| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2020-10-15 | |||
| 17:18:01 | melwitt | ok, so this issue is NOT to do with the whitelist at all. I see | |
| 17:18:29 | melwitt | so the whitelist parsing code is reused to parse the devices returned by libvirt | |
| 17:18:59 | melwitt | ok, I was missing that | |
| 17:19:17 | sean-k-mooney | so here https://github.com/openstack/nova/blob/master/nova/pci/manager.py#L118-L122 | |
| 17:20:03 | sean-k-mooney | yep we reuse https://github.com/openstack/nova/blob/2c3ff1004e8812da0bc4e5502762e6b9a7650846/nova/pci/whitelist.py#L80-L88 | |
| 17:20:05 | melwitt | ok, this way we're sure we're not parsing a whitelist | |
| 17:20:54 | sean-k-mooney | so what we do is we build a device spec object for every entry in the whitelist | |
| 17:21:16 | sean-k-mooney | then for every device returned by the vert driver we constuct a second one and see if tey match | |
| 17:21:28 | sean-k-mooney | its the libvirt address that is 32bit | |
| 17:22:16 | melwitt | but the exception guessed raised in __init__ not in the match() call? | |
| 17:22:30 | melwitt | *is | |
| 17:23:07 | sean-k-mooney | from here https://github.com/openstack/nova/blob/2c3ff1004e8812da0bc4e5502762e6b9a7650846/nova/pci/devspec.py#L217-L238 | |
| 17:23:28 | melwitt | ohhhh | |
| 17:23:30 | melwitt | guh | |
| 17:23:43 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: Fix the vGPU dynamic options race https://review.opendev.org/758470 | |
| 17:24:44 | melwitt | so devices_json is a list of whitelist pci objects | |
| 17:26:01 | melwitt | maybe not... I'll look. this is so confusing to me | |
| 17:26:34 | sean-k-mooney | its a json blob we construct for the data we get from libvirt | |
| 17:26:43 | sean-k-mooney | the nodedev xmls | |
| 17:28:02 | sean-k-mooney | it looks like a list of these https://github.com/openstack/nova/blob/131f7606c179bd08bde3b7ae0e6bb0b59acf1545/nova/tests/unit/pci/test_manager.py#L32-L41 | |
| 17:28:07 | melwitt | yeah, just that device_assignable is from the Whitelist class | |
| 17:29:40 | sean-k-mooney | i have to look this up everythim but it generated here https://opendev.org/openstack/nova/src/branch/master/nova/virt/libvirt/driver.py#L7127-L7150 | |
| 17:30:42 | sean-k-mooney | i rewrote this recently | |
| 17:31:02 | sean-k-mooney | but we get all the nodedev form libvirt that are pci device ro nics | |
| 17:31:08 | melwitt | thanks | |
| 17:31:55 | melwitt | oh it's the dev_filter that's the whitelist, facepalm. sorry | |
| 17:32:20 | melwitt | I have gone through this before too, but still mess it up | |
| 17:33:07 | sean-k-mooney | dev_filter is an instance of the Whitelist class yes | |
| 17:33:47 | sean-k-mooney | whic has one filed specs which holds all the whitelist entries | |
| 17:34:02 | sean-k-mooney | then we just do | |
| 17:34:04 | sean-k-mooney | for spec in self.specs: | |
| 17:34:06 | sean-k-mooney | if spec.match(dev): | |
| 17:34:08 | sean-k-mooney | return True | |
| 17:34:10 | sean-k-mooney | return False | |
| 17:34:33 | sean-k-mooney | so if any of the entries match the json seriasition of the pci device form the virt driver w return true | |
| 17:35:01 | melwitt | ok | |
| 17:35:39 | sean-k-mooney | that mach is the on defiend in WhitelistPciAddress | |
| 17:36:09 | sean-k-mooney | which does https://github.com/openstack/nova/blob/2c3ff1004e8812da0bc4e5502762e6b9a7650846/nova/pci/devspec.py#L238 | |
| 17:37:07 | sean-k-mooney | whic eventually expodes | |
| 17:37:51 | sean-k-mooney | it actully expoding here https://github.com/openstack/nova/blob/2c3ff1004e8812da0bc4e5502762e6b9a7650846/nova/pci/devspec.py#L237 | |
| 17:38:20 | sean-k-mooney | when we contuct the dev object to compare agaisnt form teh adress we get form libvirt | |
| 17:38:55 | sean-k-mooney | that why its expoliding in init | |
| 17:39:21 | sean-k-mooney | we constuct the object form the string to reuse the parsing that was written for teh whitelist | |
| 17:39:30 | sean-k-mooney | so we dont have to have it twice | |
| 17:40:13 | melwitt | is pci_phys_addr the whitelist entry? | |
| 17:41:24 | sean-k-mooney | no that is the pci adress form libvirt | |
| 17:41:39 | sean-k-mooney | self.pci_address_spec is the object form the whitelist | |
| 17:41:47 | sean-k-mooney | arent the names great | |
| 17:41:57 | sean-k-mooney | so easy to follow and not confusing at all | |
| 17:42:04 | melwitt | /o\ | |
| 17:43:01 | sean-k-mooney | melwitt: its not taking this branch https://github.com/openstack/nova/blob/2c3ff1004e8812da0bc4e5502762e6b9a7650846/nova/pci/devspec.py#L231-L234 | |
| 17:43:10 | sean-k-mooney | but it would explode if it did | |
