| *** catintheroof has joined #senlin | 00:23 | |
| *** zhurong has joined #senlin | 01:00 | |
| *** hanwei has quit IRC | 01:01 | |
| *** XueFengLiu has quit IRC | 01:07 | |
| *** catintheroof has quit IRC | 01:13 | |
| *** catintheroof has joined #senlin | 01:14 | |
| *** hanwei has joined #senlin | 01:14 | |
| *** catintheroof has quit IRC | 01:14 | |
| *** yanyanhu has joined #senlin | 01:25 | |
| *** Drago1 has quit IRC | 01:28 | |
| *** ruijie_ has joined #senlin | 01:57 | |
| *** chohoor has joined #senlin | 01:59 | |
| *** root_____ has joined #senlin | 02:05 | |
| *** root_____ is now known as XueFeng | 02:05 | |
| openstackgerrit | Merged openstack/senlin: Fix cluster-list failed of ResourceNotFound problem https://review.openstack.org/421008 | 02:27 |
|---|---|---|
| *** shu-mutou has joined #senlin | 02:37 | |
| *** chohoor has quit IRC | 02:41 | |
| *** Jeffrey4l_ has joined #senlin | 03:00 | |
| *** Jeffrey4l has quit IRC | 03:04 | |
| *** chohoor has joined #senlin | 03:19 | |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin: revise tempest api for cluster 2 https://review.openstack.org/422369 | 03:30 |
| *** zhurong has quit IRC | 03:34 | |
| *** elynn has joined #senlin | 04:38 | |
| *** yanyanhu has quit IRC | 04:45 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Fix various problems in doc tree https://review.openstack.org/422386 | 04:58 |
| openstackgerrit | miaohb proposed openstack/senlin: Add db purge in senlin manage https://review.openstack.org/420666 | 05:04 |
| *** Drago1 has joined #senlin | 05:09 | |
| elynn | Hi Qiming | 05:15 |
| Qiming | hi | 05:18 |
| elynn | I met a problem when loading cluster data in profile.do_create | 05:18 |
| elynn | http://paste.openstack.org/show/595401/ | 05:18 |
| elynn | Seems the profile.context didn't have project | 05:19 |
| Qiming | yes | 05:19 |
| Qiming | I was worrying about that | 05:19 |
| Qiming | so you may turn to service context with user and project explicitly specified | 05:20 |
| elynn | Any suggestions on how to load cluster data from db? | 05:20 |
| openstackgerrit | Merged openstack/senlin: unify exceptions' message https://review.openstack.org/421109 | 05:20 |
| Qiming | use the code I pasted you | 05:20 |
| elynn | okay... I could have a try | 05:21 |
| elynn | I lost the code you pasted me before.. | 05:21 |
| openstackgerrit | Ethan Lynn proposed openstack/python-senlinclient: Parse nested dict for metadata parameter https://review.openstack.org/422392 | 05:22 |
| Qiming | ctx_dict = senlin_context.get_service_context(user=obj.user, project=obj.cluster) | 05:22 |
| Qiming | ctx = senlin_context.RequestContext.from_dict(ctx_dict) | 05:22 |
| Qiming | or something similar to this: http://git.openstack.org/cgit/openstack/senlin/tree/senlin/engine/health_manager.py#n72 | 05:23 |
| elynn | Thanks Qiming | 05:24 |
| Qiming | np | 05:24 |
| elynn | Another thing is about 'config' in metadata | 05:24 |
| elynn | See my patch https://review.openstack.org/422392 | 05:25 |
| elynn | client didn't parse config='{"foo": "bar"}' as a dict. | 05:25 |
| elynn | When directly call API, it's fine. | 05:26 |
| elynn | Saw your comments. Qiming | 05:29 |
| elynn | How are we gonna to handle it at server side? | 05:30 |
| Qiming | when we handles it at server side, we have the freedom to cope it case by case | 05:31 |
| elynn | Let me see if I can handle it at server side. | 05:32 |
| Qiming | if it is a value for metadata, we can do test and a conversion from the value when getting it from API request | 05:32 |
| elynn | yes, it would be great if senlin supports this. | 05:34 |
| elynn | If we support it at server side, what's the format of metadata that's looks like? | 05:38 |
| elynn | Qiming, for {config: {foo:bar}}, we could pass config=foo=bar, but for many parameters like {config: {f1:b1,f2:b2}} | 05:39 |
| elynn | What's your suggestion to pass them? | 05:39 |
