| *** shu-mutou has quit IRC | 00:05 | |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make node-update return a location header and body https://review.openstack.org/254032 | 00:07 |
|---|---|---|
| openstackgerrit | Merged openstack/senlin: Fix service setup script for new endpoints https://review.openstack.org/254074 | 00:12 |
| openstackgerrit | Merged openstack/senlin: Fix policy json file to delete project checking https://review.openstack.org/254075 | 00:23 |
| openstackgerrit | Merged openstack/senlin: Make node-update return a location header and body https://review.openstack.org/254032 | 00:36 |
| *** jdandrea has quit IRC | 00:39 | |
| *** jdandrea has joined #senlin | 00:40 | |
| *** Qiming has joined #senlin | 00:48 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Delete tenant_id from location header https://review.openstack.org/254458 | 01:12 |
| *** zhenguo has joined #senlin | 01:23 | |
| *** Yanyanhu has joined #senlin | 01:26 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Fix user documentation for schema changes https://review.openstack.org/253862 | 01:35 |
| Qiming | Yanyanhu, regarding the functional test broken | 01:36 |
| Yanyanhu | yes | 01:36 |
| Qiming | I spent some time looking into the logs | 01:37 |
| Qiming | and I also tried to do tox -e functional in my local environment | 01:37 |
| Qiming | during that experiment, I found something need fixes -- already proposed | 01:37 |
| Qiming | but the fixes are not targeting why the test was still not passing | 01:38 |
| Qiming | when I look back at the changes we have made, I'm suspecting maybe the recent changes to the scheduler is no good | 01:38 |
| Qiming | I'm doubting if that change has bring in some race conditions | 01:39 |
| Yanyanhu | you mean the change about supporting initiative action scheduling | 01:39 |
| Qiming | yep | 01:39 |
| Yanyanhu | hmm, that's possible | 01:39 |
| Qiming | other than that, we were not changing action scheduling and dependency thing | 01:39 |
| Yanyanhu | yes, we didn't change any other part | 01:40 |
| Yanyanhu | ok, will recheck that change to find possible issue | 01:40 |
| Yanyanhu | and will also spend some time on DB transaction problem | 01:41 |
| Qiming | right, maybe the new db function is seeing cached data? maybe the change it made was not written back? | 01:41 |
| Yanyanhu | yes | 01:41 |
| Qiming | anyway, solving db transaction problem using oslo_concurrency is bad | 01:42 |
| Yanyanhu | will work it today. hope can find some clue | 01:42 |
| Yanyanhu | yep | 01:42 |
| Qiming | they work at different layers | 01:42 |
| Yanyanhu | agree with this | 01:42 |
| Qiming | great | 01:42 |
| Qiming | by the way, you don't have to push code to gate for functional tests | 01:46 |
| Qiming | it can be checked in your local environment | 01:46 |
| Yanyanhu | yes, now I can reproduce this error locally | 01:47 |
| Qiming | just apply https://review.openstack.org/254273 and the tenant_id related changes | 01:47 |
| Qiming | change cloud_backend in conf and restart services | 01:47 |
| Yanyanhu | yes | 01:48 |
| xuhaiwei | Qiming, about cluster/node delete api, should engine return action_id to API? | 01:52 |
| xuhaiwei | https://review.openstack.org/#/c/253980/ | 01:52 |
| xuhaiwei | yanyan has reviewed it, also need your opinion about it | 01:53 |
| Qiming | xuhaiwei, I'd suggest we remove action_id from engine return | 01:54 |
| Qiming | it is not used for these operations | 01:54 |
| Qiming | we will need action_id somewhere else | 01:54 |
| Qiming | for example, the 'action' handler for clusters | 01:55 |
| Qiming | all those handlers should return a location pointing to the action | 01:55 |
| openstackgerrit | Merged openstack/senlin: Delete tenant_id from location header https://review.openstack.org/254458 | 01:55 |
| xuhaiwei | ok, I will remove it | 01:56 |
| *** elynn has joined #senlin | 02:04 | |
| Yanyanhu | hi, Qiming, need I update openstackclient before using the latest setup-service tools to regenerate senlin service and endpoint? | 02:09 |
| Yanyanhu | since I got this error when running the tool | 02:09 |
| elynn | morning | 02:09 |
| Yanyanhu | openstack endpoint create: error: argument --publicurl is required | 02:09 |
| Yanyanhu | morning, elynn | 02:09 |
