| *** openstack has joined #openstack-watcher | 00:02 | |
| *** harlowja has joined #openstack-watcher | 00:19 | |
| *** thorst_ has joined #openstack-watcher | 00:33 | |
| *** sballe_ has quit IRC | 00:48 | |
| *** openstack has joined #openstack-watcher | 00:53 | |
| *** hvprash has joined #openstack-watcher | 01:09 | |
| *** jwcroppe has quit IRC | 01:40 | |
| *** jwcroppe has joined #openstack-watcher | 01:40 | |
| *** jwcroppe has quit IRC | 01:45 | |
| *** openstackgerrit has quit IRC | 01:49 | |
| *** openstackgerrit has joined #openstack-watcher | 01:56 | |
| *** openstackgerrit has quit IRC | 01:56 | |
| *** openstackgerrit has joined #openstack-watcher | 01:57 | |
| openstackgerrit | Jinquan Ni proposed openstack/watcher: Destination selectors plugin rst https://review.openstack.org/319176 | 01:57 |
|---|---|---|
| openstackgerrit | junjie.huang proposed openstack/watcher: Workload balance migration strategy implementation https://review.openstack.org/292188 | 02:02 |
| *** hvprash has quit IRC | 02:33 | |
| *** openstack has joined #openstack-watcher | 02:38 | |
| *** thorst_ has quit IRC | 02:51 | |
| *** thorst_ has joined #openstack-watcher | 02:52 | |
| *** thorst_ has quit IRC | 03:00 | |
| openstackgerrit | Edwin Zhai proposed openstack/watcher: Enable strategy parameters https://review.openstack.org/297556 | 03:48 |
| *** thorst_ has joined #openstack-watcher | 03:58 | |
| *** thorst_ has quit IRC | 04:05 | |
| *** hvprash has joined #openstack-watcher | 04:34 | |
| *** hvprash has quit IRC | 04:39 | |
| *** thorst_ has joined #openstack-watcher | 05:03 | |
| *** thorst_ has quit IRC | 05:09 | |
| *** ecelik has joined #openstack-watcher | 05:44 | |
| *** jwcroppe has joined #openstack-watcher | 05:47 | |
| *** jwcroppe has quit IRC | 05:52 | |
| *** hvprash has joined #openstack-watcher | 06:02 | |
| *** thorst_ has joined #openstack-watcher | 06:07 | |
| openstackgerrit | Edwin Zhai proposed openstack/python-watcherclient: Enable strategy parameter https://review.openstack.org/297557 | 06:13 |
| *** thorst_ has quit IRC | 06:14 | |
| *** hvprash has quit IRC | 06:15 | |
| *** vtech_ has quit IRC | 06:43 | |
| *** openstack has joined #openstack-watcher | 06:53 | |
| *** vtech has joined #openstack-watcher | 07:02 | |
| *** thorst_ has joined #openstack-watcher | 07:12 | |
| openstackgerrit | junjie.huang proposed openstack/watcher: Workload balance migration strategy implementation https://review.openstack.org/292188 | 07:17 |
| *** thorst_ has quit IRC | 07:20 | |
| *** jed56 has joined #openstack-watcher | 07:20 | |
| *** vincentfrancoise has joined #openstack-watcher | 07:23 | |
| openstackgerrit | Edwin Zhai proposed openstack/watcher: Enable strategy parameters https://review.openstack.org/297556 | 07:32 |
| *** jwcroppe has joined #openstack-watcher | 07:50 | |
| *** jwcroppe has quit IRC | 07:55 | |
| *** vtech has quit IRC | 08:07 | |
| *** vtech_ has joined #openstack-watcher | 08:07 | |
| *** thorst_ has joined #openstack-watcher | 08:17 | |
| *** thorst_ has quit IRC | 08:25 | |
| openstackgerrit | Alexander Chadin proposed openstack/watcher: Add continuously optimization https://review.openstack.org/321411 | 08:31 |
| *** alexchadin has joined #openstack-watcher | 08:39 | |
| openstackgerrit | Vladimir Ostroverkhov proposed openstack/python-watcherclient: Add support continuously-optimization https://review.openstack.org/321421 | 08:52 |
| jinquan | alexchadin hello? | 08:53 |
| alexchadin | jinquan: hi! | 08:53 |
| jinquan | I have a question about vm_workload_consolidatin | 08:54 |
| jinquan | https://review.openstack.org/#/c/305596/ | 08:54 |
| jinquan | i'm work on this BP, now I need refactor the select_destination in all startegies | 08:55 |
| jinquan | but i'm not sure the vm_fits need refactor in vm_workload_consolidation | 08:56 |
| jinquan | do you have some suggest? | 08:56 |
| alexchadin | wait a sec please | 08:57 |
| jinquan | OK thank you very much!:) | 08:57 |
| alexchadin | Name of my strategy is workload_stabilization | 08:58 |
| alexchadin | Do you mean this or vm_workload_consolidation? | 08:58 |
| vincentfrancoise | jinquan, alexchadin: both will need refactoring at some point | 08:59 |
| jinquan | I am mean vm_workload_consolidation | 09:00 |
| jinquan | and workload_stabilization has merged? | 09:00 |
