| *** XueFeng has quit IRC | 00:05 | |
| *** yanyanhu has joined #senlin | 00:46 | |
| *** ruijie has joined #senlin | 01:03 | |
| *** yuanying has quit IRC | 01:11 | |
| *** yuanying has joined #senlin | 01:11 | |
| *** zhurong has joined #senlin | 01:15 | |
| *** zhurong has quit IRC | 01:20 | |
| *** zhurong has joined #senlin | 01:22 | |
| *** chohoor has joined #senlin | 01:42 | |
| *** elynn has joined #senlin | 01:51 | |
| *** elynn has quit IRC | 01:56 | |
| *** elynn has joined #senlin | 01:56 | |
| openstackgerrit | miaohb proposed openstack/senlin: Doc update for node replace https://review.openstack.org/409676 | 02:05 |
|---|---|---|
| *** yuanying has quit IRC | 02:11 | |
| fabian4 | Qiming, how can i use senlin create ceilometer alarm ? | 02:15 |
| Qiming | fabian4, senlin doesn't have such a proxy | 02:16 |
| fabian4 | Qiming, I find ceilometer code create alarm in senlin driver , senlin do not support ? | 02:18 |
| Qiming | ceilometer code create alarm? | 02:25 |
| fabian4 | Qiming, senlin/drivers/openstack/ceilometer_v2.py def alarm_create | 02:28 |
| Qiming | that is an internal interface for senlin to talk to ceilometer | 02:28 |
| Qiming | we are not proxying that to senlin users | 02:29 |
| Qiming | it makes no sense | 02:29 |
| Qiming | imaging this | 02:29 |
| Qiming | imagine this | 02:29 |
| Qiming | if we support "senlin alarm-create --type ceilometer --metric cpu.util --threshold 70 --operator great_than --period 1 --num_periods 3 ... " | 02:30 |
| Qiming | then senlin calls the ceilometer to create an alarm for you ... | 02:31 |
| Qiming | why don't you call ceilometer API directly, or aodh CLI directly? | 02:31 |
| *** guoshan has joined #senlin | 02:41 | |
| openstackgerrit | miaohb proposed openstack/senlin: Doc update for node replace https://review.openstack.org/409676 | 02:42 |
| fabian4 | Qiming, I known, I was begin study senlin . thanks give me a advise | 02:51 |
| Qiming | never mind | 02:52 |
| Qiming | we have been there | 02:52 |
| Qiming | we did considered whether senlin should provide an abstraction of all alarms we will use, e.g. ceilometer alarm, monasca alarm, surveil alarm ... whatever | 02:52 |
| Qiming | but we found we are complicating things unnecessarily | 02:53 |
| Qiming | users know what to do with alarm services, we are not suppose to just add a proxy layer to the backend service | 02:53 |
| openstackgerrit | Merged openstack/python-senlinclient: Client support for node replace https://review.openstack.org/409420 | 03:22 |
| openstackgerrit | Merged openstack/python-senlinclient: Add metadata clean help https://review.openstack.org/409431 | 03:24 |
| *** zhurong has quit IRC | 03:34 | |
| ruijie | hi, yanyanhu, around? | 03:35 |
| openstackgerrit | Merged openstack/python-senlinclient: Delete pbr generated files https://review.openstack.org/407486 | 03:47 |
| *** dinghh has joined #senlin | 04:00 | |
| openstackgerrit | miaohb proposed openstack/senlin: Doc update for node replace https://review.openstack.org/409676 | 04:09 |
| *** elynn has quit IRC | 04:29 | |
| *** yuanying has joined #senlin | 04:33 | |
| *** elynn has joined #senlin | 04:49 | |
| *** elynn has quit IRC | 04:54 | |
| *** elynn has joined #senlin | 04:54 | |
| *** dinghh has quit IRC | 05:07 | |
| *** dinghh has joined #senlin | 05:09 | |
| yanyanhu | hi, ruijie, just came back | 05:14 |
| ruijie | yanyan, I just check the rally-jobs problem | 05:16 |
| ruijie | all the tests failed due to the private network | 05:16 |
| ruijie | but they passed after I change the network to public | 05:16 |
| yanyanhu | ruijie, that means there is no such network named as "private" in rally test environment. | 05:17 |
