| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-03-08 | |||
| 07:13:54 | Qiming | not the key issue I want to raise here | |
| 07:14:15 | alex_xu | Qiming: but thanks btw | |
| 07:14:48 | Qiming | when I tried 'curl -H 'openstack-api-version: compute 2.3' .... I can get the response with "OS-EXT-SRV-ATTR:user_data" returned | |
| 07:14:49 | openstackgerrit | Chris Friesen proposed openstack/nova-specs master: add spec for instance flavor API https://review.openstack.org/265282 | |
| 07:15:11 | alex_xu | yea | |
| 07:15:26 | Qiming | when I tried openstackclient with --os-compute-api 2.3 --debug, this field and other fields added since 2.3 are missing | |
| 07:15:49 | alex_xu | Qiming: i don't openstackclient support microvresion now | |
| 07:16:03 | alex_xu | s/i don't/i don't think/ | |
| 07:17:03 | Qiming | em... nova responded this in the header: Content-Length: 1464 Content-Type: application/json Openstack-Api-Version: compute 2.3 X-Openstack-Nova-Api-Version: 2.3 Vary: OpenStack-API-Version, X-OpenStack-Nova-API-Version X-Compute-Request-Id: req-efc2a8c9-b37b-43e9-a170-73e1658b3177 Date: Wed, 08 Mar 2017 07:16:20 GMT Connection: keep-alive | |
| 07:17:20 | Qiming | this means ... it should work, right? | |
| 07:18:00 | alex_xu | Qiming: yes...that means OSC request 2.3, and the server response with 2.3 | |
| 07:18:22 | alex_xu | Qiming: but I'm not sure OSC handle 2.3 in the CLI display | |
| 07:18:40 | Qiming | so I think nova-api understands the micro-version api request properly | |
| 07:19:09 | Qiming | I don't care about the CLI output, so I turned on --debug, I want to check the raw response body | |
| 07:19:13 | alex_xu | Qiming: there is talk about implement microversion OSC, so i think OSC didn't implement that | |
| 07:19:32 | alex_xu | Qiming: no new fields in the raw response body? | |
| 07:19:40 | Qiming | yup | |
| 07:19:55 | alex_xu | strange, let me check | |
| 07:20:27 | Qiming | so I am here complaining to you, instead of complaining in another channel | |
| 07:20:57 | Qiming | I was suspecting that nova-api treats 'User-Agent' header in different (buggy) ways | |
| 07:21:07 | Qiming | but that is not true | |
| 07:21:47 | Qiming | I was doubting the 'X-Auth-Token' is causing some problems ... seems that is relevant | |
| 07:22:19 | Qiming | If I'm passing token in this format: -H "X-Auth-Token: {SHA1}e216eff934c3484e8bc2cf1642c736e215c8c369" | |
| 07:22:29 | alex_xu | Qiming: Openstack-API-Version header returned, that means nova-api already know the requested version | |
| 07:22:32 | Qiming | I can never get the fields from the response | |
| 07:22:42 | alex_xu | emm... | |
| 07:22:46 | alex_xu | Qiming: By curl | |
| 07:22:48 | alex_xu | ? | |
| 07:23:17 | Qiming | if I'm passing token in this way: 'X-Auth-Token: 6c9bce0e3a514ad7981fa03f990e5cfc' | |
| 07:23:24 | Qiming | I can get the correct result | |
| 07:23:29 | Qiming | see the difference? | |
| 07:23:30 | alex_xu | Qiming: I guess you can't pass taken with hash token by curl | |
| 07:23:59 | Qiming | that is what I'm seeing from the keystoneauth1 log | |
| 07:24:28 | alex_xu | Qiming: the hash is done by the keystoneauth, for hide the secret thing. if you paste that into the curl, you can't pass the auth | |
| 07:24:39 | Qiming | yes, I realized that | |
| 07:25:02 | Qiming | I was comparing the ways requests are sent from OSC and from cURL | |
| 07:25:38 | Qiming | I believe you can reproduce this easily? | |
| 07:26:34 | alex_xu | Qiming: yea, i'm trying | |
| 07:26:58 | Qiming | thanks! | |
| 07:28:01 | arne_r | If someone could help me get this change into newton, newton is broken on our system without it: https://review.openstack.org/#/c/436868/ | |
| 07:28:09 | openstackgerrit | jichenjc proposed openstack/nova master: Clean up metadata param in doc https://review.openstack.org/442981 | |
| 07:29:51 | alex_xu | Qiming: your command like this "openstack --os-compute-api-version 2.3 --debug server show 940562c1-71c6-4dcf-83be-eeb2a22b5ca0"? | |
| 07:30:02 | Qiming | yep | |
| 07:30:18 | alex_xu | Qiming: I can see... the user_data field | |
| 07:31:17 | Qiming | oh? what's your magic? | |
| 07:31:26 | alex_xu | Qiming: both from the cli display and raw data | |
| 07:32:17 | alex_xu | Qiming: can you sure you request is send to a v2.1 endpoint, not the legacy v2 endpoint | |
| 07:32:59 | Qiming | from debug log, I can see the request sent (by OSC) this way: curl -g -i -X GET http://192.168.42.11:8774/v2.1/servers/59a17d5a-c27a-4301-b6f2-19dd128f24df -H "openstack-api-version: compute 2.3" -H "User-Agent: openstacksdk/0.9.14 keystoneauth1/2.18.0 python-requests/2.12.4 CPython/2.7.12" -H "X-Auth-Token: {SHA1}e216eff934c3484e8bc2cf1642c736e215c8c369" | |
