:09 | |
*** LindaWang has joined #openstack-rally | 00:44 | |
openstackgerrit | Boris Pavlovic proposed openstack/rally master: WIP: Add server console log to scenario output data https://review.openstack.org/483150 | 01:32 |
---|---|---|
openstackgerrit | Boris Pavlovic proposed openstack/rally master: WIP: Add server console log to scenario output data https://review.openstack.org/483150 | 01:37 |
*** psuriset has quit IRC | 01:42 | |
*** ilbot3 has quit IRC | 01:45 | |
a8K8U1LOP | [From Gitter] geyao0619 : Hi, I failed the 'gate-rally-dsvm-rally-heat’ for jenkins. | 01:48 |
a8K8U1LOP | [From Gitter] geyao0619 : I got ‘HTTPNotFound’. just wondering if‘recheck’ will solve it?http://logs.openstack.org/94/476694/9/check/gate-rally-dsvm-rally-heat/6f52057/rally-plot/results.html.gz#/HeatStacks.create_stack_and_scale-2/failures | 01:48 |
*** ilbot3 has joined #openstack-rally | 01:49 | |
*** itlinux has joined #openstack-rally | 02:10 | |
LindaWang | Hi guys | 02:31 |
LindaWang | When I run "rally verify configure-verifier" in rally 0.9.1 | 02:32 |
LindaWang | the param lock_path is missing in tempest.conf | 02:34 |
LindaWang | Cause there is no dir named .rally/verification/tempest/data/lock_files_uuid any more. | 02:34 |
LindaWang | So I am wondering what the param is for. | 02:35 |
LindaWang | 2 tests (tempest.api.compute.test_quotas.QuotasTestJSON.test_get_default_quotas[id-9bfecac7-b966-4f47-913f-1a9e2c12134a] | 02:35 |
LindaWang | tempest.api.compute.test_quotas.QuotasTestJSON.test_get_quotas[id-f1ef0a97-dbbb-4cca-adc5-c9fbc4f76107]) failed due to this issue | 02:35 |
LindaWang | Anybody would like to help? Thanks | 02:36 |
openstackgerrit | Merged openstack/rally master: [Install] Fix postgresql installation https://review.openstack.org/481998 | 02:36 |
openstackgerrit | chenhb proposed openstack/rally master: Fix an issue of database migration https://review.openstack.org/483163 | 02:49 |
*** aojea has joined #openstack-rally | 02:53 | |
*** aojea has quit IRC | 02:58 | |
a8K8U1LOP | [From Gitter] zhangzhihui1 : in my tempest.conf | 02:58 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : [oslo_concurrency] | 02:58 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : lock_path = /root/.rally/verification/verifier-0a571769-31d4-4ab9-ab28-34c88b9ac4c2/for-deployment-0052293e-7103-4bcb-8147-a97960e50c50/lock_files | 02:58 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : 'lock_files' is a dir. | 02:59 |
*** chhavi has joined #openstack-rally | 03:02 | |
openstackgerrit | chenhb proposed openstack/rally master: Fix an issue of database migration https://review.openstack.org/483163 | 03:17 |
*** dave-mccowan has quit IRC | 03:24 | |
*** psuriset has joined #openstack-rally | 03:52 | |
LindaWang | zhangzhihui1 Which version of rally do you use? | 04:00 |
LindaWang | a8K8U1LOP: zhangzhihui1 When I run tests in rally 0.8.1, the dir lock_files is there and the param lock_path is right | 04:02 |
LindaWang | zhangzhihui1 The issue is only for rally 0.9.1 | 04:03 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : @ LindaWang # rally version | 04:50 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : 0.9.1~dev280 | 04:50 |
*** anshul has joined #openstack-rally | 04:59 | |
*** aojea has joined #openstack-rally | 05:28 | |
*** psuriset has quit IRC | 05:30 | |
*** rcernin has joined #openstack-rally | 05:40 | |
openstackgerrit | chenhb proposed openstack/rally master: Fix an issue of database migration https://review.openstack.org/483163 | 05:46 |
*** psuriset has joined #openstack-rally | 05:47 | |
*** aojea has quit IRC | 05:53 | |
*** lpetrut has joined #openstack-rally | 06:06 | |
*** psuriset has quit IRC | 06:48 | |
*** pcaruana has joined #openstack-rally | 06:52 | |
*** tesseract has joined #openstack-rally | 06:56 | |
*** psuriset has joined #openstack-rally | 07:01 | |
*** lpetrut has quit IRC | 07:07 | |
LindaWang | zhangzhihui1 I doubt if the way I installed rally was wrong. | 07:14 |
LindaWang | zhangzhihui1 I did not use the install_rally.sh to install | 07:16 |