| yanyanhu | maybe there were some changes in rally gate | 05:17 |
| yanyanhu | since it did work before... | 05:17 |
| yanyanhu | and in default devstack environment, a network named as "private" will be created by default | 05:18 |
| ruijie | yup, they did work just several days ago | 05:18 |
| ruijie | that is the problem... I can create nova server manually, but failed when run rally task | 05:19 |
| yanyanhu | what exception was raised? | 05:19 |
| ruijie | no network private found | 05:19 |
| yanyanhu | have you changed the network referred in rally job file? | 05:20 |
| ruijie | nope | 05:21 |
| yanyanhu | you may need to change the network name defined in rally job file as well | 05:22 |
| ruijie | the devstack environment have public and private network | 05:23 |
| ruijie | the test failed when I use private network in the profile | 05:23 |
| ruijie | but it works when I change the network to public in profile | 05:23 |
| yanyanhu | yes. devstack-gate will create both "private" and "public" networks by default | 05:24 |
| yanyanhu | just not sure whether rally gate makes change on it | 05:24 |
| yanyanhu | another possible reason is network "private" is invisible for the user/tenant performs the test operations, e.g. cluster create | 05:25 |
| yanyanhu | for private network could not be "shared" | 05:25 |
| ruijie | but we are using admin_client to trigger these actions | 05:26 |
| yanyanhu | ruijie, admin_client means the tenant(s) userd for test have "admin" role | 05:27 |
| yanyanhu | not it is "admin" user :) | 05:27 |
| yanyanhu | so not sure whether that is the problem | 05:27 |
| yanyanhu | so maybe a temporary workaround is using "public" rather than "private" network in rally test | 05:28 |
| yanyanhu | another possible way is creating a specific network and using it for rally test | 05:29 |
| yanyanhu | not sure whether rally has provided such context to support creating specific network | 05:29 |
| *** guoshan has quit IRC | 05:29 | |
| yanyanhu | if so, we can directly using it | 05:29 |
| yanyanhu | if not, there could be some effort to support it... | 05:30 |
| yanyanhu | so maybe we can apply the workaround first, and then make some investigation here. I will try to take a look at it | 05:30 |
| ruijie | thanks yanyanhu, will check it with rally team | 05:32 |
| yanyanhu | ruijie, thanks a lot :) | 05:32 |
| *** Drago1 has quit IRC | 05:36 | |
| *** guoshan has joined #senlin | 06:04 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Use util function in cluster API (1) https://review.openstack.org/410050 | 06:06 |
| *** yuanying has quit IRC | 06:12 | |
| *** yuanying has joined #senlin | 06:12 | |
| *** yuanying has quit IRC | 06:12 | |
| *** yuanying has joined #senlin | 06:13 | |
| *** yuanying has quit IRC | 06:17 | |
| *** chohoor has quit IRC | 06:38 | |
| *** chohoor has joined #senlin | 06:41 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Use util function in cluster API (2) https://review.openstack.org/410062 | 06:56 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Use util function in cluster API (2) https://review.openstack.org/410062 | 07:14 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Use util function in cluster API (1) https://review.openstack.org/410050 | 07:16 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Use util function in cluster API (2) https://review.openstack.org/410062 | 07:18 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Use util function in cluster API (2) https://review.openstack.org/410062 | 07:30 |
| *** yuanying has joined #senlin | 07:51 | |
| openstackgerrit | Aaron Ding proposed openstack/senlin: Set max_size to cfg.CONF.max_nodes_per_cluster if -1. https://review.openstack.org/410083 | 08:14 |
