| *** hidekazu has joined #openstack-watcher | 00:31 | |
| *** thorst has joined #openstack-watcher | 00:41 | |
| *** thorst has quit IRC | 00:45 | |
| *** thorst has joined #openstack-watcher | 00:46 | |
| *** thorst has quit IRC | 00:46 | |
| *** wootehfoot has joined #openstack-watcher | 00:58 | |
| *** thorst has joined #openstack-watcher | 02:47 | |
| *** thorst has quit IRC | 02:52 | |
| *** atuly_ has quit IRC | 03:12 | |
| *** thorst has joined #openstack-watcher | 03:14 | |
| *** thorst has quit IRC | 03:14 | |
| *** atuly_ has joined #openstack-watcher | 03:25 | |
| openstackgerrit | Cao Xuan Hoang proposed openstack/watcher: Fix a typo in watcher/objects/base.py https://review.openstack.org/420496 | 04:21 |
|---|---|---|
| *** diga has joined #openstack-watcher | 04:30 | |
| *** thorst has joined #openstack-watcher | 05:15 | |
| *** thorst has quit IRC | 05:20 | |
| *** atuly_ has quit IRC | 06:42 | |
| *** atuly_ has joined #openstack-watcher | 06:55 | |
| *** thorst has joined #openstack-watcher | 07:16 | |
| *** thorst has quit IRC | 07:21 | |
| *** atuly_ has quit IRC | 07:33 | |
| openstackgerrit | Merged openstack/watcher: Fix a typo in watcher/objects/base.py https://review.openstack.org/420496 | 07:44 |
| *** atuly_ has joined #openstack-watcher | 07:45 | |
| *** atuly_ has quit IRC | 07:54 | |
| *** zhurong has joined #openstack-watcher | 08:05 | |
| *** atuly_ has joined #openstack-watcher | 08:08 | |
| openstackgerrit | David TARDIVEL proposed openstack/watcher: Added tempest test for workload_stabilization https://review.openstack.org/393355 | 08:20 |
| *** vincentfrancoise has joined #openstack-watcher | 08:33 | |
| *** vincentfrancoise has left #openstack-watcher | 08:34 | |
| *** alexchadin has joined #openstack-watcher | 08:34 | |
| *** akuleshov1 has joined #openstack-watcher | 08:34 | |
| *** vincentfrancoise has joined #openstack-watcher | 08:34 | |
| *** adisky_ has joined #openstack-watcher | 08:35 | |
| openstackgerrit | Vincent Françoise proposed openstack/watcher-specs: Specs for action-plan-versioned-notifications-api https://review.openstack.org/420576 | 08:39 |
| openstackgerrit | Alexander Chadin proposed openstack/watcher: New default planner https://review.openstack.org/406991 | 08:55 |
| openstackgerrit | Alexander Chadin proposed openstack/python-watcherclient: Add functional tests to watcherclient https://review.openstack.org/389181 | 09:03 |
| openstackgerrit | Alexander Chadin proposed openstack/python-watcherclient: Support parents field along with planner changes https://review.openstack.org/409126 | 09:03 |
| *** thorst has joined #openstack-watcher | 09:17 | |
| *** thorst has quit IRC | 09:22 | |
| *** zhurong has quit IRC | 09:37 | |
| openstackgerrit | Merged openstack/watcher: Added Monasca Helper https://review.openstack.org/348891 | 09:54 |
| openstackgerrit | Merged openstack/watcher: Multi datasource support for Basic Consolidation https://review.openstack.org/414617 | 09:54 |
| alexchadin | vincentfrancoise: hi | 10:02 |
| vincentfrancoise | alexchadin: morning | 10:02 |
| alexchadin | I test your weight planner and there is strange behaviour: I set parallelization parameter to 3 for migrate, resize and nop types. So, nop's are going first and both of them have no parents (it's ok), but migrate actions still have sequence links | 10:06 |
| alexchadin | vincentfrancoise: http://pastebin.com/imehmT9d | 10:07 |
| openstackgerrit | Merged openstack/watcher-dashboard: Add auto_trigger option in Audit creation form https://review.openstack.org/411832 | 10:08 |
| *** zhurong has joined #openstack-watcher | 10:08 | |
| vincentfrancoise | alexchadin: let me see | 10:09 |
| vincentfrancoise | first column is the UUID and 2nd are the parents right? | 10:10 |
| *** dtardivel has joined #openstack-watcher | 10:10 | |
| alexchadin | right | 10:11 |
| alexchadin | parallelization = { | 10:12 |
| alexchadin | 'turn_host_to_acpi_s3_state': 2, | 10:12 |
