| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2021-05-04 | |||
| 14:08:21 | bauzas | honestly, instead of having different AZs between nova and cinder, the cinder-scheduler should look at Nova AZs | |
| 14:08:21 | bauzas | honestly, instead of having different AZs between nova and cinder, the cinder-scheduler should look at Nova AZs | |
| 14:08:34 | bauzas | (hint : placement can help) | |
| 14:08:34 | bauzas | (hint : placement can help) | |
| 14:09:47 | tobias-urdin | yeah - as I said, it's a messy case. Since spawning new ones works fine because nova when creating volumes will honor cross_az_attach https://github.com/openstack/nova/blob/master/nova/virt/block_device.py#L53 | |
| 14:09:47 | tobias-urdin | yeah - as I said, it's a messy case. Since spawning new ones works fine because nova when creating volumes will honor cross_az_attach https://github.com/openstack/nova/blob/master/nova/virt/block_device.py#L53 | |
| 14:10:16 | tobias-urdin | it just don't care on resize and can select compute node in az2 that cannot access storage in az1 for example | |
| 14:10:16 | tobias-urdin | it just don't care on resize and can select compute node in az2 that cannot access storage in az1 for example | |
| 14:12:40 | tobias-urdin | consider this 1) you have a cloud with one az, nothing much to care about you spawn a instance requestspec gets availability_zone=None you are happy 2) you add a availability zone for nova and cinder same name, you set nova.conf cross_az_attach=false because compute nodes in az1 is not addressed on storage network for az2 | |
| 14:12:40 | tobias-urdin | consider this 1) you have a cloud with one az, nothing much to care about you spawn a instance requestspec gets availability_zone=None you are happy 2) you add a availability zone for nova and cinder same name, you set nova.conf cross_az_attach=false because compute nodes in az1 is not addressed on storage network for az2 | |
| 14:13:14 | tobias-urdin | now you try to resize a instance created before that has requestspec availability_zone=None it will still try to select a compute node in az2 since it doesn't care about not attaching across AZs | |
| 14:13:14 | tobias-urdin | now you try to resize a instance created before that has requestspec availability_zone=None it will still try to select a compute node in az2 since it doesn't care about not attaching across AZs | |
| 14:13:54 | tobias-urdin | my only point, it only cares about handling that volume connection upon creation, everything else is just not implemented, nothing more so i'll dig a see what i can figure out :) | |
| 14:13:54 | tobias-urdin | my only point, it only cares about handling that volume connection upon creation, everything else is just not implemented, nothing more so i'll dig a see what i can figure out :) | |
| 14:16:52 | bauzas | tobias-urdin: again, I understand your problem, as I said, I'm not happy that cinder provided this option without thinking about move operations | |
| 14:16:52 | bauzas | tobias-urdin: again, I understand your problem, as I said, I'm not happy that cinder provided this option without thinking about move operations | |
| 14:17:10 | tobias-urdin | agree | |
| 14:17:10 | tobias-urdin | agree | |
| 14:17:27 | tobias-urdin | thanks for clearing things up, was a little bit confused but i understand the issues atleast | |
| 14:17:27 | tobias-urdin | thanks for clearing things up, was a little bit confused but i understand the issues atleast | |
| 14:17:43 | bauzas | fixing this would need us to discuss between nova and cinder to see what to do | |
| 14:17:43 | bauzas | fixing this would need us to discuss between nova and cinder to see what to do | |
| 14:18:15 | bauzas | but I wouldn't be happy with only nova fixing this because of cinder | |
| 14:18:15 | bauzas | but I wouldn't be happy with only nova fixing this because of cinder | |
| 14:18:44 | bauzas | and now we have placement, maybe we should discuss how to provide a query that would ask for this | |
| 14:18:44 | bauzas | and now we have placement, maybe we should discuss how to provide a query that would ask for this | |
| 14:19:11 | bauzas | that's a huge feature and this shouldn't be backported | |
| 14:19:11 | bauzas | that's a huge feature and this shouldn't be backported | |
| 14:19:33 | bauzas | for example, I could imagine some way to map between Nova and Cinder AZs | |
| 14:19:33 | bauzas | for example, I could imagine some way to map between Nova and Cinder AZs | |
| 14:19:44 | bauzas | one Nova AZ could map to say 3 Cinder AZs | |
| 14:19:44 | bauzas | one Nova AZ could map to say 3 Cinder AZs | |
| 14:20:10 | bauzas | and then you could write a query saying "I want my instance to use Cinder AZs" | |
| 14:20:10 | bauzas | and then you could write a query saying "I want my instance to use Cinder AZs" | |