| 17:43:31 | melwitt | well, what I mean is, IIUC (which I probably don't) one of those construct of PhysicalPciAddress is "correctly" parsing a whitelist entry and the other one is parsing an address from libvirt. is my understanding wrong? | |
| 17:43:34 | sean-k-mooney | that is for pci device that are capable fo sriov | |
| 17:43:46 | melwitt | oh, ok. nevermind my question then | |
| 17:44:09 | sean-k-mooney | for PF if you list a PF we allow all the VF if they are allocated to be used | |
| 17:44:15 | sean-k-mooney | that is what that branch is for | |
| 17:44:38 | sean-k-mooney | so if whitelist a PF its a shortcut to whitelist the VFs | |
| 17:44:55 | melwitt | that makes sense | |
| 17:46:59 | sean-k-mooney | anyway the tl;dr for gibi is we shoudl catch the excpetion here https://github.com/openstack/nova/blob/2c3ff1004e8812da0bc4e5502762e6b9a7650846/nova/pci/whitelist.py#L85-L87 instead | |
| 17:47:18 | sean-k-mooney | to that if should be in a try | |
| 17:48:00 | sean-k-mooney | and we retrun fause if we get the PciConfigInvalidWhitelist excption | |
| 17:48:15 | sean-k-mooney | whcih is what he is doing here https://review.opendev.org/#/c/756697/4/nova/pci/manager.py | |
| 17:48:54 | sean-k-mooney | back to your orginal question | |
| 17:48:57 | sean-k-mooney | https://review.opendev.org/#/c/756697/4/nova/pci/devspec.py | |
| 17:49:18 | sean-k-mooney | is still going to raise and excption if you input an invalid whitelist | |
| 17:49:36 | sean-k-mooney | so i think the compute agent will still fail to start in that case | |
| 17:50:13 | sean-k-mooney | but we support 3 different type of adresses. | |
| 17:51:08 | sean-k-mooney | gibi only exeteded this one since its the one we use for the virt driver objcts | |
| 17:52:28 | sean-k-mooney | since gibi is not catching the excption it will still fail here | |
| 17:52:30 | sean-k-mooney | https://github.com/openstack/nova/blob/2c3ff1004e8812da0bc4e5502762e6b9a7650846/nova/pci/whitelist.py#L52-L79 | |
| 17:52:54 | sean-k-mooney | when we are parsing the whitelist if it had a 32bit domain | |
| 17:53:25 | sean-k-mooney | this will raise the new excption | |
| 17:53:47 | sean-k-mooney | spec = devspec.PciDeviceSpec(ds) | |
| 17:53:53 | sean-k-mooney | line 75 | |
| 18:21:27 | ganso | hi folks! If anyone could please spare a few minutes to look at this stable/train backport, should be really simple and quick as the other ones: https://review.opendev.org/#/c/757870/ | |
| 18:22:30 | ganso | melwitt, lyarwood, elod ^ | |
| 18:24:12 | melwitt | ganso: ack will look later today | |
| 18:24:22 | ganso | melwitt: thank you! | |
| 20:12:33 | elod | ganso: the backport looks good to me, just as it was ok to ussuri. +2 | |
| 20:12:50 | ganso | elod: thanks! | |
| #openstack-nova - 2020-10-16 | |||
| 08:00:14 | yingjisun | gibi: are you here? | |
| 08:08:29 | gibi | yingjisun: hi! | |
| 08:14:15 | yingjisun | gibi: I have a question about updating nova db. I would like to change the length of internal_access_path in console_auth_tokens, starting from Train. I want to modify nova/db/sqlalchemy/models.py and the migration file under nova/db/sqlalchemy/migrate_repo/versions. However I did not find some placeholder in Train. 402_add_resources.py is the last one in Train and 403_ is "Add reserved schema migrations for Ussuri". So how can I do it in Train ? | |
| 08:15:03 | yingjisun | Here is the repo link. https://github.com/openstack/nova/tree/master/nova/db/sqlalchemy/migrate_repo/versions | |
| 08:19:19 | gibi | yingjisun: good question, let me think about it | |
| 08:24:26 | yingjisun | Thank you very much. If it is not a question at hand, I can ask it in openstack-discuss alias. I do not want you to spend much time on this. :) | |
| 08:25:46 | gibi | yingjisun: I think you can use 403 for the backport to Train. As 402 is train and and no newer migration merged yet | |
| 08:26:56 | yingjisun | ok. Then I will open a bug first and invovle more people to review it. Thank you. | |
| 08:27:05 | gibi | but it would be nice to ask dansmith about this ^^ I've never done db migration backports | |
| 08:27:24 | yingjisun | sure. I will. | |
| 08:27:35 | gibi | thnks | |
| 08:27:38 | gibi | thanks | |
| 08:28:46 | gibi | link me the bug when you open it | |
| 08:28:57 | yingjisun | ok. | |
| 09:11:40 | gibi | dansmith: could you please update the topic of the channel with the wallaby runways etherpad link? https://etherpad.opendev.org/p/nova-runways-wallaby ? | |
| 10:17:26 | frickler | gibi: ^^ like this? | |
| 10:22:36 | gibi | frickler: thanks | |
| 10:22:44 | gibi | is there a way I can do it next time? | |
| 10:26:12 | openstackgerrit | Andrew Bonney proposed openstack/nova master: Handle disabled CPU features to fix live migration failures https://review.opendev.org/757577 | |