| Qiming | it should be config='{"foo": "bar"}' | 05:47 |
| Qiming | I mean .. expose the JSON nature to end users | 05:47 |
| elynn | config='{"f1": "b1", "f2":"b2"}' ? | 05:47 |
| Qiming | yes | 05:47 |
| Qiming | doing another level of translation is gonna get us into a lot of troubles | 05:48 |
| Qiming | e.g. the 'f1' in your example could be another nested dict | 05:48 |
| elynn | If the parameters are too many, writing such a metadata would be a little boring.. | 05:48 |
| Qiming | agreed | 05:49 |
| Qiming | but incomplete solution is not a solution, right? | 05:49 |
| Qiming | {"f1": {"k1": ["v1", 2], "k2": {"v3": "v4"}}, "f2": "b2"} | 05:50 |
| Qiming | such a structure is very difficult to be flattened | 05:50 |
| elynn | okay, at least there's a way to parse json first, then we can improve and figure out a better way. | 05:50 |
| Qiming | exactly | 05:51 |
| elynn | Qiming, yes.. that's what struggles me... | 05:51 |
| wllabs | senlin create receiver after | 06:06 |
| wllabs | how to create monitor trigger alarm url | 06:06 |
| wllabs | call os::heat ? | 06:07 |
| wllabs | how to check the cluster when only transfer a alarm url and a cluster id | 06:08 |
| wllabs | Qiming | 06:19 |
| Qiming | don't understand what you are asking ... | 06:21 |
| Qiming | what do you mean by "check the cluster" ? | 06:22 |
| Qiming | what do you mean by "monitor" ? | 06:22 |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: Add security_group_find to neutron driver https://review.openstack.org/422407 | 06:26 |
| elynn | Hi Qiming , do you have time to review this patch for heat ? https://review.openstack.org/#/c/418339/ | 06:35 |
| Qiming | done | 06:35 |
| elynn | zan | 06:36 |
| lixinhui | elynn | 06:43 |
| lixinhui | In heat, can we use yaml to create senlin heat type profile? | 06:44 |
| lixinhui | never tried that | 06:44 |
| lixinhui | but assume it works | 06:44 |
| elynn | lixinhui, hi | 06:46 |
| elynn | senile heat type profile? | 06:46 |
| elynn | u mean use heat yaml and write down a senile profile type which is another heat yaml? | 06:46 |
| lixinhui | yes | 06:47 |
| elynn | I guess it could. | 06:47 |
| lixinhui | how can I point to the inner yaml | 06:47 |
| elynn | https://github.com/lynic/templates/blob/master/senlin/templates/asg-heat.yaml#L28-L42 | 06:48 |
| lixinhui | thanks | 06:48 |
| elynn | Something like this. | 06:48 |
| lixinhui | I think I wanna know | 06:48 |
| elynn | I used that yaml to test senile resources in heat 1 year ago. | 06:49 |
| elynn | thanks god I didn't delete it :) | 06:49 |
| lixinhui | if I can define something like this in the properties: template: nova_server_template.yaml | 06:50 |
| lixinhui | :) | 06:50 |
| elynn | What are you going to do? | 06:51 |
| lixinhui | use a heat template to create heat typed senlin profile | 06:51 |
| lixinhui | in inner yaml describe the trivial information | 06:52 |
| lixinhui | do not know if we can point some property to a ymal file | 06:52 |
| lixinhui | then automatically read all the definitions in that ymal | 06:53 |
| openstackgerrit | Merged openstack/senlin: Add port_create/port_delete to neutron driver https://review.openstack.org/421709 | 06:53 |
| lixinhui | ymal | 06:53 |
| lixinhui | yaml | 06:53 |
| elynn | lixinhui, not quite sure what you're going to do, call me if you need help. | 06:53 |
| lixinhui | ok | 06:54 |
| ruijie_ | Hi, Qiming, I am revising the tempest test. | 07:04 |
| ruijie_ | One question is that service layer will raise an exception if you detach a policy that not been attached to the cluster | 07:05 |
| ruijie_ | But it will accept requests that attach one policy to a cluster | 07:06 |
| *** zhurong has joined #senlin | 07:07 | |
| wllabs | When we create the receiver, we need to create a monitor to trigger it | 07:07 |
| wllabs | I would like to ask is how the monitor is detected | 07:08 |
| wllabs | because I like only the cluster ID and alarm url to the monitor | 07:08 |