| Qiming | Yanyanhu, I think so | 02:09 |
| Yanyanhu | ok | 02:09 |
| Qiming | I think I'm using 1.7.1 | 02:10 |
| Yanyanhu | ok, its version is 1.5 in my local env | 02:10 |
| Yanyanhu | it's not a pip package? | 02:10 |
| Yanyanhu | weird, it is | 02:11 |
| Yanyanhu | it is not shown using pip list | 02:11 |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make cluster-delete retuan a location header https://review.openstack.org/253980 | 02:19 |
| Yanyanhu | hi, Qiming, after upgrading openstackclient version, that error still happened | 02:23 |
| Yanyanhu | the help msg from openstack cmd is: | 02:23 |
| Yanyanhu | usage: openstack endpoint create [-h] | 02:23 |
| Yanyanhu | [-f {html,json,json,shell,table,value,yaml,yaml}] | 02:23 |
| Yanyanhu | [-c COLUMN] [--max-width <integer>] | 02:23 |
| Yanyanhu | [--noindent] [--prefix PREFIX] --publicurl | 02:23 |
| Yanyanhu | <url> [--adminurl <url>] | 02:23 |
| Yanyanhu | [--internalurl <url>] [--region <region-id>] | 02:23 |
| Yanyanhu | <service> | 02:23 |
| xuhaiwei | yes, got the same error | 02:27 |
| elynn | Maybe the commads are not correct? | 02:27 |
| Qiming | looks like the help message is not correct? | 02:28 |
| Qiming | you can try openstack endpoint create for real | 02:28 |
| xuhaiwei | it seems '--publicurl --adminurl -- internalurl' options are required | 02:28 |
| Yanyanhu | hmm, the running of cmd also gave this error | 02:29 |
| Qiming | that is weird | 02:29 |
| Qiming | after I removed tenant_id | 02:29 |
| Qiming | I was recreating the endpoint | 02:29 |
| Qiming | but the --publicurl ... all doesn't work | 02:29 |
| Qiming | I see, you have to use keystone v3 | 02:30 |
| Qiming | kv3 endpoint create | 02:30 |
| Qiming | usage: openstack endpoint create [-h] | 02:30 |
| Qiming | [-f {html,json,json,shell,table,value,yaml,yaml}] | 02:30 |
| Qiming | [-c COLUMN] [--max-width <integer>] | 02:30 |
| Qiming | [--noindent] [--prefix PREFIX] | 02:30 |
| Qiming | [--region <region-id>] [--enable | --disable] | 02:30 |
| Qiming | <service> <interface> <url> | 02:30 |
| *** Qiming has quit IRC | 02:31 | |
| Yanyanhu | oh, so by default, openstackclient will talk with keystone v2 | 02:31 |
| *** Qiming has joined #senlin | 02:32 | |
| Qiming | kv3 here is my local alias to 'openstack --os-username=$OS_USERNAME --os-password=$OS_PASSWORD --os-url=$KEYSTONE_V3 --os-identity-api-version=3' | 02:32 |
| Qiming | then it means the script needs a fix again ... | 02:33 |
| Yanyanhu | yes, seems the default env setting of devstack is still for keystone v2 | 02:33 |
| Qiming | yep | 02:37 |
| Qiming | so we have two choices here | 02:39 |
| Qiming | use v3 command in setup-service | 02:39 |
| Qiming | or we stick to v2 | 02:39 |
| xuhaiwei | in setup_service, are we trying to create 3 senlin endpoint? | 02:39 |
| Qiming | yes | 02:39 |
| xuhaiwei | why need 3 endpoint? | 02:40 |
| xuhaiwei | I think the command it wrong | 02:40 |
| Qiming | because you need at least 3 | 02:41 |
| xuhaiwei | but changing the command to 'openstack endpoint create --region RegionOne senlin --adminurl xxx --publicurl xxx --internalurl xxx ', it works for me now | 02:44 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Revert to use keystone v2 based OSC https://review.openstack.org/254510 | 02:45 |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make cluster-delete return a location header https://review.openstack.org/253980 | 02:46 |
| Qiming | xuhaiwei, because you are using keystone v2 authentication | 02:46 |
| Qiming | keystone v3 command line is different | 02:46 |
| xuhaiwei | ohh, I see | 02:46 |
| Qiming | <Qiming> use v3 command in setup-service | 02:46 |
| Qiming | <Qiming> or we stick to v2 | 02:46 |
| Qiming | I'm reverting back to v2 now | 02:46 |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make cluster-delete return a location header https://review.openstack.org/253980 | 02:55 |
| openstackgerrit | lvdongbing proposed openstack/python-senlinclient: Make desired_capacity default to min_size if min_size specified https://review.openstack.org/254512 | 02:56 |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make cluster-delete return a location header https://review.openstack.org/253980 | 03:06 |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make node-delete return a location header https://review.openstack.org/253984 | 03:18 |
