| hemanth | Hi nova team, appreciate for discussing compute memory monitor during weekly meeting (could not attend due to TZ diff) and thanks sean for detail explanation on the PR! | 01:32 |
|---|---|---|
| opendevreview | zhou zhong proposed openstack/nova master: Add auto_converge parameters for invoking the migration https://review.opendev.org/c/openstack/nova/+/935614 | 08:25 |
| mikal | sean-k-mooney: I did some reading today, and I am left with the distinct impression that we should only add XHCI USB controllers for now. This makes it simpler to predict how many ports are available, but the performance of the older controllers is also not great. So much like we don't support floppies, we should probably not support USB controllers from the 1990s. | 08:34 |
| mikal | sean-k-mooney: that said, I think we should probably support both qemu-xhci and nec-xhci, because I suspect (without testing) that Windows wont know about the qemu version without additional drivers. | 08:35 |
| mikal | sean-k-mooney: it would be trivial to add other controllers back if someone could make a case for them later, but we should probably remove them before the USB redir extra spec lands. | 08:35 |
| sean-k-mooney | ya so the diffent types of port in the list i proved were basicly usb 1 2 and 3 | 11:48 |
| sean-k-mooney | xhci being usb 3 | 11:49 |
| sean-k-mooney | i think that makes sense as a baseline | 11:49 |
| opendevreview | Dan Smith proposed openstack/nova master: DNM: Test with latest oslo.utils https://review.opendev.org/c/openstack/nova/+/933444 | 15:44 |
| opendevreview | Dan Smith proposed openstack/nova master: WIP Inspect snapshot images and adjust disk_format https://review.opendev.org/c/openstack/nova/+/939815 | 15:44 |
| opendevreview | Balazs Gibizer proposed openstack/nova master: [CI][nova-next]test with placement ac breadth-first https://review.opendev.org/c/openstack/nova/+/937275 | 16:08 |
| opendevreview | James Page proposed openstack/nova master: Restore nova.virt.__init__ https://review.opendev.org/c/openstack/nova/+/939827 | 16:49 |
| mikal | sean-k-mooney: so, mostly. If I am reading this stuff right, OHCI / UHCI are USB 1.1. If you want USB 2.0 you need both EHCI _and_ one of OHCI / UCHI as a "companion controller". None of these controllers are particularly virt (or 64 bit!) friendly. XHCI supports USB 1.1 / 2 / 3 without a companion controller, while also being 64 bit and virt friendly. | 18:27 |
| mikal | sean-k-mooney: XHCI 1.0 was released in 2010, so I would hope that basically anything people would want to run would support it. | 18:28 |
| opendevreview | Merged openstack/nova master: Run nova-next without periodic cache healing https://review.opendev.org/c/openstack/nova/+/939453 | 18:30 |
| sean-k-mooney | mikal: yep it should | 18:33 |
| sean-k-mooney | its a good baseline if nothing else | 18:33 |
| mikal | Honestly, adding the others back would be relatively trivial (less than ten lines of code), so if you think that's reasonable I'll tweak that PR to just do XHCI and we can limit the number of redirected channels to the number of root ports on a XHCI controller. | 18:36 |
| mikal | I would like to test to see if qemu-xhci works for me because for reasons I no longer remember I used nec-xhci in my testing last time. | 18:37 |
| sean-k-mooney | ack sounds like a plan | 18:38 |
| mikal | Cool. I personally don't think any of that should hold up reviews of the earlier patches, I see this as a relatively minor tweak to the final patch in the series. | 18:40 |
| mikal | While I think of it, for reference oVirt appears to use a qemu-xhci controller with 8 ports. So that should be a reasonably safe choice. | 19:46 |
| sean-k-mooney | mikal: i run my windows gaming pc in a vm | 19:56 |
| sean-k-mooney | mikal: and it has | 19:56 |
| sean-k-mooney | <controller type="usb" index="0" model="qemu-xhci" ports="15"> | 19:56 |
| sean-k-mooney | <address type="pci" domain="0x0000" bus="0x02" slot="0x00" function="0x0"/> | 19:56 |
| sean-k-mooney | </controller> | 19:56 |
| mikal | Yeah, there is some ovirt code I'm reading which seems to change the controller type based on guest OS, but if qemu-xhci works for Linux and modern Windows I think we're good to go. | 19:57 |
| sean-k-mooney | i may have installed virtio drivers for it but that was the default model that virt-manager selected | 19:57 |
| sean-k-mooney | it worked fine when i was passhtihg though a phsical mouse and keyboard to the vm | 19:59 |
| sean-k-mooney | again i dont recall if installed a dirver but i dont think i had to | 19:59 |
| mikal | I think https://github.com/oVirt/ovirt-engine/blob/master/backend/manager/modules/utils/src/test/java/org/ovirt/engine/core/utils/OsRepositoryImplTest.java#L79 is effectively saying that RHEL 6 and 7 need nec-xhci but maybe nothing else? I still need to find the code which does that thing, instead of the test which tests it. | 20:00 |
| mikal | But Java is not a thing I've read in quite a while now... | 20:00 |
| mikal | I need to wander off for now, but I will keep poking at this and let you know if I find anything interesting. | 20:00 |
| -opendevstatus- NOTICE: The Gerrit service on review.opendev.org will be offline momentarily for a restart to put some database compaction config changes into effect, and will return within a few minutes | 22:54 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!