*** ianychoi has quit IRC | 01:17 | |
*** ianychoi has joined #openstack-watcher | 01:28 | |
*** ianychoi has quit IRC | 01:42 | |
*** ianychoi has joined #openstack-watcher | 02:19 | |
*** TuanVu has joined #openstack-watcher | 03:21 | |
*** logan- has quit IRC | 07:19 | |
*** logan- has joined #openstack-watcher | 07:32 | |
*** alexchadin has joined #openstack-watcher | 07:40 | |
*** logan- has quit IRC | 08:13 | |
*** alexchadin has quit IRC | 08:21 | |
*** logan- has joined #openstack-watcher | 08:27 | |
*** logan- has quit IRC | 08:51 | |
*** logan- has joined #openstack-watcher | 09:05 | |
*** alexchadin has joined #openstack-watcher | 09:12 | |
TuanVu | Hi Team | 09:20 |
---|---|---|
TuanVu | have a nice day | 09:20 |
TuanVu | I would like to have a question related to Live Migration | 09:20 |
alexchadin | TuanVu: hi | 09:21 |
*** openstackgerrit has joined #openstack-watcher | 09:22 | |
openstackgerrit | Tetiana Lashchova proposed openstack/watcher master: Do not pass www_authenticate_uri to RequestContext https://review.openstack.org/607144 | 09:22 |
TuanVu | Hi Alex, nice to see you :) | 09:22 |
alexchadin | TuanVu: nice to see you too! | 09:22 |
TuanVu | I have deployed Watcher with Devstack in multi-node environment: 1 controller (compute as well) & another compute node | 09:23 |
TuanVu | I have created 1 VM, currently it's located on compute1 | 09:24 |
TuanVu | now I want to live migrate it to compute2 | 09:24 |
TuanVu | as far as I understand, I'll need to create a new audit template first | 09:25 |
*** kholkina has joined #openstack-watcher | 09:25 | |
TuanVu | the goal is: Hardware Maintenance, right? | 09:25 |
alexchadin | TuanVu: wait a sec | 09:38 |
TuanVu | sure, no problem, Alex :) | 09:39 |
alexchadin | TuanVu: you will need cluster_maintaining goal | 09:40 |
TuanVu | thanks, let me test it right now | 09:41 |
TuanVu | what about strategy? | 09:42 |
TuanVu | host maintainance, right? | 09:42 |
alexchadin | TuanVu: Host Maintenance strategy | 09:42 |
alexchadin | TuanVu: yep, right | 09:42 |
TuanVu | thanks | 09:42 |
TuanVu | and is that enough? | 09:43 |
TuanVu | do I need anything for the scope? | 09:43 |
alexchadin | TuanVu: you may specify your two nodes in the scope | 09:45 |
alexchadin | TuanVu: just to make sure that this strategy won't affect other resources | 09:46 |
TuanVu | don't worry, I have only 2 nodes and it's test environment :) | 09:46 |
alexchadin | TuanVu: get back soon, going to lunch | 09:46 |
TuanVu | please enjoy your lunch :) | 09:46 |
TuanVu | thanks for your great help, Alex | 09:46 |
alexchadin | TuanVu: thanks a lot! | 09:46 |
*** alexchadin has quit IRC | 09:46 | |
*** alexchadin has joined #openstack-watcher | 10:08 | |
alexchadin | TuanVu: so what's up with your tests? | 10:09 |
TuanVu | oh, that's so quick | 10:12 |
TuanVu | have you finished your lunch yet? | 10:12 |
alexchadin | yep | 10:14 |
TuanVu | that's great :D | 10:14 |
TuanVu | btw, I've just followed this | 10:14 |
TuanVu | https://docs.openstack.org/watcher/latest/strategies/host_maintenance.html#how-to-use-it | 10:14 |
TuanVu | somehow, watcher-decision-engine.service raised error: | 10:15 |
TuanVu | ComputeNodeNotFound: The compute node compute could not be found | 10:15 |
TuanVu | I've already made sure that its name (compute) is correct | 10:15 |
TuanVu | $openstack compute service list | 10:15 |
TuanVu | nova-compute | devstack | nova | enabled | up | 10:16 |
TuanVu | nova-compute | compute | nova | enabled | up | 10:16 |
TuanVu | _______________________________ | 10:16 |
TuanVu | here's my command: | 10:16 |
TuanVu | $ openstack optimize audit create -g cluster_maintaining -s host_maintenance -p maintenance_node=devstack -p backup_node=compute --auto-trigger | 10:16 |
*** alexchadin has quit IRC | 10:17 | |
*** alexchadin has joined #openstack-watcher | 10:21 | |
alexchadin | TuanVu: could you please send me your log of decision engine? | 10:21 |