| *** elynn has quit IRC | 03:48 | |
| openstackgerrit | Merged openstack/senlin: Revert to use keystone v2 based OSC https://review.openstack.org/254510 | 03:49 |
| openstackgerrit | Merged openstack/senlin: Updated from global requirements https://review.openstack.org/254481 | 03:49 |
| *** elynn has joined #senlin | 03:56 | |
| *** elynn has quit IRC | 04:00 | |
| openstackgerrit | Merged openstack/python-senlinclient: Make desired_capacity default to min_size if min_size specified https://review.openstack.org/254512 | 04:19 |
| *** zhangguoqing has joined #senlin | 04:22 | |
| openstackgerrit | Merged openstack/senlin: Make cluster-delete return a location header https://review.openstack.org/253980 | 04:40 |
| *** elynn has joined #senlin | 04:40 | |
| openstackgerrit | Merged openstack/senlin: Make node-delete return a location header https://review.openstack.org/253984 | 04:41 |
| *** pratikmallya has joined #senlin | 04:43 | |
| *** elynn has quit IRC | 04:49 | |
| *** elynn has joined #senlin | 05:02 | |
| Yanyanhu | hi, Qiming, I noticed the version number has been removed from endpoint url, is that intended? | 05:11 |
| *** elynn_ has joined #senlin | 05:11 | |
| *** elynn_ has quit IRC | 05:11 | |
| *** elynn_ has joined #senlin | 05:11 | |
| Qiming | yes | 05:11 |
| Qiming | an endpoint is not supposed to be bound to a specific version number | 05:12 |
| Yanyanhu | so I guess we don't need version_negotiation middleware anymore? | 05:12 |
| Qiming | in future, we will need to add version negotiation | 05:12 |
| Qiming | no, that middleware is not doing useful work today, because we only have one api version | 05:13 |
| Qiming | however, it doesn't hurt | 05:13 |
| Yanyanhu | ok | 05:13 |
| Yanyanhu | it now gives an error like following when handling webhook trigger request | 05:13 |
| *** elynn has quit IRC | 05:13 | |
| Yanyanhu | senlin.api.middleware.version_negotiation [-] Returning HTTP 404 due to unknown Accept header: application/json | 05:13 |
| Qiming | you haven't applied my patch | 05:13 |
| Yanyanhu | also the webhook middleware need to be reworked for endpoint url change | 05:14 |
| Qiming | webhook has to change, right | 05:14 |
| Yanyanhu | I just pull the code about 30min ago | 05:14 |
| Yanyanhu | yes, will try to fix it | 05:14 |
| Qiming | this one is not merged | 05:15 |
| Qiming | https://review.openstack.org/#/c/254273/ | 05:15 |
| Qiming | senlin/tests/functional/common/client.py line 147 | 05:15 |
| Qiming | you need to explicitly specify the api version you want to talk to | 05:16 |
| Yanyanhu | I made the test manually without using functional test | 05:16 |
| Yanyanhu | ok. so if I want to trigger a webhook using curl, which part should I add to specify the version? | 05:17 |
| Qiming | if it is coming in from a v1 api call, we should add v1 into the url | 05:18 |
| Yanyanhu | oh, you mean add the version number behind the endpoint url? | 05:18 |
| Yanyanhu | like endpointurl/v1/ | 05:18 |
| Qiming | certainly | 05:18 |
| Yanyanhu | ok. so for webhook, I think we should add this part to webhook_url when creating it? | 05:19 |
| Qiming | just heads up, http://lists.openstack.org/pipermail/openstack-dev/2015-December/081612.html | 05:19 |
| Qiming | this is solving your previous concern that bugs are not automatically closed | 05:19 |
| Qiming | I wasn't closing them because I know this is coming | 05:19 |
| Yanyanhu | I see | 05:20 |
| Yanyanhu | Qiming, actually, I ran the functional test without change senlin/tests/functional/common/client.py line 147 by adding version number. So that is also because we have only one API version now, right? | 05:21 |
| Qiming | yes, we have only one | 05:23 |
| Yanyanhu | ok, understand | 05:24 |
| Yanyanhu | will fix the webhook issue | 05:24 |
| Qiming | http://git.openstack.org/cgit/openstack/senlin/tree/senlin/api/openstack/versions.py | 05:25 |
| Qiming | the version controller today only support v1 | 05:26 |
| Yanyanhu | ok, just felt a little weird since all other API requests except webhook trigger request can pass version_negotiation middleware | 05:27 |
| Yanyanhu | go downstairs, back soon | 05:28 |