| vincentfrancoise | jinquan: both names are similar so I got confused when you asked me earlier | 09:01 |
| alexchadin | No, workload_stabilization hasn't been merged yet | 09:01 |
| vincentfrancoise | sorry | 09:01 |
| vincentfrancoise | for vm_workload_consolidation, you can ask vtech_ | 09:01 |
| alexchadin | +1 | 09:01 |
| jinquan | nothing , i find the author in code, ok i will ask vtech | 09:02 |
| vincentfrancoise | but alexchadin, if you can have a look at what jinquan is doing, that would ease the refactoring ;) | 09:02 |
| jinquan | vincentfrancoise,alexchadin thank you !! | 09:03 |
| alexchadin | I will once I have a free time, ok, jinquan? | 09:03 |
| jinquan | vtech_: hello? are you online? | 09:03 |
| jinquan | alexchadin: of course!:) | 09:04 |
| vtech_ | hello guys, yep I am here. | 09:04 |
| vincentfrancoise | hi | 09:04 |
| alexchadin | vincentfrancoise: continuously-optimization patches have been commited for watcher and python-watcherclient | 09:04 |
| vtech_ | let me catch up what's going on here ;) | 09:05 |
| jinquan | can you read the info above and give us some help ? | 09:05 |
| jinquan | ok | 09:05 |
| vincentfrancoise | alexchadin: I'm doing reviews this morning, so I'll try to review it as well | 09:06 |
| alexchadin | vincentfrancoise: great :) | 09:06 |
| vtech_ | jinquan, ok, I will check it out, but I will need few minutes. | 09:08 |
| jinquan | vtech_: OK | 09:09 |
| jinquan | vincentfrancoise: can i propose new patch set that already refactor outlet_temp_consolidation? | 09:10 |
| vincentfrancoise | jinquan: go ahead but I'll review alexchadin's code first so I certainly won't review it today anyway | 09:15 |
| vtech_ | jinquan, it might be a little bit complicated here "vm_fits" (https://github.com/openstack/watcher/blob/master/watcher/decision_engine/strategy/strategies/vm_workload_consolidation.py#L374) checks whether a is a hypervisor able to accomodate VM considering cpu, ram and disk metrics including capacity coefficients (for underbooking / overbooking) | 09:16 |
| jinquan | vincentfrancoise: ok, the new patch need's more info form vtech_. then we will konw the patch will go | 09:17 |
| vtech_ | jinquan, the algorithm iterates through the destination hypervisors sorted ASC in the offloading and DSC in the consolidation phase based on PREDICTED hypervisor CPU load | 09:18 |
| jinquan | vtech_:yeah , it check one host one time and it's based on vm cpu/mem/disk util from ceilometer . | 09:19 |
| jinquan | vtech: am i right? | 09:19 |
| vtech_ | the strategy internally has a model which predicts the cpu load of hypervisor load as the migrations are added to the solution (before they are actually applied in applier) | 09:19 |
| vtech_ | the whole strategy creates a deep copy of the model in the very beginning and then operates on a "prediction model" | 09:21 |
| vtech_ | e.g. considering two host environment (host A, host B) with VM1 (on A) and VM2 (on B) | 09:22 |
| *** thorst_ has joined #openstack-watcher | 09:22 | |
| jinquan | vetch_: and you does this beacuse you need avoid the situation like "pingpong"? | 09:24 |
| jinquan | vtech_:So it seems like refactor to use select_destination are diffcult? | 09:24 |
| vtech_ | once I add a migration to the solution (lets say migrate VM1 from A->B), the resource utilization is predicted so host A util = host A util - VM A util and host B util = host B util + vm A util | 09:25 |
| vtech_ | yes IMO it might be very difficult, and I don't think it's a good idea to only have one unified way how to select destination hypervisor | 09:26 |
| jinquan | vtech_:oh, we can implement a new way in new destinations selector | 09:28 |
| vtech_ | so in the strategy we query ceilometer only once at the beginning, because as you are adding migrations you can expect the load on the host to change accordingly but this is not reflected in ceilometer metrics until you actually move the VMs arround | 09:28 |
| *** thorst_ has quit IRC | 09:30 | |
| vtech_ | ie. let's assume you have a hypervisor with 0 load. If you only query ceilometer during the computation (solution generation) this hypervisor will still appear with 0 load. But this is extremely as it doesn't reflect migration which will destinate (and thus will affect) hypervisor's load. | 09:33 |