*** aojea has joined #openstack-rally | 07:17 | |
LindaWang | zhangzhihui1 What I use is these lines below: | 07:19 |
LindaWang | mkdir -p /etc/rally && \ | 07:19 |
LindaWang | printf "[database]\nconnection = 'sqlite:////var/lib/rally/database/rally.sqlite'" > /etc/rally/rally.conf && \ | 07:19 |
LindaWang | mkdir -p /var/lib/rally/database && rally-manage db create | 07:19 |
LindaWang | zhangzhihui1 Do you know if it is right? | 07:20 |
*** alexchadin has joined #openstack-rally | 07:25 | |
LindaWang | zhangzhihui1 Or, could you help test rally 0.9.1 locally? I see your rally version is master. | 07:27 |
openstackgerrit | sudhir agarwal proposed openstack/rally master: Replaces uuid.uuid4 with uuidutils.generate_uuid() https://review.openstack.org/483239 | 07:29 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : do you want to let me to run tempest.api.compute.test_quotas.QuotasTestJSON.test_get_default_quotas? | 07:32 |
*** lpetrut has joined #openstack-rally | 07:34 | |
LindaWang | No, after deployment, just run "rally verify configrue-verifier" | 07:34 |
LindaWang | To check is the dir .rally/verification/tempest/data/lock_files_uuid exists | 07:35 |
LindaWang | So also check if the param lock_path in tempest.conf has been configured correctly. | 07:35 |
*** jose_lausuch has joined #openstack-rally | 07:37 | |
*** psuriset has quit IRC | 07:38 | |
*** lpetrut has quit IRC | 07:43 | |
*** itlinux has quit IRC | 07:47 | |
openstackgerrit | Andrey Kurilin proposed openstack/rally master: [ci] Turn off glance scenarios https://review.openstack.org/483247 | 07:49 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : ok | 07:54 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : .rally/verification/tempest/data/lock_files_uuid: No such file or directory, but test is success | 07:55 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : lock_path = /root/.rally/verification/verifier-0a571769-31d4-4ab9-ab28-34c88b9ac4c2/for-deployment-804455f2-9fec-497c-a840-72186052cb94/lock_files | 07:55 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : tempest.api.compute.test_quotas.QuotasTestJSON.test_get_quotas ... success [2.223s] | 07:56 |
LindaWang | Then, could you help check if the way I insatlled rally is correct? | 07:57 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : and i don't know if the way you installed rally is right. we can ask @andreykurilin | 07:59 |
LindaWang | zhangzhihui1 Fine. Thanks | 07:59 |
LindaWang | andreykurilin: Could you help check the way I installed rally is right? I did not user install_rally.sh | 08:05 |
LindaWang | andreykurilin: | 08:06 |
LindaWang | mkdir -p /etc/rally && \ | 08:06 |
LindaWang | printf "[database]\nconnection = 'sqlite:////var/lib/rally/database/rally.sqlite'" > /etc/rally/rally.conf && \ | 08:06 |
LindaWang | mkdir -p /var/lib/rally/database && rally-manage db create | 08:06 |
*** serlex has joined #openstack-rally | 08:18 | |
*** alexchadin has quit IRC | 08:29 | |
LindaWang | zhangzhihui1 There is no such dir in my system: /root/.rally/verification/verifier-uuid/for-deployment-uuid/lock_files | 08:32 |
LindaWang | zhangzhihui1 Actually there is no dir named lock_files anywhere. | 08:32 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : when run tests, the dir will be created | 08:34 |
LindaWang | I thought it was created by the cli "rally verify configrue-verifier" | 08:35 |
LindaWang | But why it failed to run tempest.api.compute.test_quotas.QuotasTestJSON.test_get_default_quotas, complaining that: oslo_config.cfg.RequiredError: value required for option lock_path in group [DEFAULT] | 08:38 |
*** e0ne has joined #openstack-rally | 08:38 | |
a8K8U1LOP | [From Gitter] andreykurilin : @geyao0619 hi! sorry, several other OpenStack projects broke our gates. We are trying to localize problems now. | 08:40 |
a8K8U1LOP | [From Gitter] andreykurilin : @LindaWang, @zhangzhihui1 : hi! | 08:42 |
LindaWang | Hi andreykurilin | 08:43 |
a8K8U1LOP | [From Gitter] zhangzhihui1 : Actually we can set lock_path by ourself, if without setting lock_path, rally will auto create. | 08:43 |