| ruijie_ | wllabs, that is out the scope of Senlin | 07:09 |
| Qiming | wllabs, have you checked this doc? http://docs.openstack.org/developer/senlin/scenarios/autoscaling_ceilometer.html | 07:09 |
| wllabs | ok ,thanks a lot for you | 07:11 |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: Add security_group_find to neutron driver https://review.openstack.org/422407 | 07:12 |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: [WIP] Add a NFV VDU profile in contrib https://review.openstack.org/422422 | 07:12 |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin: revise tempest api for cluster 3 https://review.openstack.org/422429 | 07:28 |
| *** elynn has quit IRC | 07:34 | |
| *** zhangkun_ has joined #senlin | 07:35 | |
| *** zhangkun_ has left #senlin | 07:36 | |
| *** elynn has joined #senlin | 07:36 | |
| *** Drago1 has quit IRC | 08:09 | |
| *** openstackgerrit has quit IRC | 08:33 | |
| Wu_ | How to use senlin to create a container cluster? Does it depend on Magnum? | 08:33 |
| *** zhurong has quit IRC | 08:35 | |
| Qiming | no senlin directly invokes docker-py | 08:38 |
| wllabs | I only passed the cluster id to aodh, aodh know how to detect which nodes | 08:42 |
| Qiming | did you read the doc I sent you? | 08:43 |
| wllabs | I immediately under a closer look | 08:43 |
| wllabs | aodh will do metric aggregation based on the specified metadata. | 08:53 |
| wllabs | To make aodh aware of the cluster_id metadata senlin injects into each and every VM server created, you may need to add the following line into your /etc/ceilometer/ceilometer.conf file: | 08:53 |
| wllabs | i understand，thank you | 08:54 |
| *** openstackgerrit has joined #senlin | 08:59 | |
| *** ChanServ sets mode: +v openstackgerrit | 08:59 | |
| openstackgerrit | Merged openstack/senlin: Add security_group_find to neutron driver https://review.openstack.org/422407 | 08:59 |
| *** shu-mutou is now known as shu-mutou-AWAY | 08:59 | |
| wllabs | but only for a sinle node, not cluster | 09:00 |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: [WIP] Add a NFV VDU profile in contrib https://review.openstack.org/422422 | 09:05 |
| openstackgerrit | miaohb proposed openstack/senlin: Add db purge in senlin manage https://review.openstack.org/420666 | 09:10 |
| openstackgerrit | Merged openstack/senlin: revise tempest api for cluster 2 https://review.openstack.org/422369 | 09:10 |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin: revise tempest api for cluster 3 https://review.openstack.org/422429 | 09:11 |
| openstackgerrit | Hongbin Li proposed openstack/python-senlinclient: Add deprecation of cluster-run cli https://review.openstack.org/422498 | 09:22 |
| openstackgerrit | Hongbin Li proposed openstack/python-senlinclient: Add deprecation of cluster-run cli https://review.openstack.org/422498 | 09:30 |
| ruijie_ | Hi, Qiming, shell we provide an API to do action based recovering? | 09:34 |
| ruijie | e. g I have a cluster with 2 Active node. Then do cluster-scale- in -c 2. But one of these node-delete actions failed. | 09:47 |
| *** hanwei has quit IRC | 10:02 | |
| wllabs | I would like to ask Senlin in the back version will have a lot of feature to join ? | 10:26 |
| *** zhurong has joined #senlin | 10:30 | |
| *** ruijie has left #senlin | 10:59 | |
| *** zhurong has quit IRC | 12:05 | |
| *** catintheroof has joined #senlin | 12:43 | |
| *** elynn has quit IRC | 15:41 | |
| *** Drago1 has joined #senlin | 16:06 | |
| *** catinthe_ has joined #senlin | 18:18 | |
| *** catintheroof has quit IRC | 18:18 | |
| *** openstackgerrit has quit IRC | 19:33 | |
| *** Drago1 has quit IRC | 21:19 | |
| *** Drago1 has joined #senlin | 21:34 | |
| *** Drago2 has joined #senlin | 21:34 | |
| *** Jeffrey4l_ has quit IRC | 21:35 | |
| *** Jeffrey4l_ has joined #senlin | 21:36 | |
| *** Drago1 has quit IRC | 21:38 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!