| Qiming | http://docs-draft.openstack.org/45/252245/7/check/gate-api-site-tox-doc-publish-checkbuild/16a9e4d//publish-docs/api-ref/api-ref-clustering-v1.html | 05:31 |
| Qiming | this is how the API docs will look like | 05:31 |
| Qiming | guys please browse through it and see if there are bugs | 05:31 |
| Yanyanhu | wonderful | 05:32 |
| Qiming | still wondering why it is not generating pdf for me | 05:33 |
| Qiming | http://docs-draft.openstack.org/45/252245/8/check/gate-api-site-tox-doc-publish-checkbuild/8ec6078//publish-docs/ | 05:33 |
| openstackgerrit | Merged openstack/senlin: Misc fix to functional tests https://review.openstack.org/254273 | 05:34 |
| openstackgerrit | Merged openstack/senlin: Fix user documentation for schema changes https://review.openstack.org/253862 | 05:34 |
| Qiming | okay, got it | 05:34 |
| openstackgerrit | Ethan Lynn proposed openstack/python-senlinclient: Update related attributes after policy_create/profile_create https://review.openstack.org/254254 | 05:54 |
| openstackgerrit | Yanyan Hu proposed openstack/senlin: Rework webhook to adapt to latest API change https://review.openstack.org/254554 | 05:59 |
| openstackgerrit | Ethan Lynn proposed openstack/python-senlinclient: Update related attributes after policy_create/profile_create https://review.openstack.org/254254 | 06:00 |
| *** pratikmallya has quit IRC | 06:18 | |
| openstackgerrit | Ethan Lynn proposed openstack/python-senlinclient: Update related attributes after policy_create/profile_create https://review.openstack.org/254254 | 06:26 |
| Qiming | fighting with the api doc for almost a whole day now | 07:16 |
| Yanyanhu | almost done? | 07:17 |
| Qiming | debugging is very difficult | 07:17 |
| Qiming | the output only tells you there are errors | 07:17 |
| Yanyanhu | yes, since the xml is very very difficult to read... | 07:17 |
| Qiming | there is no useful information about line number or the error itself | 07:18 |
| Qiming | the only thing you get is a full screen of Java exceptions | 07:18 |
| Yanyanhu | sigh... | 07:18 |
| Qiming | this is the most useful line: Failed to execute goal com.rackspace.cloud.api:clouddocs-maven-plugin:2.1.4:generate-pdf (os-api-ref-clustering-v1) on project openstack-api-ref-guides: Execution os-api-ref-clustering-v1 of goal com.rackspace.cloud.api:clouddocs-maven-plugin:2.1.4:generate-pdf failed: net.sf.saxon.s9api.SaxonApiException: Processing terminated by xsl:message at line -1 in null -> [Help 1] | 07:19 |
| Yanyanhu | -_-! | 07:19 |
| Qiming | I don't have line -1 | 07:20 |
| Yanyanhu | what the hell is this... | 07:20 |
| Qiming | then you go back to the doc, try recall which line you changed ... | 07:21 |
| Yanyanhu | ... | 07:22 |
| Qiming | lessons learnt | 07:22 |
| Qiming | maybe we should provide a profile-validate and a policy-validate api | 07:22 |
| *** zhangguoqing has quit IRC | 07:22 | |
| Qiming | if something goes wrong, we should tell users the exact line number | 07:23 |
| Yanyanhu | yes | 07:23 |
| Yanyanhu | that will be very helpful for enduser | 07:23 |
| Qiming | heat had such an intent a long time ago, but no one is working on that | 07:23 |
| Qiming | people tend to fix whatever bugs they want to fix, not features that would be useful for users | 07:23 |
| Yanyanhu | this is an useful feature obviously | 07:24 |
| Qiming | the contributors don't care, they care only about their core status, their number of patches ... | 07:25 |
| Yanyanhu | yea, this is a problem that could happen when a project become very big | 07:26 |
| openstackgerrit | lvdongbing proposed openstack/python-senlinclient: Cluster's timeout is in seconds not minutes https://review.openstack.org/254601 | 08:04 |
| *** shu-mutou has joined #senlin | 08:11 | |
| *** elynn_ has quit IRC | 08:33 | |
| *** elynn_ has joined #senlin | 08:40 | |
| *** elynn__ has joined #senlin | 08:44 | |
| *** elynn_ has quit IRC | 08:44 | |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make cluster-resize return a location header https://review.openstack.org/254616 | 08:47 |
| *** elynn__ has quit IRC | 08:56 | |
| *** elynn__ has joined #senlin | 08:56 | |
| openstackgerrit | Yanyan Hu proposed openstack/senlin: [WIP]Fix DB sync issue happened during action dependency updating https://review.openstack.org/252231 | 09:26 |