| *** XueFeng has joined #senlin | 08:24 | |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Use util function in cluster API (3) https://review.openstack.org/410086 | 08:24 |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: [WIP] Remove retry logic from lock_acquire https://review.openstack.org/410095 | 08:42 |
| openstackgerrit | Merged openstack/senlin: Doc update for node replace https://review.openstack.org/409676 | 08:54 |
| openstackgerrit | Shan Guo proposed openstack/senlin: Add parameter in CLUSTER_DEL_NODES action https://review.openstack.org/410109 | 09:05 |
| Qiming | guoshan, online? | 09:11 |
| guoshan | yes | 09:11 |
| Qiming | just noticed you and your colleage are both working on adding 'destroy_after_deletion' parameter to ClusterDelNodesRequest | 09:12 |
| Qiming | this has to be done carefully | 09:12 |
| Qiming | I have posted comments to your (abandoned) patch and the patch from your colleague | 09:12 |
| guoshan | yes, I has assigned this bug long time ago | 09:12 |
| Qiming | strictly speaking, this is not a bug | 09:13 |
| Qiming | it is a wishlist | 09:13 |
| Qiming | but never mind, if we have such a request, we can get it doen | 09:13 |
| Qiming | s/doen/done/ | 09:13 |
| Qiming | the correct way of doing this is to split that patch into several steps | 09:14 |
| Qiming | the first is to improve the engine/action to understand a 'destroy' parameter if present | 09:14 |
| Qiming | we need to review if there are things missing in such a patch | 09:14 |
| Qiming | the next patch could be bumping the version of ClusterDelNodesRequest to '1.1', its default VERSION is '1.0', adding a new field means its version has to be bumped | 09:15 |
| Qiming | you will add a "VERSION='1.1'" class property to the ClusterDelNodesRequest class | 09:16 |
| Qiming | and you will add a classmethod obj_make_compatible() to convert a 1.1 version of request object to a 1.0 version | 09:16 |
| Qiming | which means you will delete the 'destroy' field in such a conversion | 09:16 |
| Qiming | the 3rd patch is about API micro_version bumping | 09:17 |
| guoshan | yes, i know senlin is using api micro_version | 09:18 |
| Qiming | you will change the microversion to '1.4' here: http://git.openstack.org/cgit/openstack/senlin/tree/senlin/api/openstack/v1/version.py | 09:18 |
| Qiming | also add a note here: http://git.openstack.org/cgit/openstack/senlin/tree/senlin/api/openstack/history.rst | 09:19 |
| Qiming | then you add a VERSION_MAP class property to ClusterDelNodesRequest: | 09:19 |
| Qiming | VERSION_MAP = { | 09:19 |
| Qiming | '1.5': '1.1', | 09:20 |
| Qiming | } | 09:20 |
| Qiming | which means the 1.5 api version will use the revised verion of cluster-del-nodes API, all API versions below that will use the 1.0 version of the request object | 09:20 |
| Qiming | your last patch, the 4th one, will be updating api-ref and user, developer docs | 09:21 |
| Qiming | it is not a single patch thing | 09:21 |
| Qiming | but I hope the above outline helps you understand why and how to use versioned objects to manage API compatiblity in Senlin | 09:22 |
| guoshan | okay, will do it, thank you for your thorough guiding | 09:22 |
| *** elynn has quit IRC | 09:36 | |
| *** yuanying has quit IRC | 09:47 | |
| openstackgerrit | Merged openstack/senlin: Utility function for API request conversion https://review.openstack.org/409651 | 09:54 |
| openstackgerrit | Aaron Ding proposed openstack/senlin: Set max_size to cfg.CONF.max_nodes_per_cluster if -1. https://review.openstack.org/410083 | 10:02 |
| openstackgerrit | Merged openstack/senlin: Use util function in cluster API (1) https://review.openstack.org/410050 | 10:07 |