a8K8U1LOP | [From Gitter] andreykurilin : rally can be installed manually and by the helper script. it doesnt metter for rally code itself | 08:43 |
a8K8U1LOP | [From Gitter] andreykurilin : about lock_path: it is the same as image and some other options. Rally creates the directory right before running actual verification | 08:44 |
a8K8U1LOP | [From Gitter] andreykurilin : can you share an error of failed test? | 08:44 |
LindaWang | andreykurilin: http://paste.openstack.org/raw/615232/ | 08:48 |
*** ushkalim has joined #openstack-rally | 08:55 | |
*** alexchadin has joined #openstack-rally | 08:58 | |
openstackgerrit | chenhb proposed openstack/rally master: Print the messages what can be traced back to real address https://review.openstack.org/483271 | 09:02 |
*** oanufrii has joined #openstack-rally | 09:03 | |
a8K8U1LOP | [From Gitter] andreykurilin : @LindaWang: do you use the latest tempest? | 09:09 |
LindaWang | andreykurilin: tempest 14.0.0 | 09:09 |
LindaWang | rally 0.9.1 | 09:10 |
a8K8U1LOP | [From Gitter] andreykurilin : ok, it looks like tempest team changed the location of the options in 15+ and our code sets the option in new way. Can you check that you have oslo_concurrency section with lock_path variable in the config? | 09:11 |
*** tosky has joined #openstack-rally | 09:12 | |
a8K8U1LOP | [From Gitter] andreykurilin : Folks: our gates should be unblocked right after https://review.openstack.org/483247 | 09:13 |
LindaWang | a8K8U1LOP: andreykurilin: the var lock_path in session oslo_concurrency is in tempest.conf | 09:20 |
*** a8K8U1LOP has quit IRC | 09:20 | |
*** rallydev-bot has joined #openstack-rally | 09:21 | |
rallydev-bot | [From Gitter] andreykurilin : In this case, I have to solutions: use tempest 15+ :) propose a patch, which will set an old variable too here - https://github.com/openstack/rally/blob/master/rally/plugins/openstack/verification/tempest/context.py#L69-L70 | 09:22 |
LindaWang | andreykurilin: The reason we Functest team uses tempest 14.0.0 is that the upper stream openstack Ocata recommends that version | 09:26 |
LindaWang | andreykurilin: https://git.openstack.org/cgit/openstack/requirements/plain/upper-constraints.txt?h=stable/ocata | 09:26 |
*** mvk has quit IRC | 09:28 | |
rallydev-bot | [From Gitter] andreykurilin : LindaWang: can you update propose a patch to rally for setting the right option? | 09:36 |
LindaWang | andreykurilin: I have no idea how rally handles that option with different versions of tempest? | 09:39 |
rallydev-bot | [From Gitter] andreykurilin : I think any version of tempest will no fail if additional options,which are not used, will be specified | 09:40 |
rallydev-bot | [From Gitter] andreykurilin : so we can specify two options | 09:40 |
rallydev-bot | [From Gitter] andreykurilin : old one and new one | 09:40 |
rallydev-bot | [From Gitter] andreykurilin : and everything should work | 09:40 |
LindaWang | So could you help with that when you are available? I'd really appreciate it. | 09:50 |
*** mvk has joined #openstack-rally | 09:55 | |
*** alexchadin has quit IRC | 10:06 | |
rallydev-bot | [From Gitter] chenhb-zte : @andreykurilin yup, saw the patch, expect gates will be right. | 10:18 |
rallydev-bot | [From Gitter] andreykurilin : LindaWang: I'm available now and ready to help you to make a patch to rally :) | 10:20 |
LindaWang | andreykurilin: Thanks | 10:20 |
openstackgerrit | Merged openstack/rally master: [ci] Turn off glance scenarios https://review.openstack.org/483247 | 10:21 |
LindaWang | andreykurilin: A quick question: Is the var floating_network_name in tempest.conf supported by rally 0.9.1+? | 10:24 |
rallydev-bot | [From Gitter] zhangzhihui1 : @andreykurilin https://review.openstack.org/#/c/433013/ give us some advice | 10:26 |
rallydev-bot | [From Gitter] MaXinjian : @andreykurilin https://review.openstack.org/#/c/421088/ I have beem stuck on this case for a long time... | 10:28 |
rallydev-bot | [From Gitter] andreykurilin : LndaWang: yes. if rally finds a public network while config generation, it puts it into the config as floating_network_name | 10:28 |