| openstackgerrit | Yanyan Hu proposed openstack/senlin: [WIP]Fix DB sync issue happened during action dependency updating https://review.openstack.org/252231 | 09:28 |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make cluster actions API return a location header https://review.openstack.org/254636 | 09:29 |
| Qiming | Yanyanhu, I'm looking at the code about concurrency fix | 09:29 |
| Qiming | suddenly I noticed that there are locations we are not saving the changed object | 09:30 |
| Yanyanhu | Qiming, thanks. I think it still has some problems. I will test it to see whether deadlock could happen | 09:30 |
| Qiming | for example, in function action_marke_succeeded | 09:30 |
| Qiming | we are changing some properties of an action | 09:30 |
| Qiming | but we did not store it | 09:31 |
| Yanyanhu | oh, that's true | 09:31 |
| Yanyanhu | just noticed it | 09:31 |
| Qiming | the should be an equivalent call to action.save(...) as in function action_create() | 09:31 |
| Qiming | maybe this is the route cause | 09:31 |
| Yanyanhu | right | 09:32 |
| Qiming | it looks like a stupid error | 09:32 |
| Yanyanhu | will revise related it code to check the result | 09:32 |
| Qiming | yes, maybe it could be just a simple fix by adding one line | 09:32 |
| Qiming | sigh ... | 09:33 |
| Yanyanhu | ... | 09:33 |
| *** pratikmallya has joined #senlin | 09:34 | |
| Yanyanhu | oh, I remeber I checked the status of those node actions which are dependent by cluster action. their status are set to SUCCEEDED correctly | 09:34 |
| Yanyanhu | does that mean those updated action attributes are actually stored back to DB? | 09:34 |
| Qiming | we were doing some .refresh() calls previously | 09:35 |
| Qiming | you will never know | 09:35 |
| Yanyanhu | oh, yes | 09:35 |
| Qiming | we are raising a challenging question to sqlalchemy and I don't think it will have an answer | 09:35 |
| Yanyanhu | Qiming, yes, I think this is a tradeoff between consistency and performance... | 09:39 |
| Yanyanhu | hi, Qiming, just added action.save(), problem is still there... Will add it to patch 252231 to see what will happen | 09:46 |
| *** shu-mutou has quit IRC | 09:46 | |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make node join/leave return a location header https://review.openstack.org/254641 | 09:48 |
| *** pratikmallya has quit IRC | 09:54 | |
| lixinhui | /join ##openstack-meeting | 09:59 |
| elynn__ | ? | 10:00 |
| Yanyanhu | lixinhui :P | 10:01 |
| lixinhui | :P | 10:02 |
| Yanyanhu | meeting night tonight | 10:02 |
| Yanyanhu | prepare to go back home | 10:02 |
| lixinhui | yeah, see you then | 10:02 |
| Yanyanhu | see U | 10:02 |
| Yanyanhu | Qiming, plz leave comment on 252231 if any. Thanks | 10:03 |
| *** zhenguo has quit IRC | 10:03 | |
| Qiming | okay, I'm still feeling that we are not saving changes to actions | 10:04 |
| Qiming | some refresh() calls were removed | 10:05 |
| Qiming | I'm not sure if the execution_option params will help | 10:05 |
| *** elynn__ has quit IRC | 10:07 | |
| *** Yanyanhu has quit IRC | 10:07 | |
| openstackgerrit | Merged openstack/python-senlinclient: Specify keystoneauth1 exception more https://review.openstack.org/253303 | 10:36 |
| openstackgerrit | Merged openstack/senlin: Make cluster-resize return a location header https://review.openstack.org/254616 | 10:49 |
| *** openstackgerrit has quit IRC | 11:02 | |
| *** openstackgerrit has joined #senlin | 11:03 | |
| *** Qiming has quit IRC | 11:39 | |
| *** pratikmallya has joined #senlin | 11:57 | |
| *** Liuqing has joined #senlin | 12:22 | |
| *** Yanyanhu has joined #senlin | 12:25 | |
| *** elynn__ has joined #senlin | 12:29 | |
| *** Qiming has joined #senlin | 12:30 | |
| openstackgerrit | Yanyan Hu proposed openstack/senlin: Rework webhook to adapt to latest API change https://review.openstack.org/254554 | 12:59 |
| Qiming | Yanyanhu, I think we don't need a '+' in generating location | 13:00 |
| openstackgerrit | Yanyan Hu proposed openstack/senlin: Rework webhook to adapt to latest API change https://review.openstack.org/254554 | 13:00 |
| openstackgerrit | Yanyan Hu proposed openstack/senlin: [WIP]Fix DB sync issue happened during action dependency updating https://review.openstack.org/252231 | 13:00 |