| jed56 | hello | 09:34 |
| jinquan | vtech_ vincentfrancoise: Ok, i see, IMO, selector a destination has two aspects: watcher and nova . to watcher this is decide by difference stategy. to nova we need check cpu/mem/disk and so on. I think we should mix the two aspects | 09:34 |
| jed56 | vtech_: I agree, that every strategy can have their own concern. | 09:34 |
| jed56 | However, the strategies have to respect the policies Group (Affinity, no-affiniy, etc) | 09:34 |
| jed56 | IMHO, the stragey can propose a list of possible "destinations" and the select_destination is in charge of check the constraints. | 09:35 |
| vtech_ | jed56, true, this is not implemented in our strategy to date | 09:35 |
| jinquan | s/should/should not | 09:35 |
| vtech_ | jed56, this sounds sensible to me. | 09:36 |
| jed56 | The main aim of select_destination is verify only the constraints | 09:36 |
| jed56 | we are working with nova | 09:36 |
| jed56 | to improve the selection_destination | 09:36 |
| jed56 | https://review.openstack.org/#/c/276840/ | 09:37 |
| vtech_ | jed56, I am aware of the affinity constraints, it's a very good point | 09:37 |
| jed56 | We wants that nova accept multiple hosts when calling live migration | 09:37 |
| jed56 | this is the reason why we want to propose a "common way" to do it in watcher | 09:37 |
| jed56 | we will to adapt every strategy to call this "super checker" | 09:38 |
| jed56 | we will need to adapt every strategy to call this "super checker" | 09:39 |
| vtech_ | jed56, can you prioritize them? | 09:39 |
| vtech_ | that would solve the problem I think | 09:39 |
| vtech_ | so you only pass nova hypervisors with desired priority for where you want your VM to land | 09:40 |
| jinquan | jed56: Ok , i see . "super checker" is that i said "selector for nova aspect "? am i right? | 09:40 |
| jed56 | jinquan: yes | 09:41 |
| vtech_ | and nova scheduler will go through them by the priority and lands returns you the destination of the first match. | 09:41 |
| jed56 | vtech_: yes this is the idea | 09:42 |
| jed56 | but i'm not sure that nova will accept a list with priority :) | 09:42 |
| jed56 | the other solution is to retrieve the constraints from nova through the RequestSpec or from congress | 09:43 |
| jed56 | but I'm not sure that nova will appreciate this solution | 09:43 |
| vtech_ | jed56, yeah, but I am sure nova guys will be happier when the nova scheduler can make "the final decision" | 09:43 |
| jed56 | yes, this why they should adapt their API | 09:44 |
| vtech_ | I like the idea of sending multiple destination options, once this support prioritization, you can still make some impact on the final placement in strategy and nova guys will stay happy as the nova scheduler "approves" your choice | 09:46 |
| jed56 | vtech_: me too :) | 09:47 |
| vtech_ | but now i can see an uncertainity computing the solution in my strategy. | 09:48 |
| jed56 | yes, I agree | 09:50 |
| vtech_ | as nova decides the location when migration actually start (in applier), there is no way how to predict where the VM land during the computation. | 09:50 |
| jed56 | jinquan: I didn't take a look to your code, are you align with this discussions ? | 09:50 |
| vtech_ | and to be frank I am not 100% sure how to modify our strategy to this. | 09:51 |
| jed56 | vtech_: we can force the destination in the applier or write a plugin in nova | 09:51 |
| jed56 | vtech_: I have to go we continue the discussion later | 09:51 |
| jinquan | i'm reading your discussion | 09:51 |
| jed56 | this is a very hot topic | 09:51 |
| jed56 | We have two priority in watcher scalability and constraints | 09:52 |
| vtech_ | ok, if we can force this, there is no problem with the strategy, but then (again) we are taking some responsibillity from nova | 09:52 |
| vtech_ | I will think about this... | 09:53 |
| *** jwcroppe has joined #openstack-watcher | 09:53 | |
| jed56 | vtech_: no, if we ask nova if this is a possible solution through the select_destination | 09:53 |
| jinquan | you mean target destinations with live-migration ? | 09:53 |
| jed56 | I'm sorry i have to go | 09:53 |
| jed56 | bye | 09:53 |
| jinquan | bye | 09:54 |
| vtech_ | bye | 09:54 |
| vtech_ | jinquan, yes | 09:54 |
| jinquan | vtech_: do we have a commit bp in nova? | 09:55 |