| 14:20:34 | bauzas | if so, when resizing, we could continue to look at placement for knowing the existing AZ | |
| 14:20:34 | bauzas | if so, when resizing, we could continue to look at placement for knowing the existing AZ | |
| 14:20:39 | bauzas | AZs even | |
| 14:20:39 | bauzas | AZs even | |
| 14:22:36 | sean-k-mooney | bauzas: im not sure that is true | |
| 14:22:36 | sean-k-mooney | bauzas: im not sure that is true | |
| 14:22:54 | sean-k-mooney | i dont think one nova AZ can map to multiplle cinder AZs | |
| 14:22:54 | sean-k-mooney | i dont think one nova AZ can map to multiplle cinder AZs | |
| 14:23:15 | sean-k-mooney | i think all our assumtion for schduling is tthat there is a 1:1 mapping | |
| 14:23:15 | sean-k-mooney | i think all our assumtion for schduling is tthat there is a 1:1 mapping | |
| 14:23:25 | bauzas | sean-k-mooney: I'm not telling about the existing behaviour | |
| 14:23:25 | bauzas | sean-k-mooney: I'm not telling about the existing behaviour | |
| 14:23:29 | sean-k-mooney | espcailly when it comes to placmenet and azs | |
| 14:23:30 | sean-k-mooney | espcailly when it comes to placmenet and azs | |
| 14:23:48 | bauzas | but about what we *could* do to support related AZs between Nova and Cinder | |
| 14:23:48 | bauzas | but about what we *could* do to support related AZs between Nova and Cinder | |
| 14:23:58 | sean-k-mooney | bauzas: ok but i dont think it would be desireable to support that | |
| 14:23:58 | sean-k-mooney | bauzas: ok but i dont think it would be desireable to support that | |
| 14:24:04 | bauzas | but, cross-project spec, that's it | |
| 14:24:04 | bauzas | but, cross-project spec, that's it | |
| 14:24:13 | bauzas | sean-k-mooney: sure, maybe a nit | |
| 14:24:13 | bauzas | sean-k-mooney: sure, maybe a nit | |
| 14:24:27 | bauzas | sean-k-mooney: but my point is, nova and cinder AZs are respectively apples and bananes | |
| 14:24:27 | bauzas | sean-k-mooney: but my point is, nova and cinder AZs are respectively apples and bananes | |
| 14:24:28 | sean-k-mooney | well no let me rephrase i dont think we shoudl do that | |
| 14:24:29 | sean-k-mooney | well no let me rephrase i dont think we shoudl do that | |
| 14:24:46 | sean-k-mooney | we could supprot placement aggreate in the service | |
| 14:24:46 | sean-k-mooney | we could supprot placement aggreate in the service | |
| 14:24:47 | bauzas | sean-k-mooney: so, you could have apples that match multiple bananes | |
| 14:24:47 | bauzas | sean-k-mooney: so, you could have apples that match multiple bananes | |
| 14:24:53 | bauzas | bananas* even | |
| 14:24:53 | bauzas | bananas* even | |
| 14:25:15 | bauzas | sean-k-mooney: surely, we need to have cinder using placement AZs | |
| 14:25:15 | bauzas | sean-k-mooney: surely, we need to have cinder using placement AZs | |
| 14:25:21 | bauzas | uhu | |
| 14:25:21 | bauzas | uhu | |
| 14:25:26 | bauzas | placement s/AZs/aggregates | |
| 14:25:26 | bauzas | placement s/AZs/aggregates | |
| 14:25:37 | sean-k-mooney | cinder does not do that currently | |
| 14:25:37 | sean-k-mooney | cinder does not do that currently | |
| 14:25:45 | bauzas | again, that's a x-p spec | |
| 14:25:45 | sean-k-mooney | but i think that would be the better approch | |
| 14:25:45 | bauzas | again, that's a x-p spec | |
| 14:25:45 | sean-k-mooney | but i think that would be the better approch | |
| 14:25:53 | sean-k-mooney | yep i know | |
| 14:25:53 | sean-k-mooney | yep i know | |
| 14:25:59 | bauzas | and yeah, I know | |
| 14:25:59 | bauzas | and yeah, I know | |
| 14:26:05 | bauzas | that cinder doesn't do it | |
| 14:26:05 | bauzas | that cinder doesn't do it | |
| 14:26:34 | bauzas | honestly, this is a known bug since... mmm, Liberty ? | |
| 14:26:34 | bauzas | honestly, this is a known bug since... mmm, Liberty ? | |
| 14:26:49 | sean-k-mooney | eventhough shareing resouce proviers was invented for them hehe | |
| 14:26:49 | sean-k-mooney | eventhough shareing resouce proviers was invented for them hehe | |
| 14:26:57 | bauzas | :) | |
| 14:26:57 | bauzas | :) | |
| 14:27:03 | sean-k-mooney | well its not a bug really | |
| 14:27:03 | sean-k-mooney | well its not a bug really | |
| 14:27:07 | bauzas | correct | |
| 14:27:07 | bauzas | correct | |
| 14:27:15 | bauzas | it's a non supported feature :) | |
| 14:27:15 | bauzas | it's a non supported feature :) | |
| 14:27:30 | sean-k-mooney | to disable cross az attach you have to align the cinder an nova az and have to froce a default az | |
| 14:27:30 | tobias-urdin | :D i'm not mad, i agree | |
| 14:27:30 | sean-k-mooney | to disable cross az attach you have to align the cinder an nova az and have to froce a default az | |
| 14:27:30 | tobias-urdin | :D i'm not mad, i agree | |