| Qiming | what do you think? | 13:00 |
| Yanyanhu | yes, there is a typo in current code | 13:00 |
| Yanyanhu | have change it | 13:01 |
| Yanyanhu | don't know how it passed pep8 check... | 13:01 |
| Qiming | I mean line 200 https://review.openstack.org/#/c/254554/2/senlin/engine/webhook.py | 13:01 |
| Yanyanhu | Qiming, I see | 13:02 |
| Yanyanhu | will change it | 13:02 |
| *** cschulz has joined #senlin | 13:13 | |
| *** lkarm has joined #senlin | 13:14 | |
| openstackgerrit | Yanyan Hu proposed openstack/senlin: Rework webhook to adapt to latest API change https://review.openstack.org/254554 | 13:15 |
| openstackgerrit | Yanyan Hu proposed openstack/senlin: [WIP]Fix DB sync issue happened during action dependency updating https://review.openstack.org/252231 | 13:15 |
| *** Mayur_ has joined #senlin | 13:17 | |
| *** Qiming has quit IRC | 13:18 | |
| *** Qiming has joined #senlin | 13:18 | |
| *** Shruti_ has joined #senlin | 13:19 | |
| *** Mayur_ has quit IRC | 13:21 | |
| *** Shruti_ has quit IRC | 13:22 | |
| *** Qiming has quit IRC | 13:25 | |
| *** Liuqing has quit IRC | 13:36 | |
| *** pratikmallya has quit IRC | 13:48 | |
| *** Liuqing has joined #senlin | 13:56 | |
| Yanyanhu | hi, guys, will leave a little earlier. ttyl | 13:57 |
| *** Qiming has joined #senlin | 13:59 | |
| Qiming | so .. it is some code maintained in senlinclient ? | 14:00 |
| *** haiwei_ has joined #senlin | 14:00 | |
| elynn__ | Yes, let me find them... | 14:01 |
| haiwei_ | ok I will go to bed | 14:01 |
| haiwei_ | see you tomorrow | 14:02 |
| *** Yanyanhu has quit IRC | 14:02 | |
| elynn__ | Qiming: https://github.com/openstack/python-senlinclient/blob/master/senlinclient/common/sdk.py#L94-L96 | 14:02 |
| Qiming | okay, I see | 14:02 |
| elynn__ | See U haiwei | 14:02 |
| Qiming | that is from a previous patch we used to support webhook | 14:02 |
| Qiming | when creating webhooks, the first response will contain some secrets that you will never see in future, :) | 14:03 |
| elynn__ | ok, I see | 14:03 |
| Qiming | need to revisit that if it still makes sense | 14:03 |
| Qiming | however, using this function is really discouraged | 14:03 |
| elynn__ | maybe patchset 1 is the best way. | 14:03 |
| cschulz | Qiming: Got Senlin up on my Liberty devstack. Testing now | 14:04 |
| elynn__ | Just retrieve it again after create | 14:04 |
| Qiming | cschulz, great! | 14:04 |
| Qiming | elynn__, yep, tend to agree | 14:04 |
| elynn__ | ok, will change it back tomorrow~ | 14:04 |
| Qiming | thx | 14:05 |
| Qiming | well | 14:05 |
| elynn__ | Need to take a shower now~ ttyl | 14:05 |
| Qiming | if it makes sense to have SDK return the *very first* response in the resource rather than doing another retrieval | 14:05 |
| Qiming | we can do that | 14:05 |
| Qiming | bye, elynn__ | 14:05 |
| *** lixinhui has quit IRC | 14:06 | |
| *** haiwei_ has quit IRC | 14:06 | |
| cschulz | Getting errors while trying to resize random_string_stack. | 14:06 |
| cschulz | [stack@autoscale2 profiles]$ senlin cluster-resize -a 1 cos-cluster | 14:07 |
| cschulz | Unknown exception: Not Found (HTTP 404) (Request-ID: req-172f7532-ce9e-4978-8025-9bc548564a12) | 14:07 |
| cschulz | 'NoneType' object has no attribute '__getitem__' | 14:07 |
| cschulz | [stack@autoscale2 profiles]$ senlin cluster-resize -c 2 cos-cluster | 14:07 |
| cschulz | Unknown exception: Not Found (HTTP 404) (Request-ID: req-1b6a7b0d-ccd9-4695-93a6-f03566644854) | 14:07 |
| cschulz | 'NoneType' object has no attribute '__getitem__' | 14:07 |
| cschulz | [stack@autoscale2 profiles]$ | 14:07 |
| Qiming | cschulz, I think you just jumped in the middle | 14:08 |
| Qiming | keystone endpoint list | grep senlin | 14:08 |
| Qiming | what does it say? | 14:08 |
| Qiming | I mean you have already have a cluster created? | 14:09 |
| *** pratikmallya has joined #senlin | 14:10 | |
| cschulz | Yes, have a cluster | 14:10 |
| Qiming | can you do senlin cluster-show ? | 14:10 |
| *** Liuqing has quit IRC | 14:11 | |
| Qiming | and when you are doing a 'resize', can you add the '-d' option to see where the exception was thrown? | 14:11 |
| cschulz | [stack@autoscale2 profiles]$ senlin cluster-show cos-cluster | 14:11 |
| cschulz | +------------------+--------------------------------------+ | 14:11 |