| *** jwcroppe has quit IRC | 09:58 | |
| vtech_ | jinquan, so the conclusion is: the vm consolidation strategy computes the placement on the go ie. it requires certainty in terms of desired destination will be the actual destination and there is not much room for uncertainity afaik. I will have to think about this a little bit more. | 09:59 |
| vtech_ | jinquann, I will also check the current status of nova to see whether to blueprint allowing to specify multiple destination is a solution for this particular case. | 10:01 |
| vtech_ | I will have to go now | 10:01 |
| jinquan | OK. thank you very much | 10:01 |
| jinquan | bye | 10:01 |
| jinquan | :) | 10:01 |
| vtech_ | np, thanks for bringing it up | 10:02 |
| vtech_ | jinquan, bye | 10:02 |
| openstackgerrit | Jinquan Ni proposed openstack/watcher: Select destinations filter implementation https://review.openstack.org/305596 | 10:10 |
| openstackgerrit | Jinquan Ni proposed openstack/watcher: Select destinations filter implementation https://review.openstack.org/305596 | 10:12 |
| *** thorst_ has joined #openstack-watcher | 10:28 | |
| *** thorst_ has quit IRC | 10:35 | |
| *** alexchadin has quit IRC | 10:41 | |
| openstackgerrit | Vincent Françoise proposed openstack/python-watcherclient: Replaced UUID of goal and strategy with name https://review.openstack.org/311985 | 10:54 |
| openstackgerrit | Vincent Françoise proposed openstack/python-watcherclient: Update Watcher CLI documentation https://review.openstack.org/320451 | 10:55 |
| openstackgerrit | Vincent Françoise proposed openstack/python-watcherclient: Updated CLI to display efficacy related fields https://review.openstack.org/320966 | 10:55 |
| *** thorst_ has joined #openstack-watcher | 11:49 | |
| *** jwcroppe has joined #openstack-watcher | 11:56 | |
| *** jwcroppe has quit IRC | 12:01 | |
| *** hvprash has joined #openstack-watcher | 12:51 | |
| *** jwcroppe has joined #openstack-watcher | 12:58 | |
| *** hvprash_ has joined #openstack-watcher | 12:59 | |
| *** hvprash has quit IRC | 13:03 | |
| *** jwcroppe has quit IRC | 13:05 | |
| openstackgerrit | Merged openstack/watcher-dashboard: Added Goals and Strategies to Dashboard https://review.openstack.org/301516 | 13:21 |
| *** lrensing has joined #openstack-watcher | 13:39 | |
| *** vincentfrancoise has quit IRC | 13:43 | |
| *** vincentfrancoise has joined #openstack-watcher | 13:45 | |
| *** vincentfrancois1 has joined #openstack-watcher | 13:45 | |
| *** vincentfrancois1 has quit IRC | 13:46 | |
| *** vincentfrancoise has quit IRC | 13:46 | |
| *** vincentfrancoise has joined #openstack-watcher | 13:52 | |
| *** ecelik has quit IRC | 13:59 | |
| *** jwcroppe has joined #openstack-watcher | 14:57 | |
| *** hvprash_ has quit IRC | 15:13 | |
| *** hvprash has joined #openstack-watcher | 15:13 | |
| *** hvprash has quit IRC | 15:48 | |
| *** hvprash has joined #openstack-watcher | 15:48 | |
| *** vtech_ has quit IRC | 15:56 | |
| *** vincentfrancoise has quit IRC | 16:21 | |
| *** vtech has joined #openstack-watcher | 16:26 | |
| *** wootehfoot has joined #openstack-watcher | 16:47 | |
| *** hvprash has quit IRC | 17:29 | |
| *** hvprash has joined #openstack-watcher | 17:31 | |
| *** ss4 has joined #openstack-watcher | 17:33 | |
| *** ss4 has quit IRC | 17:35 | |
| *** ss4 has joined #openstack-watcher | 17:35 | |
| *** wootehfoot has quit IRC | 17:36 | |
| *** hvprash_ has joined #openstack-watcher | 17:42 | |
| *** hvprash has quit IRC | 17:44 | |
| *** hvprash has joined #openstack-watcher | 17:46 | |
| *** hvprash_ has quit IRC | 17:49 | |
| *** ss4 has quit IRC | 18:31 | |
| *** lrensing has quit IRC | 19:14 | |
| *** esberglu has joined #openstack-watcher | 19:59 | |
| *** thorst_ has quit IRC | 21:10 | |
| *** thorst_ has joined #openstack-watcher | 21:10 | |
| *** thorst__ has joined #openstack-watcher | 21:12 | |
| *** thorst_ has quit IRC | 21:15 | |
| *** thorst__ has quit IRC | 21:16 | |
| *** hvprash_ has joined #openstack-watcher | 22:13 | |
| *** hvprash_ has quit IRC | 22:13 | |
| *** hvprash has quit IRC | 22:16 | |
| *** jwcroppe_ has joined #openstack-watcher | 23:21 | |
| *** jwcroppe has quit IRC | 23:23 | |
| *** hvprash has joined #openstack-watcher | 23:47 | |
| *** hvprash has quit IRC | 23:51 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!