rallydev-bot | [From Gitter] andreykurilin : @MaXinjian @zhangzhihui1 sorry for the delay, we spent time trying to fix our CI. will comment these patches today | 10:29 |
LindaWang | andreykurilin: Is the var floating_network_name also influenced by tempest version? | 10:29 |
*** yamamoto has quit IRC | 10:31 | |
rallydev-bot | [From Gitter] andreykurilin : I do not think so | 10:31 |
LindaWang | andreykurilin: ok | 10:34 |
*** mvk has quit IRC | 10:34 | |
LindaWang | andreykurilin: Is floating_network_name exactly the name of public_network_id in tempest.conf? | 10:35 |
LindaWang | andreykurilin: I find in some certain SUTs, public_network_id exists which floating_network_name is missing in rally 0.9.1 | 10:36 |
rallydev-bot | [From Gitter] andreykurilin : hm | 10:36 |
rallydev-bot | [From Gitter] andreykurilin : do you have any public networks there? | 10:36 |
LindaWang | andreykurilin: yes, | 10:37 |
rallydev-bot | [From Gitter] andreykurilin : so rally lists all networks via neutron, then filter them by active status and external router | 10:37 |
LindaWang | andreykurilin: Cause public_network_id is configured correctly | 10:37 |
rallydev-bot | [From Gitter] andreykurilin : if such networks are available, rally takes the first one, puts its id to public_network_id and its name into floating_network_name | 10:38 |
rallydev-bot | [From Gitter] andreykurilin : if public_network_id is correct, the floating_network_name should be configured too. they use the same if check for setting | 10:39 |
LindaWang | Ok, I will check that later. | 10:40 |
LindaWang | andreykurilin: Do you still remember the issue about admin role when running rally? | 10:41 |
LindaWang | andreykurilin: I sent you one email 4 days ago. | 10:41 |
rallydev-bot | [From Gitter] andreykurilin : wow, sorry, It looks like I forgot to send a reply | 10:42 |
LindaWang | andreykurilin: Talk to you later. A conference offline to join now. | 10:45 |
*** lpetrut has joined #openstack-rally | 10:45 | |
LindaWang | andreykurilin: I am wondering if the SUT has strict limit about resource management, and only the user with admin role can be allowed to take actions | 10:46 |
LindaWang | andreykurilin: Then how can I run rally on the SUT? | 10:47 |
rallydev-bot | [From Gitter] andreykurilin : first of all, it sounds really weird. I mean it is a hard to understand such decision(restriction). the seconde question: I do not know what is SUT, so please explain :) | 10:48 |
rallydev-bot | [From Gitter] andreykurilin : In general I see two options | 10:48 |
rallydev-bot | [From Gitter] andreykurilin : 1) you can use roles context, which allows to assign any role to new users. 2) use existing users with the right roles | 10:49 |
rallydev-bot | [From Gitter] andreykurilin : BUT, as far as I know, the error which you have ususally raises when the member role is not proper configured in the cloud | 10:49 |
*** mvk has joined #openstack-rally | 10:50 | |
rallydev-bot | [From Gitter] andreykurilin : to check this, you should try to create a new project, a new user in it and assign a member role (this step is quite important). then try to perform actions from this user | 10:50 |
*** jkilpatr has quit IRC | 11:01 | |
*** aojea has quit IRC | 11:04 | |
*** aojea has joined #openstack-rally | 11:06 | |
*** aojea has quit IRC | 11:09 | |
*** yamamoto has joined #openstack-rally | 11:13 | |
*** mvk has quit IRC | 11:13 | |
*** yamamoto has quit IRC | 11:18 | |
*** yamamoto has joined #openstack-rally | 11:21 | |
*** mvk has joined #openstack-rally | 11:26 | |
*** yamamoto has quit IRC | 11:26 | |
*** psuriset has joined #openstack-rally | 11:27 | |
*** yamamoto has joined #openstack-rally | 11:36 | |
*** e0ne_ has joined #openstack-rally | 11:37 | |
*** lpetrut_ has joined #openstack-rally | 11:38 | |
*** e0ne has quit IRC | 11:39 | |
*** jkilpatr has joined #openstack-rally | 11:39 | |
*** yamamoto has quit IRC | 11:40 | |
*** lpetrut has quit IRC | 11:41 | |
*** yamamoto has joined #openstack-rally | 11:50 | |
*** yamamoto has quit IRC | 11:54 | |
openstackgerrit | bhavani proposed openstack/rally master: Add a hacking rule for string interpolation at logging https://review.openstack.org/483314 | 11:57 |