| alexchadin | 'resize': 3, | 10:12 |
| alexchadin | 'migrate': 3, | 10:12 |
| alexchadin | 'sleep': 1, | 10:12 |
| alexchadin | 'change_nova_service_state': 1, | 10:12 |
| alexchadin | 'nop': 3, | 10:12 |
| alexchadin | } | 10:12 |
| alexchadin | vincentfrancoise: here is my settings | 10:12 |
| vincentfrancoise | alexchadin: Ok so we get something like this http://pastebin.com/ZbZDGpiS | 10:15 |
| vincentfrancoise | which is strange indeed | 10:15 |
| alexchadin | yeap | 10:16 |
| vincentfrancoise | first guess is that maybe the parallelization parameter is somehow ignored | 10:17 |
| alexchadin | vincentfrancoise: I've tested it with default parameters | 10:18 |
| alexchadin | vincentfrancoise: nop was in sequence with parallels 1 value | 10:18 |
| alexchadin | vincentfrancoise: so parameters aren't ignored | 10:18 |
| vincentfrancoise | did you have a look at what you get in https://review.openstack.org/#/c/406991/17/watcher/decision_engine/planner/weight.py@105 as N for the migrate acitons? | 10:18 |
| openstackgerrit | Merged openstack/python-watcherclient: Add auto_trigger support https://review.openstack.org/403716 | 10:18 |
| openstackgerrit | Alexander Chadin proposed openstack/python-watcherclient: Add functional tests to watcherclient https://review.openstack.org/389181 | 10:20 |
| openstackgerrit | Alexander Chadin proposed openstack/python-watcherclient: Support parents field along with planner changes https://review.openstack.org/409126 | 10:20 |
| vincentfrancoise | alexchadin: BTW, the auto trigger stuffs are all merged now so can you change it to implemented in https://blueprints.launchpad.net/watcher/+spec/automatic-triggering-audit ? | 10:20 |
| alexchadin | vincentfrancoise: nope | 10:20 |
| alexchadin | vincentfrancoise: Done | 10:21 |
| vincentfrancoise | alexchadin: good 1/5 is implemented :p | 10:22 |
| vincentfrancoise | alexchadin: did you check the value yet? | 10:32 |
| alexchadin | vincentfrancoise: in progress | 10:32 |
| vincentfrancoise | alexchadin: ok ;) | 10:32 |
| alexchadin | n = 3 | 10:37 |
| alexchadin | vincentfrancoise: * | 10:37 |
| vincentfrancoise | :s | 10:37 |
| vincentfrancoise | alexchadin: it's weird because I made a unit test with n=3 and it was working... | 10:41 |
| vincentfrancoise | alexchadin: https://review.openstack.org/#/c/406991/17/watcher/tests/decision_engine/planner/test_weight_planner.py@623 | 10:41 |
| openstackgerrit | Merged openstack/watcher-dashboard: Add page title for some panel table https://review.openstack.org/416798 | 10:46 |
| vincentfrancoise | alexchadin: just looking once more at the tests I wrote, I didn't check the end result within the DB | 10:46 |
| vincentfrancoise | alexchadin: so maybe the graph is correct but it doesn't get saved correctly | 10:46 |
| alexchadin | vincentfrancoise: leaf_groups contains [[Action(action_plan=<?>,action_plan_id=311,action_type='sleep',created_at=<?>,deleted_at=<?>,id=<?>,input_parameters={duration=5.0},parents=['e0fb6d2c-b4eb-4e60-91b7-8900ca824ec4','128da9c6-0cd9-4fc7-aa14-9e1d7278c925'] ,state=<?>,updated_at=<?>,uuid=d340e173-68f7-42ed-acd3-a891ec3c3209)]] | 10:47 |
| vincentfrancoise | alexchadin: it's difficult to debug without actually getting myself into code :( | 10:50 |
| vincentfrancoise | alexchadin: I can have a look this afternoon if you still haven't found the issue until then if you want | 10:51 |
| alexchadin | vincentfrancoise: I let you know if I get through it | 10:52 |
| vincentfrancoise | alexchadin: thx | 10:54 |
| openstackgerrit | Vincent Françoise proposed openstack/watcher-specs: Specs for action-plan-versioned-notifications-api https://review.openstack.org/420576 | 10:57 |
| openstackgerrit | Vincent Françoise proposed openstack/watcher: Fix multinode tempest test failure https://review.openstack.org/419544 | 11:00 |
| *** hidekazu has quit IRC | 11:02 | |