| 07:33:50 | alex_xu | Qiming: looks like it is 2.1 endpoint | |
| 07:33:55 | alex_xu | my one is same "curl -g -i -X GET http://hp-pc:8774/v2.1/servers/940562c1-71c6-4dcf-83be-eeb2a22b5ca0 -H "User-Agent: python-novaclient" -H "Accept: application/json" -H "X-OpenStack-Nova-API-Version: 2.3" -H "X-Auth-Token: | |
| 07:33:57 | alex_xu | gAAAAABYv7cy3iWBliuTe8Htsg6-G8bMGJPYMAAW2H55dUBMOYuNzh-uP5o5p0DqxymiA6qqaBAdNPp1bRtJmN9Vv7q_cnDVAxk3nuSA4B_1d0v4opEl3EIWwMvT1SbDlFpwIVcAONtXfz5iV_pA86ZtKqgttDuCuzwb5CA1dQ7iKXpUCufTo0s"" | |
| 07:34:28 | alex_xu | Qiming: why my agent is python-novaclient? | |
| 07:34:30 | Qiming | em, you token is not hashed | |
| 07:34:44 | alex_xu | Qiming: i have a modified version of keystoneauth | |
| 07:34:58 | alex_xu | i hack the code directly... | |
| 07:35:24 | Qiming | your osc version? | |
| 07:35:42 | alex_xu | Qiming: 3.8.1 | |
| 07:36:29 | openstackgerrit | Alin Balutoiu proposed openstack/os-vif master: vif_plug_ovs: Skip setting MTU on Windows when plugging devices https://review.openstack.org/441855 | |
| 07:37:02 | Qiming | my command dump: curl -g -i -X GET http://192.168.42.11:8774/v2.1/servers/59a17d5a-c27a-4301-b6f2-19dd128f24df -H "User-Agent: python-novaclient" -H "Accept: application/json" -H "X-OpenStack-Nova-API-Version: 2.3" -H "X-Auth-Token: {SHA1}104a24f12ee85b77464fc36037dccbf5cf2beb45" | |
| 07:38:23 | alex_xu | Qiming: that looks same | |
| 07:38:36 | Qiming | so ... the responses differ if token is provided in different ways? | |
| 07:39:03 | alex_xu | Qiming: i don't think so, I just hack the code to not hide the real token in the log | |
| 07:40:07 | alex_xu | Qiming: the client pass the real token not the hash one. the hashed token is only for the log | |
| 07:40:19 | Qiming | alright | |
| 07:40:42 | Qiming | alex_xu, keystoneauth version? | |
| 07:40:43 | alex_xu | Qiming: can you paste your raw response | |
| 07:41:18 | alex_xu | Qiming: 2.18.0 | |
| 07:41:47 | Qiming | alex_xu, http://paste.openstack.org/show/601879/ | |
| 07:43:05 | Qiming | alex_xu, keystoneauth version match | |
| 07:43:35 | alex_xu | Qiming: do you have modified api-paste.ini for nova? | |
| 07:43:42 | Qiming | no | |
| 07:43:52 | Qiming | but my nova version is ... very old | |
| 07:44:28 | alex_xu | Qiming: paste your api-paste.ini to me? | |
| 07:45:46 | Qiming | alex_xu, http://paste.openstack.org/show/601881/ | |
| 07:46:20 | openstackgerrit | zhaolihui proposed openstack/nova master: Fix keypair quota error https://review.openstack.org/403529 | |
| 07:46:26 | openstackgerrit | jichenjc proposed openstack/nova master: remove a TODO as all set for tags https://review.openstack.org/442984 | |
| 07:47:53 | alex_xu | Qiming: are you using admin user in OSC? | |
| 07:48:03 | openstackgerrit | Hidekazu Nakamura proposed openstack/nova master: Do not attempt volume swap when guest is stopped https://review.openstack.org/389798 | |
| 07:48:07 | Qiming | no, I have all kinds of admins | |
| 07:48:13 | Qiming | s/have/hate | |
| 07:49:01 | alex_xu | Qiming: looks like the fields added by 2.3 only can seeable for admin user | |
| 07:49:09 | Qiming | what?! | |
| 07:49:23 | Qiming | lemme check | |
| 07:50:41 | alex_xu | Qiming: just ensure that, it only expose to admin user | |
| 07:51:22 | Qiming | nonono | |
| 07:51:40 | Qiming | a user with admin role suffice | |
| 07:52:04 | alex_xu | Qiming: yes | |
| 07:52:48 | alex_xu | but the admin role in the nova is equal to super admin, there is only one global admin | |
| 07:52:50 | Qiming | 1 min | |
| 07:53:27 | Qiming | you are right ... | |
| 07:53:53 | alex_xu | Qiming: :) | |
| 07:54:06 | Qiming | alex_xu, I have a pure shell script doing API calls -- sometimes I don't trust any software | |
| 07:54:38 | alex_xu | Qiming: ah, that is what i'm thinking, how you construct a curl with different user than osc | |
| 07:54:42 | Qiming | alex_xu, that script always generate the expected results, because .... I set USERNAME=admin | |
| 07:54:55 | alex_xu | heh | |
| 07:56:08 | Qiming | alex_xu, so ... I created an instance with user_data set to something, later on, I have to ask the admin guy to tell me what that user_data is ... | |
| 07:56:34 | Qiming | alex_xu, sounds like I was creating something I can never read back | |
| 07:56:48 | alex_xu | Qiming: yea, i agree that is a little strange. | |
| 07:56:59 | alex_xu | Qiming: the 2.3 is added for compatible ec2 API | |
| 07:57:21 | Qiming | alex_xu, it is not a requirement related to EC2 compatibility | |