*** dave-mccowan has joined #openstack-rally | 12:01 | |
*** reveller has quit IRC | 12:12 | |
*** yamamoto has joined #openstack-rally | 12:17 | |
openstackgerrit | Andrey Kurilin proposed openstack/rally master: [cli] "rally task list" improvements https://review.openstack.org/482473 | 12:21 |
openstackgerrit | bhavani proposed openstack/rally master: Handle log message interpolation by the logger in common/ https://review.openstack.org/483314 | 12:28 |
*** alexchadin has joined #openstack-rally | 12:48 | |
*** catintheroof has joined #openstack-rally | 12:56 | |
*** deep-book-gk_ has joined #openstack-rally | 13:00 | |
*** deep-book-gk_ has left #openstack-rally | 13:01 | |
*** reveller has joined #openstack-rally | 13:07 | |
*** alexchadin has quit IRC | 13:09 | |
*** LindaWang has quit IRC | 13:13 | |
*** LindaWang has joined #openstack-rally | 13:14 | |
LindaWang | \ | 13:18 |
*** aojea has joined #openstack-rally | 13:24 | |
*** LindaWang has quit IRC | 13:25 | |
*** aojea has quit IRC | 13:25 | |
*** aojea has joined #openstack-rally | 13:25 | |
openstackgerrit | Andrey Kurilin proposed openstack/rally master: [cli] "rally task list" improvements https://review.openstack.org/482473 | 13:39 |
*** jkilpatr has quit IRC | 13:42 | |
*** jkilpatr has joined #openstack-rally | 13:47 | |
*** LindaWang has joined #openstack-rally | 14:07 | |
*** e0ne_ has quit IRC | 14:16 | |
*** aarefiev is now known as aarefiev_afk | 14:21 | |
*** itlinux_ has joined #openstack-rally | 14:22 | |
*** e0ne has joined #openstack-rally | 14:24 | |
*** LindaWang has quit IRC | 14:31 | |
*** LindaWang has joined #openstack-rally | 14:31 | |
openstackgerrit | Andrey Kurilin proposed openstack/rally master: Remove existing_users context https://review.openstack.org/458810 | 14:31 |
*** jose_lausuch has quit IRC | 14:34 | |
LindaWang | andreykurilin: The SUT I am testing is one oommercial version of Openstack. | 14:35 |
LindaWang | andreykurilin: The SUT I am testing is one commercial version of Openstack. | 14:36 |
LindaWang | andreykurilin: 1) Where is the roles context that can be used? | 14:36 |
rallydev-bot | [From Gitter] andreykurilin : Are you sure that regular user which cannot do anything is an expected behaviuor of that cloud? | 14:37 |
LindaWang | Yes, I am sure. The new user with admin role can be allowed to take actions in that SUT | 14:41 |
*** itlinux_ has quit IRC | 14:42 | |
rallydev-bot | [From Gitter] andreykurilin : I'm not talking about a new user with admin role. It is logical that a user with admin role can do anything :) I wonder just about regular users. Normally, new user without any roles cannot perform any action is the cloud (like in your case) and after applying the magic role member the new user obtains enough rights for doing something in the cloud | 14:43 |
rallydev-bot | [From Gitter] andreykurilin : ok, if you think that your case is not a bug, but a feature, I'm ok too :) | 14:44 |
*** serlex has quit IRC | 14:44 | |
rallydev-bot | [From Gitter] andreykurilin : so let's talk who we can interact with such case | 14:45 |
LindaWang | What I mean is that the new user without admin role can not be allowed to do anything | 14:45 |
rallydev-bot | [From Gitter] andreykurilin : ok | 14:45 |
*** anshul has quit IRC | 14:47 | |
rallydev-bot | [From Gitter] andreykurilin : solution #1: rally can create new users for workload or use existing users. In this case you just need to create several test users with enough rights to perform actions in the cloud and specify that users in rally deployment config. | 14:47 |
rallydev-bot | [From Gitter] andreykurilin : An example: https://github.com/openstack/rally/blob/master/samples/deployments/existing-with-predefined-users.json#L13-L23 (note: if you use keystone v3, specify users with using keystone v3 terms) | 14:48 |
rallydev-bot | [From Gitter] andreykurilin : solution #2 : use roles context. In this case rally will create users and then apply specified roles for you | 14:49 |