| openstackgerrit | Merged openstack/watcher-dashboard: Removed unnecessary utf-8 encoding https://review.openstack.org/418817 | 11:04 |
| openstackgerrit | David TARDIVEL proposed openstack/watcher-dashboard: Fix bad parsing of HTTP response https://review.openstack.org/420645 | 11:08 |
| openstackgerrit | Merged openstack/python-watcherclient: Enable coverage report in console output https://review.openstack.org/419728 | 11:09 |
| openstackgerrit | David TARDIVEL proposed openstack/watcher-dashboard: Fix bad parsing of HTTP response https://review.openstack.org/420645 | 11:09 |
| openstackgerrit | Merged openstack/watcher-dashboard: Fix bad parsing of HTTP response https://review.openstack.org/420645 | 11:18 |
| *** thorst has joined #openstack-watcher | 11:18 | |
| *** thorst has quit IRC | 11:23 | |
| alexchadin | vincentfrancoise: ping | 11:31 |
| *** atuly_ has quit IRC | 12:35 | |
| *** thorst has joined #openstack-watcher | 12:45 | |
| vincentfrancoise | alexchadin: pong | 12:47 |
| *** atuly_ has joined #openstack-watcher | 12:47 | |
| alexchadin | vincentfrancoise: I've found where the trouble is | 12:47 |
| vincentfrancoise | alexchadin: ah! what was it? did you fix it? | 12:48 |
| alexchadin | vincentfrancoise: When sleep is added, two migrate actions are the next. the first migrate action pop sleep from leaf_groups and then add itself to leaf_groups. The second migrate action pop the first migrate action from leaf_groups and make it as parents | 12:50 |
| vincentfrancoise | alexchadin: the thing is that you should have the 2 migrates into the same action chunk | 12:54 |
| vincentfrancoise | this is the weird part IMHO | 12:54 |
| vincentfrancoise | alexchadin: I'll be back in 5mn | 12:54 |
| alexchadin | vincentfrancoise: ok | 12:54 |
| *** zhurong has quit IRC | 13:03 | |
| vincentfrancoise | alexchadin: back | 13:06 |
| alexchadin | vincentfrancoise: I've fixed it | 13:07 |
| alexchadin | vincentfrancoise: one second | 13:07 |
| vincentfrancoise | alexchadin: nice!!! | 13:07 |
| alexchadin | vincentfrancoise: take a look | 13:08 |
| alexchadin | http://pastebin.com/p7LfeJHY | 13:08 |
| vincentfrancoise | oh ok so you pop the action a the end of the iteration | 13:09 |
| alexchadin | vincentfrancoise: yeap:) | 13:14 |
| vincentfrancoise | alexchadin: can you write the extra unit test that covers the error you found? | 13:15 |
| alexchadin | vincentfrancoise: all unit tests are still passes since they do not cover this error | 13:15 |
| alexchadin | vincentfrancoise: yes, will do now | 13:15 |
| vincentfrancoise | alexchadin: with this I guess we should be all good to merge :) | 13:16 |
| alexchadin | vincentfrancoise: looks good to me :D | 13:16 |
| vincentfrancoise | alexchadin: why don't you have the servionica license in watcher/applier/actions/resize.py ? | 13:17 |
| alexchadin | vincentfrancoise: oh, you know, sometimes I added licenses to files and someone (I don't remember who it was) said that I should remove company license along with First/Last name | 13:19 |
| vincentfrancoise | alexchadin: really? As long as it's apache 2 I don't see any problem | 13:19 |
| alexchadin | vincentfrancoise: me too, but since that time I don't place any license stuff | 13:20 |
| alexchadin | vincentfrancoise: well, if it is good, I will place it:) | 13:20 |
| vincentfrancoise | alexchadin: I just made a review on the code | 13:25 |
| alexchadin | vincentfrancoise: let me see | 13:26 |
| vincentfrancoise | alexchadin: only nits overall, expect some TODOs I left in https://review.openstack.org/#/c/415505/2/watcher/decision_engine/planner/weight.py that you didn't see | 13:26 |
| alexchadin | vincentfrancoise: I've seen them, they just aren't pushed yet | 13:27 |
| vincentfrancoise | oh ok great then :) | 13:27 |
| *** diga has quit IRC | 13:59 | |
| *** jwcroppe_ has quit IRC | 14:00 | |
| *** diga has joined #openstack-watcher | 14:00 | |
| *** tmetsch has joined #openstack-watcher | 14:03 | |
| acabot | diga : hi | 14:05 |