| cschulz | | Property | Value | | 14:11 |
| cschulz | +------------------+--------------------------------------+ | 14:11 |
| cschulz | | created_time | 2015-12-08T13:49:02 | | 14:11 |
| cschulz | | data | {} | | 14:11 |
| cschulz | | deleted_time | None | | 14:11 |
| cschulz | | desired_capacity | 1 | | 14:11 |
| cschulz | | domain | None | | 14:11 |
| cschulz | | id | 733eeaca-c526-4208-b89e-6b1d06015cdb | | 14:11 |
| cschulz | | max_size | 3 | | 14:11 |
| cschulz | | metadata | {} | | 14:11 |
| cschulz | | min_size | 0 | | 14:11 |
| cschulz | | name | cos-cluster | | 14:11 |
| cschulz | | nodes | b54da470-99b8-4d9c-8a43-093a79942ab4 | | 14:11 |
| cschulz | | parent | None | | 14:12 |
| cschulz | | profile_id | 9a25dd16-a591-4e56-8151-1bd7deb986f4 | | 14:12 |
| cschulz | | profile_name | cos-stack | | 14:12 |
| cschulz | | project | 91e1117d33a24536be8e8f4cfb167db9 | | 14:12 |
| cschulz | | status | ACTIVE | | 14:12 |
| cschulz | | status_reason | Cluster creation succeeded. | | 14:12 |
| cschulz | | timeout | 3600 | | 14:12 |
| cschulz | | updated_time | None | | 14:12 |
| cschulz | | user | b0d7b45503cb4d86b742107d91e4347d | | 14:12 |
| cschulz | +------------------+--------------------------------------+ | 14:12 |
| Qiming | em, you have a cluster with a heat stack node, cool! | 14:12 |
| cschulz | INFO (shell) 1 | 14:13 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('v2token = keystoneauth1.loading._plugins.identity.v2:Token') | 14:13 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('admin_token = keystoneauth1.loading._plugins.admin_token:AdminToken') | 14:13 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('v3oidcauthcode = keystoneauth1.loading._plugins.identity.v3:OpenIDConnectAuthorizationCode') | 14:13 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('v2password = keystoneauth1.loading._plugins.identity.v2:Password') | 14:13 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('v3password = keystoneauth1.loading._plugins.identity.v3:Password') | 14:13 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('v3oidcpassword = keystoneauth1.loading._plugins.identity.v3:OpenIDConnectPassword') | 14:13 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('token = keystoneauth1.loading._plugins.identity.generic:Token') | 14:14 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('v3token = keystoneauth1.loading._plugins.identity.v3:Token') | 14:14 |
| cschulz | DEBUG (extension) found extension EntryPoint.parse('password = keystoneauth1.loading._plugins.identity.generic:Password') | 14:14 |
| cschulz | DEBUG (session) REQ: curl -g -i -X GET http://10.114.223.72:5000/v2.0 -H "Accept: application/json" -H "User-Agent: python-senlinclient" | 14:14 |
| cschulz | INFO (connectionpool) Starting new HTTP connection (1): 10.114.223.72 | 14:14 |
| cschulz | DEBUG (connectionpool) "GET /v2.0 HTTP/1.1" 200 339 | 14:14 |
| cschulz | DEBUG (session) RESP: [200] Content-Length: 339 Vary: X-Auth-Token Keep-Alive: timeout=5, max=100 Server: Apache/2.4.6 (CentOS) OpenSSL/1.0.1e-fips mod_wsgi/3.4 Python/2.7.5 Connection: Keep-Alive Date: Tue, 08 Dec 2015 14:12:46 GMT Content-Type: application/json x-openstack-request-id: req-b8883581-e007-405a-af82-b11800881a13 | 14:14 |
| cschulz | RESP BODY: {"version": {"status": "stable", "updated": "2014-04-17T00:00:00Z", "media-types": [{"base": "application/json", "type": "application/vnd.openstack.identity-v2.0+json"}], "id": "v2.0", "links": [{"href": "http://10.114.223.72:5000/v2.0/", "rel": "self"}, {"href": "http://docs.openstack.org/", "type": "text/html", "rel": "descr | 14:14 |
| cschulz | Yes everything was working fine until I tried to do a scaling action. | 14:15 |
| Qiming | good news is that keystone v3 works on liberty, keystoneauth1 works also | 14:16 |
| Qiming | liberty is pretty new | 14:16 |
| Qiming | the trace you pasted was not complete | 14:16 |
| cschulz | Yes the Liberty seems to be working fine. I see the stack and resource created in Heat | 14:17 |
| Qiming | I'm not sure you have successfully got a token or not | 14:17 |
| Qiming | you can check the sl-api screen session and see if something happened there | 14:18 |
| cschulz | That's everything I got back on the console. The command was just above the INFO line | 14:18 |