rallydev-bot | [From Gitter] andreykurilin : https://rally.readthedocs.io/en/latest/plugins/plugin_reference.html#roles-context | 14:49 |
rallydev-bot | [From Gitter] andreykurilin : An example: https://github.com/openstack/rally/blob/master/samples/tasks/contexts/roles.json#L17 | 14:49 |
rallydev-bot | [From Gitter] andreykurilin : NOTE: the first solution will work for all rally scenarios and context. The second solution can led to some issues at validation step | 14:50 |
LindaWang | Actually I use only on user(admin) in the creds when I created rally deployment | 14:50 |
LindaWang | Actually I use only one user(admin) in the creds when I created rally deployment | 14:50 |
rallydev-bot | [From Gitter] andreykurilin : yes. specifing just one user(admin) is a case when rally creates new users for workload | 14:52 |
LindaWang | I think the user named admin is only for creating the rally deployment. | 14:54 |
LindaWang | When it comes to different tests, new different users will be created then. | 14:55 |
rallydev-bot | [From Gitter] andreykurilin : yes | 14:55 |
*** ushkalim has quit IRC | 14:56 | |
LindaWang | For solution 1, the additional added users in the file existing-with-predefined-users.json, will be used when running all scenarios? If so, there will be no new users created, right? | 14:58 |
rallydev-bot | [From Gitter] andreykurilin : yes | 14:58 |
LindaWang | So in my case here, i have to assign admin role to these users beforehand? | 14:59 |
rallydev-bot | [From Gitter] andreykurilin : yup | 14:59 |
LindaWang | How to guarantee the number of additional added users in the file existing-with-predefined-users.json is enough for all these tests? | 15:00 |
*** e0ne has quit IRC | 15:01 | |
rallydev-bot | [From Gitter] andreykurilin : LindaWang: ok. so rally takes a random user from the list. If it will be just one users, all scenarios and all iterations will be executed from it. If you have more users, the usage will be fuzzy. But again, even one user will be enough. NOTE: you need to remove "users" context section from the task config in such case. If "users" sections is present there, rally eng | 15:02 |
rallydev-bot | ine will force to create new users | 15:02 |
LindaWang | Now I prefer to solution 2 | 15:04 |
rallydev-bot | [From Gitter] andreykurilin : heh | 15:04 |
LindaWang | But I have no idea when I run these command: http://paste.openstack.org/raw/615280/ | 15:05 |
rallydev-bot | [From Gitter] andreykurilin : you need to modify /home/opnfv/repos/functest/functest/opnfv_tests/openstack/rally/task.yaml | 15:06 |
LindaWang | How can I configure the roles.json file? | 15:06 |
*** serlex has joined #openstack-rally | 15:06 | |
LindaWang | https://github.com/opnfv/functest/blob/master/functest/opnfv_tests/openstack/rally/task.yaml | 15:06 |
LindaWang | No roles can be configured in this file. | 15:07 |
*** serlex has left #openstack-rally | 15:07 | |
rallydev-bot | [From Gitter] andreykurilin : Actually, it would be much easier to you to return to solution #1 | 15:08 |
LindaWang | Well, fine :) | 15:09 |
*** e0ne has joined #openstack-rally | 15:09 | |
rallydev-bot | [From Gitter] andreykurilin : Each of workload(scenario) can require theoretically different roles, so the roles context should be specified for each workload separately. Based on the fact that your task config is splitted accross a lot of files (see https://github.com/opnfv/functest/tree/master/functest/opnfv_tests/openstack/rally/scenario), it will take much time | 15:09 |
LindaWang | I am just wondering how to realize solution #2 in functest | 15:09 |
rallydev-bot | [From Gitter] andreykurilin : you will need to add roles context like in example to each of context sections here - https://github.com/opnfv/functest/blob/master/functest/opnfv_tests/openstack/rally/scenario/opnfv-authenticate.yaml#L3-L4 | 15:10 |
rallydev-bot | [From Gitter] andreykurilin : and to other files | 15:10 |
rallydev-bot | [From Gitter] andreykurilin : As you can see, you task config uses macros and it supports existing users. It means that in case of solution #1, you will need just to create new users transmit them into rally deployment and transmit into task a argument to use existing users | 15:12 |