| diga | acabot: Hi | 14:05 |
| acabot | diga : how are you ? | 14:05 |
| diga | acabot: Good, how r u ? | 14:05 |
| acabot | diga : good thx | 14:06 |
| diga | acabot: Happy New year to you! | 14:06 |
| acabot | diga : I'm still waiting for your changes to be abandonned https://review.openstack.org/#/q/owner:digambarpatil15%2540yahoo.co.in+status:open+project:openstack/watcher | 14:06 |
| acabot | diga : thanks happy new year | 14:06 |
| acabot | diga : as alexchadin already implemented this feature | 14:06 |
| diga | acabot: I thought that work is taken by alexchadin | 14:07 |
| diga | yes | 14:07 |
| acabot | diga : yes but you still need to click on the "abandon" button | 14:07 |
| acabot | diga : I can't do it myself | 14:07 |
| diga | acabot: okay, I will abandon | 14:07 |
| acabot | diga : thx | 14:07 |
| diga | acabot: Done! | 14:08 |
| acabot | good thank you | 14:08 |
| acabot | tmetsch : hi | 14:08 |
| tmetsch | acabot: howdy :-) | 14:09 |
| acabot | tmetsch : happy new year ;-) | 14:10 |
| acabot | tmetsch : I heard that you will be in Atlanta ? | 14:10 |
| tmetsch | acabot: same to you :-) not 100% sure yet, but looks good. | 14:10 |
| tmetsch | acabot: I was actually about to chat with you about the graph model. | 14:11 |
| acabot | yes this is the point I'd like to discuss | 14:11 |
| acabot | vincentfrancoise told me about your discussion a couple of days ago | 14:11 |
| tmetsch | acabot: cool - what are your thoughts on this? how can we progress best from here? | 14:12 |
| acabot | as you may know, Watcher rely only on Nova to create the current cluster data model | 14:12 |
| acabot | its very limited but we really want to avoid having to deal with Neutron & Cinder API changes | 14:13 |
| acabot | the graph model Greg suggested and we discussed in Austin include compute, network and storage resources | 14:13 |
| *** jwcroppe has joined #openstack-watcher | 14:13 | |
| tmetsch | yeah I can see that for sure. Part of the BP & specs was however to introduce the dependencies/breaking out of the compute domain so we can do more fine grained strategies. | 14:14 |
| acabot | it would really improve Watcher in the long run but we need to maintain storage & network info by catching cinder & neutron notifications | 14:14 |
| tmetsch | agreed - so I would love to do some stuff on that for the next release cycle. | 14:14 |
| acabot | and I think this is something quite big that we cannot handle on our side | 14:14 |
| tmetsch | going to be hard to do it for ocata I think :-) | 14:15 |
| acabot | I will have to reduce bcom resources involved in Watcher during the next cycle | 14:15 |
| acabot | so this is something that would be great for Watcher but we need more contributors involved | 14:15 |
| acabot | vincentfrancoise suggested to release the graph model with network & storage resources as "experimental" | 14:16 |
| acabot | but I dont really like that because we are targeting a v1.0 for Watcher at the end of the ocata cycle | 14:16 |
| acabot | and want to avoid having experimental features in a 1.0 :-) | 14:17 |
| tmetsch | that is a good point indeed. | 14:17 |
| acabot | so I would suggest to comment all code related to network & storage, merge the graph model, submit new blueprints regarding neutron & cinder notifications for the next cycle | 14:18 |
| acabot | and then uncomment your code when we will have notifications handled | 14:19 |
| acabot | what do you think ? | 14:19 |
| tmetsch | well if that is the fastest way to get started I'm fine I guess. It'll put some of our plans for the next cycle a bit under pressure as we wanted to work with a graph model which has net & storage. | 14:21 |
| tmetsch | but maybe we can figure that out somehow. | 14:21 |
| tmetsch | so we would +2 this PS? https://review.openstack.org/#/c/418434/ | 14:23 |
| acabot | yes this is the mapping of your graph model to our existing cluster data model | 14:24 |