| cschulz | Oops! maybe the copy didn't work completely | 14:19 |
| Qiming | can you help paste the whole trace into paste.openstack.org? | 14:19 |
| cschulz | DEBUG (session) Request returned failure status: 404 | 14:19 |
| cschulz | Unknown exception: Not Found (HTTP 404) (Request-ID: req-9dab3b95-f330-48ab-b21a-326b66be1b0a) | 14:19 |
| cschulz | Traceback (most recent call last): | 14:19 |
| cschulz | File "/usr/bin/senlin", line 11, in <module> | 14:19 |
| cschulz | sys.exit(main()) | 14:19 |
| cschulz | File "/usr/lib/python2.7/site-packages/senlinclient/shell.py", line 308, in main | 14:19 |
| cschulz | SenlinShell().main(args) | 14:19 |
| cschulz | File "/usr/lib/python2.7/site-packages/senlinclient/shell.py", line 294, in main | 14:19 |
| cschulz | args.func(sc, args) | 14:19 |
| cschulz | File "/usr/lib/python2.7/site-packages/senlinclient/v1/shell.py", line 825, in do_cluster_resize | 14:19 |
| cschulz | print('Request accepted by action: %s' % resp['action']) | 14:19 |
| cschulz | TypeError: 'NoneType' object has no attribute '__getitem__' | 14:19 |
| cschulz | That's what was after the RESP BODY | 14:19 |
| Qiming | okay, ... | 14:20 |
| Qiming | that is a new bug, cschulz | 14:20 |
| Qiming | :) | 14:20 |
| Qiming | let me try explain this to you | 14:20 |
| Qiming | we are revising the senlin api design, you can get a look and feel of it here: http://docs-draft.openstack.org/45/252245/10/check/gate-api-site-tox-doc-publish-checkbuild/d1f26a1//publish-docs/api-ref-guides/bk-api-ref-clustering-v1.pdf | 14:21 |
| Qiming | one of the big changes we are making recently is to have all request that are performed asynchronously return a 202 status code | 14:22 |
| Qiming | this is a requirement from the api-wg project (which is trying to uniform how apis behave across projects) | 14:22 |
| Qiming | the server side changes have just landed | 14:23 |
| Qiming | we have not yet finished the client side change | 14:23 |
| openstackgerrit | lvdongbing proposed openstack/python-senlinclient: Make policy-create show result policy correctly https://review.openstack.org/254777 | 14:23 |
| Qiming | in the case of 202 return code, SDK is now setting the response body to {} | 14:23 |
| Qiming | that is why we are no longer getting the 'action' key from the response body | 14:24 |
| Qiming | anyway, if the resize logic was correct, you may have your cluster already resized, :) | 14:24 |
| cschulz | Oh let me check | 14:25 |
| cschulz | Nope, resize didn't happen. | 14:27 |
| *** elynn__ has quit IRC | 14:27 | |
| cschulz | Should I open a bug report? | 14:29 |
| Qiming | that would help | 14:29 |
| cschulz | How do I do that? | 14:30 |
| Qiming | https://bugs.launchpad.net/senlin/ | 14:30 |
| Qiming | upper right corner, "Report a bug" | 14:30 |
| Qiming | in the bug report, please paste the trace you got | 14:31 |
| cschulz | OK | 14:31 |
| Qiming | bed time | 14:33 |
| Qiming | see you, guys | 14:34 |
| cschulz | Good night. Thanks. | 14:34 |
| *** Qiming has quit IRC | 14:34 | |
| *** pratikmallya has quit IRC | 15:05 | |
| openstackgerrit | lvdongbing proposed openstack/senlin: Format time to cut microsecond https://review.openstack.org/254814 | 15:09 |
| *** pratikmallya has joined #senlin | 15:44 | |
| *** pratikmallya has quit IRC | 16:19 | |
| *** pratikmallya has joined #senlin | 16:30 | |
| *** openstackgerrit has quit IRC | 16:47 | |
| *** openstackgerrit has joined #senlin | 16:48 | |
| *** pratikmallya has quit IRC | 17:36 | |
| *** lkarm has quit IRC | 17:42 | |
| *** pratikmallya has joined #senlin | 18:07 | |
| *** lkarm has joined #senlin | 18:41 | |
| *** openstackgerrit has quit IRC | 19:02 | |
| *** openstackgerrit has joined #senlin | 19:03 | |
| *** pratikmallya has quit IRC | 19:33 | |
| *** pratikmallya has joined #senlin | 19:36 | |
| *** pratikma_ has joined #senlin | 22:10 | |
| *** pratikmallya has quit IRC | 22:14 | |
| *** pratikma_ has quit IRC | 22:15 | |
| *** pratikmallya has joined #senlin | 22:43 | |
| *** yuanying has joined #senlin | 23:00 | |
| *** lkarm has quit IRC | 23:02 | |
| openstackgerrit | xu-haiwei proposed openstack/senlin: Make node join/leave return a location header https://review.openstack.org/254641 | 23:57 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!