LindaWang | What is the purpose of the var tmpl_dir and sup_dir? ('tmpl_dir': '/home/opnfv/repos/functest/functest/opnfv_tests/openstack/rally/scenario/templates', 'sup_dir': '/home/opnfv/repos/functest/functest/opnfv_tests/openstack/rally/scenario/support') | 15:12 |
rallydev-bot | [From Gitter] andreykurilin : I do not know what is it) | 15:13 |
LindaWang | https://github.com/opnfv/functest/tree/master/functest/opnfv_tests/openstack/rally/scenario/templates | 15:13 |
rallydev-bot | [From Gitter] andreykurilin : I think it is templates for heat | 15:14 |
LindaWang | I thought these 2 params (tmpl_dir and sup_dir) were required by rally | 15:14 |
rallydev-bot | [From Gitter] andreykurilin : possibly, these parameters are required by you script, but not rally itself | 15:19 |
*** yingjun has joined #openstack-rally | 15:20 | |
*** jkilpatr has quit IRC | 15:23 | |
*** rcernin has quit IRC | 15:27 | |
LindaWang | Do you mean these two lines should be removed for solution #1 https://github.com/opnfv/functest/blob/master/functest/opnfv_tests/openstack/rally/scenario/opnfv-authenticate.yaml#L3-L4 | 15:27 |
rallydev-bot | [From Gitter] andreykurilin : no | 15:28 |
LindaWang | Remove the session of context for all scenario? | 15:28 |
rallydev-bot | [From Gitter] andreykurilin : let me try to explain in another way | 15:28 |
LindaWang | Thanks, I just can not understand this sentence: you need to remove "users" context section from the task config in such case. If "users" sections is present there, rally engine will force to create new users | 15:29 |
rallydev-bot | [From Gitter] andreykurilin : an example of rally task which should launch two scenarios - http://xsnippet.org/362640/ | 15:32 |
rallydev-bot | [From Gitter] andreykurilin : both of scenarios has context section with specified users ctx | 15:33 |
rallydev-bot | [From Gitter] andreykurilin : if you launch that scenario, rally will not even look at existance of real users in deployment config, the new users will be created, because it is specified in the json | 15:34 |
LindaWang | YEP | 15:34 |
rallydev-bot | [From Gitter] andreykurilin : "users": {"tenants": 2, "users_per_tenant": 2} means that user asks rally to create 2 tenants with 2 users in it, so rally will do it for you. | 15:35 |
rallydev-bot | [From Gitter] andreykurilin : an another example - http://xsnippet.org/362641/ | 15:35 |
rallydev-bot | [From Gitter] andreykurilin : the first workload is not changed | 15:35 |
rallydev-bot | [From Gitter] andreykurilin : rally will again ignore "users" section of deployment config and create new users | 15:36 |
rallydev-bot | [From Gitter] andreykurilin : BUT the config for the second workload is changed | 15:36 |
rallydev-bot | [From Gitter] andreykurilin : there is no users context | 15:36 |
rallydev-bot | [From Gitter] andreykurilin : in this case rally will check a deployment and if "users" are there, they will be used for workload. If not -> the new ones will be created | 15:37 |
LindaWang | So my question is how to modify this file https://github.com/opnfv/functest/blob/master/functest/opnfv_tests/openstack/rally/scenario/opnfv-authenticate.yaml#L3-L4 to avoid new user created by rally? | 15:38 |
*** pcaruana has quit IRC | 15:38 | |
rallydev-bot | [From Gitter] andreykurilin : as from the previous example, it is become obvious that each workload have separated contexts, so you need(do not do it before reading the next my message) to remove all contexts sections - https://github.com/opnfv/functest/blob/master/functest/opnfv_tests/openstack/rally/scenario/opnfv-authenticate.yaml#L3-L4 and https://github.com/opnfv/functest/blob/master/functest | 15:40 |
rallydev-bot | /opnfv_tests/openstack/rally/scenario/opnfv-authenticate.yaml#L14-L15 and | 15:40 |
rallydev-bot | [From Gitter] andreykurilin : so on | 15:40 |