| acabot | I think at the end we will keep only the graph model, as it is a lot easier to use and store | 14:24 |
| acabot | for HA reason | 14:24 |
| tmetsch | jah I fully support that. | 14:24 |
| tmetsch | the graph model is pretty powerful for sure :-) | 14:25 |
| tmetsch | esp to breaking the compute, network & storage domain...let me check with Kevin and see if he can take a look at the PS. | 14:25 |
| tmetsch | We'll add comments if we see sth we could improve. | 14:26 |
| tmetsch | would that be a good way forward? | 14:26 |
| acabot | yes and again I fully agree with you that we need network and storage to build more advanced strategies, my point is on dev support | 14:26 |
| tmetsch | jah got that, we'll figure that out collectively :-) | 14:26 |
| acabot | yes it would be great if you can review https://review.openstack.org/#/c/418434/ | 14:26 |
| acabot | vincentfrancoise told me this morning that its almost complete, so feel free to comment | 14:27 |
| tmetsch | ok I'll take that as an action item with Kevin. | 14:27 |
| acabot | thanks a lot | 14:27 |
| acabot | we should merge it by ocata-3 - January 23rd | 14:28 |
| tmetsch | (y) thanks for the support, vincentfrancoise did a good job there too. | 14:28 |
| acabot | so please try to be responsive ;-) | 14:28 |
| tmetsch | for sure we'll tackle that before that date :-) | 14:28 |
| tmetsch | acabot: kevin takes a look now - we might have some minor comments (if any) on the PS by Wednesday... | 14:50 |
| alexchadin | vincentfrancoise: ping | 14:56 |
| *** alexchad_ has joined #openstack-watcher | 15:12 | |
| *** alexchadin has quit IRC | 15:12 | |
| *** alexchad_ has quit IRC | 15:29 | |
| acabot | tmetsch : thx a lot to both of you | 15:38 |
| *** alexchadin has joined #openstack-watcher | 15:39 | |
| tmetsch | jo | 15:40 |
| tmetsch | np :-) | 15:40 |
| *** wootehfoot has quit IRC | 15:44 | |
| openstackgerrit | David TARDIVEL proposed openstack/watcher: Update Server Consolidation global efficacy https://review.openstack.org/417867 | 15:56 |
| vincentfrancoise | alexchadin: pong | 16:15 |
| alexchadin | vincentfrancoise: I'm reviewing weight planner | 16:16 |
| alexchadin | vincentfrancoise: get_leaf_groups method is not necessary | 16:16 |
| alexchadin | vincentfrancoise: :| | 16:17 |
| *** tmetsch has quit IRC | 16:19 | |
| vincentfrancoise | alexchadin: I don't have it in front of me but I weote this code in 1-2 days so that doesn't surprise me much :p | 16:20 |
| dtardivel | PC portable pour un déplacement pro | 16:57 |
| openstackgerrit | Alexander Chadin proposed openstack/watcher: New default planner https://review.openstack.org/406991 | 17:03 |
| openstackgerrit | Alexander Chadin proposed openstack/python-watcherclient: Support parents field along with planner changes https://review.openstack.org/409126 | 17:05 |
| openstackgerrit | Alexander Chadin proposed openstack/python-watcherclient: Add functional tests to watcherclient https://review.openstack.org/389181 | 17:08 |
| *** alexchadin has quit IRC | 17:11 | |
| *** akuleshov1 has left #openstack-watcher | 17:17 | |
| *** vincentfrancoise has quit IRC | 17:23 | |
| *** acabot has quit IRC | 17:24 | |
| *** diga has quit IRC | 17:39 | |
| *** wootehfoot has joined #openstack-watcher | 17:47 | |
| openstackgerrit | Alexander Chadin proposed openstack/watcher: New default planner https://review.openstack.org/406991 | 18:06 |
| openstackgerrit | Alexander Chadin proposed openstack/watcher: New default planner https://review.openstack.org/406991 | 20:12 |
| *** dtardivel has quit IRC | 20:14 | |
| *** adisky_ has quit IRC | 20:19 | |
| *** absubram_ has joined #openstack-watcher | 21:46 | |
| *** wootehfoot has quit IRC | 21:51 | |
| *** thorst has quit IRC | 21:55 | |
| *** absubram_ has quit IRC | 22:00 | |
| *** thorst has joined #openstack-watcher | 22:24 | |
| *** thorst has quit IRC | 22:28 | |
| *** jwcroppe has quit IRC | 23:26 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!