TuanVu | Hi Alex, sure | 10:22 |
TuanVu | http://paste.openstack.org/show/731347/ | 10:23 |
TuanVu | here it is | 10:23 |
alexchadin | TuanVu: just tried this strategy in my env, it worked | 10:32 |
alexchadin | let's see | 10:32 |
TuanVu | thanks for your confirmation | 10:33 |
alexchadin | TuanVu: what is your ModelRoot? | 10:33 |
TuanVu | sorry, could you please explain more? | 10:33 |
alexchadin | TuanVu: could you please provide 100-200 lines of decision engine log? | 10:33 |
alexchadin | 100-200 last lines | 10:33 |
*** alexchadin has quit IRC | 10:36 | |
TuanVu | Hi Alex, here's the 200 last lines | 10:40 |
TuanVu | http://paste.openstack.org/show/731349/ | 10:40 |
*** alexchadin has joined #openstack-watcher | 10:54 | |
alexchadin | TuanVu: look at 157-160 lines | 10:58 |
alexchadin | TuanVu: there is only one compute node | 10:59 |
TuanVu | thanks for your analysis, Alex | 11:23 |
TuanVu | do I need to enable Watcher on Compute Node? | 11:24 |
TuanVu | as far as I understand, we only need Watcher on Controller, right? | 11:24 |
alexchadin | TuanVu: that's right, only on controller | 11:27 |
TuanVu | it's kind of strange that Watcher only recognize 1 compute node while "openstack compute service list" is showing 2 nova-compute properly | 11:28 |
alexchadin | let me think | 11:29 |
alexchadin | TuanVu: what nova api do you have? | 11:31 |
TuanVu | I've just restarted watcher*.service and result is still the same | 11:32 |
TuanVu | ServiceService Endpoint | 11:32 |
TuanVu | Computehttp://10.0.0.1/compute/v2.1 | 11:32 |
TuanVu | Compute_Legacyhttp://10.0.0.1/compute/v2/c708a7c280f649dc88cefffdaddfb562 | 11:32 |
TuanVu | this is what you mean, right? | 11:32 |
alexchadin | TuanVu: curl -g -i -X GET http://10.0.0.1/compute/ -H "User-Agent: python-novaclient" -H "Accept: application/json" -H "X-Auth-Token: your_token_here" | 11:37 |
alexchadin | TuanVu: you can get token by "openstack token issue" | 11:37 |
TuanVu | thanks | 11:42 |
TuanVu | by "openstack token issue" I got id of the token | 11:43 |
TuanVu | here it is | 11:43 |
TuanVu | gAAAAABbtKq7V9YaAE8dxyIudCxzEIsEUk0FA_hkF8Lr4oHF_eiGdciEJMNeCEspQX9w9t96BFAYNHtn25IZ_i47FsHgWQL3YJiRvkq-ajtVazuaoeb0rZm1BG_Gc8nifkYe-uw3Y3PUK3GJH-45L5wTCRttvJmuYHRT0uv-SQ7MEvPB1GCZ_tQ | 11:43 |
TuanVu | then I apply it to your "curl" command | 11:43 |
TuanVu | but it does not return anything | 11:43 |
TuanVu | it looks like problem with network, I've just replace the IP by 127.0.0.1 | 11:44 |
TuanVu | and it returns result | 11:44 |
TuanVu | let me try to fix this | 11:44 |
TuanVu | thanks a lot :) | 11:45 |
alexchadin | TuanVu: you should see something like {"versions": [{"status": "SUPPORTED", "updated": "2011-01-21T11:33:21Z", "links": [{"href": "http://controller:8774/v2/", "rel": "self"}], "min_version": "", "version": "", "id": "v2.0"}, {"status": "CURRENT", "updated": "2013-07-23T11:33:21Z", "links": [{"href": "http://controller:8774/v2.1/", "rel": "self"}], "min_version": "2.1", "version": "2.53", "id": "v2.1"}]} | 11:52 |
TuanVu | yes, I got it :) | 11:52 |
alexchadin | I have 2.53 version | 11:52 |
TuanVu | there's some problem with network proxy in my local environment | 11:52 |
alexchadin | what version do you have? | 11:52 |
TuanVu | I'm fixing it :) | 11:52 |
TuanVu | I got 2.66 | 11:52 |
TuanVu | thank you very much :D | 11:52 |
alexchadin | TuanVu: try to add section [nova_client] in Watcher config file with option api_version = 2.66 | 11:57 |
*** alexchadin has quit IRC | 11:58 | |
*** alexchadin has joined #openstack-watcher | 12:09 | |
*** TuanVu has quit IRC | 12:20 | |
*** kholkina has quit IRC | 14:05 | |
*** hoangcx has quit IRC | 14:06 | |
*** alexchadin has quit IRC | 14:11 | |
*** hoangcx has joined #openstack-watcher | 14:22 | |
*** hoangcx has quit IRC | 16:12 | |
*** hoangcx has joined #openstack-watcher | 16:13 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!