| openstackgerrit | Qiming Teng proposed openstack/senlin: Use util function in cluster API (4) https://review.openstack.org/410130 | 10:07 |
| openstackgerrit | Aaron Ding proposed openstack/senlin: Set max_size to cfg.CONF.max_nodes_per_cluster if -1. https://review.openstack.org/410083 | 10:09 |
| *** yanyanhu has quit IRC | 10:14 | |
| *** guoshan has quit IRC | 10:24 | |
| *** guoshan has joined #senlin | 10:39 | |
| *** chohoor has quit IRC | 11:00 | |
| *** guoshan has quit IRC | 11:06 | |
| *** dinghh1 has joined #senlin | 11:13 | |
| *** dinghh has quit IRC | 11:14 | |
| *** guoshan has joined #senlin | 11:14 | |
| *** guoshan has quit IRC | 11:41 | |
| *** guoshan has joined #senlin | 12:07 | |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: Remove retry logic from lock_acquire https://review.openstack.org/410095 | 12:17 |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: Lookup a random action to execute https://review.openstack.org/409805 | 12:17 |
| *** guoshan has quit IRC | 12:27 | |
| *** guoshan has joined #senlin | 12:27 | |
| *** guoshan has quit IRC | 12:32 | |
| openstackgerrit | Aaron Ding proposed openstack/senlin: Set max_size to cfg.CONF.max_nodes_per_cluster if -1. https://review.openstack.org/410083 | 12:35 |
| *** dinghh1 has left #senlin | 12:46 | |
| *** yanyanhu has joined #senlin | 12:50 | |
| yanyanhu | hi, guys, meeting will start in #openstack-meeting channel in minutes | 12:53 |
| *** lvdongbing has joined #senlin | 12:57 | |
| *** elynn has joined #senlin | 12:58 | |
| yanyanhu | hi, guys, meeting has started | 13:00 |
| *** XueFengLiu has joined #senlin | 13:06 | |
| *** XueFengLiu has quit IRC | 13:33 | |
| *** XueFengLiu has joined #senlin | 13:33 | |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: Remove retry logic from lock_acquire https://review.openstack.org/410095 | 14:07 |
| openstackgerrit | Ethan Lynn proposed openstack/senlin: Lookup a random action to execute https://review.openstack.org/409805 | 14:07 |
| *** guoshan has joined #senlin | 14:11 | |
| openstackgerrit | XueFeng Liu proposed openstack/senlin: Remove dead code about action-delete https://review.openstack.org/410235 | 14:18 |
| *** yanyanhu has quit IRC | 14:27 | |
| *** XueFengLiu has quit IRC | 14:44 | |
| *** lvdongbing has quit IRC | 15:04 | |
| *** guoshan has quit IRC | 15:05 | |
| *** guoshan has joined #senlin | 15:06 | |
| *** guoshan has quit IRC | 15:11 | |
| *** Drago1 has joined #senlin | 15:41 | |
| *** XueFengLiu has joined #senlin | 15:42 | |
| *** elynn has quit IRC | 16:11 | |
| *** elynn has joined #senlin | 16:21 | |
| -openstackstatus- NOTICE: Launchpad SSO is not currently working, so logins to our services like review.openstack.org and wiki.openstack.org are failing; the admins at Canonical are looking into the issue but there is no estimated time for a fix yet. | 16:27 | |
| *** ChanServ changes topic to "Launchpad SSO is not currently working, so logins to our services like review.openstack.org and wiki.openstack.org are failing; the admins at Canonical are looking into the issue but there is no estimated time for a fix yet." | 16:27 | |
| *** elynn has quit IRC | 16:34 | |
| *** ChanServ changes topic to "IRCLog: http://eavesdrop.openstack.org/irclogs/%23senlin/ | Bugs: bugs.launchpad.net/senlin | Review: https://review.openstack.org/#/q/project:openstack/senlin,n,z" | 17:03 | |
| -openstackstatus- NOTICE: Canonical admins have resolved the issue with login.launchpad.net, so authentication should be restored now. | 17:03 | |
| *** Drago1 has quit IRC | 21:04 | |
| *** Drago1 has joined #senlin | 21:10 | |
| *** Drago1 has quit IRC | 21:47 | |
| *** Drago1 has joined #senlin | 21:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!