rallydev-bot | [From Gitter] andreykurilin : Here is another thing that you should know about rally -> task configs are not just regular JSON/YAML files, they are actually jinja2 templates, which allows to optimize some things, split one big task config into separate files, add if...else blocks, etc | 15:41 |
rallydev-bot | [From Gitter] andreykurilin : the author of https://github.com/opnfv/functest/tree/master/functest/opnfv_tests/openstack/rally knew it and make the task config configurable. A lot of things can be configured there. And one of those things - where to add users context to all workloads or not | 15:42 |
LindaWang | Ok, got it. For authenticate sceanrio, remove all contexts sections in opnfv-authenticate.yaml, right? | 15:43 |
rallydev-bot | [From Gitter] andreykurilin : so you do not need to remove anything, you just need to transmit the right argument into the task, so existing users will be used | 15:43 |
rallydev-bot | [From Gitter] andreykurilin : It looks like you need just to transmit "use_existing_users" into task-args here - http://paste.openstack.org/raw/615280/ | 15:44 |
LindaWang | ok, I will try that | 15:45 |
LindaWang | Thanks for your detailed explanation and patience. :) | 15:47 |
openstackgerrit | Andrey Kurilin proposed openstack/rally master: [wip][cli] "rally task list" improvements https://review.openstack.org/482473 | 15:52 |
LindaWang | I will try to update the patch about lock_path, if we have to use tempest 14.0.0 | 15:55 |
*** e0ne has quit IRC | 15:55 | |
*** aojea has quit IRC | 15:55 | |
*** aojea has joined #openstack-rally | 16:02 | |
*** LindaWang has quit IRC | 16:04 | |
*** LindaWang has joined #openstack-rally | 16:04 | |
*** LindaWang has quit IRC | 16:06 | |
*** aojea has quit IRC | 16:07 | |
*** aojea has joined #openstack-rally | 16:12 | |
*** aojea has quit IRC | 16:16 | |
*** tesseract has quit IRC | 16:19 | |
*** aojea has joined #openstack-rally | 16:21 | |
*** aojea has quit IRC | 16:25 | |
*** aojea has joined #openstack-rally | 16:30 | |
*** aojea has quit IRC | 16:36 | |
*** aojea has joined #openstack-rally | 16:39 | |
*** lpetrut_ has quit IRC | 16:42 | |
*** aojea has quit IRC | 16:44 | |
*** itlinux has joined #openstack-rally | 16:45 | |
*** aojea has joined #openstack-rally | 16:48 | |
*** aojea has quit IRC | 16:52 | |
*** aojea has joined #openstack-rally | 16:57 | |
*** aojea has quit IRC | 17:02 | |
*** oanufrii has quit IRC | 17:02 | |
*** aojea has joined #openstack-rally | 17:06 | |
*** e0ne has joined #openstack-rally | 17:09 | |
*** aojea has quit IRC | 17:10 | |
*** aojea has joined #openstack-rally | 17:15 | |
*** jkilpatr has joined #openstack-rally | 17:18 | |
*** aojea has quit IRC | 17:20 | |
*** aojea has joined #openstack-rally | 17:24 | |
*** jkilpatr has quit IRC | 17:26 | |
*** harlowja has quit IRC | 17:27 | |
*** aojea has quit IRC | 17:29 | |
*** yingjun has quit IRC | 17:29 | |
*** tosky has quit IRC | 17:34 | |
*** mvk has quit IRC | 18:01 | |
*** jkilpatr has joined #openstack-rally | 18:16 | |
*** e0ne has quit IRC | 18:16 | |
*** mvk has joined #openstack-rally | 18:35 | |
*** pcaruana has joined #openstack-rally | 18:38 | |
*** harlowja has joined #openstack-rally | 18:54 | |
*** chandankumar has quit IRC | 19:00 | |
*** tosky has joined #openstack-rally | 19:06 | |
*** chhavi has quit IRC | 19:38 | |
*** pcaruana has quit IRC | 19:52 | |
*** harlowja has quit IRC | 20:19 | |
*** lpetrut_ has joined #openstack-rally | 20:21 | |
*** harlowja has joined #openstack-rally | 20:22 | |
*** aojea has joined #openstack-rally | 20:35 | |
*** aojea has quit IRC | 20:41 | |
*** lpetrut_ has quit IRC | 22:06 | |
*** dpaterson has quit IRC | 22:17 | |
*** dpaterson has joined #openstack-rally | 22:20 | |
*** itlinux has quit IRC | 22:58 | |
*** tosky has quit IRC | 23:00 | |
*** catintheroof has quit IRC | 23:03 | |
*** aojea has joined #openstack-rally | 23:28 | |
*** aojea has quit IRC | 23:33 | |
*** aojea has joined #openstack-rally | 23:37 | |
*** aojea has quit IRC | 23:41 | |
*** aojea has joined #openstack-rally | 23:46 | |
*** aojea has quit IRC | 23:51 | |
*** itlinux has joined #openstack-rally | 23:56 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!