| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-08-23 | |||
| 12:13:36 | mriedem | sahid: you should take a look at moshele's patch | |
| 12:13:46 | mriedem | moshele: what sets the driver_name to vhost? | |
| 12:14:54 | moshele | mriedem: the macvtap port which is virtio tap connected to VF so for that case it should be fine | |
| 12:14:54 | sahid | mriedem ah yes this bug has been reported downstream yesterday i think | |
| 12:15:00 | sahid | moshele: thanks for your patch | |
| 12:15:18 | moshele | sahid: np :) | |
| 12:15:44 | mriedem | moshele: sorry i don't understand that statement, | |
| 12:15:46 | mriedem | looking at your paste | |
| 12:15:46 | mriedem | : libvirtError: unsupported configuration: Unknown PCI device <driver name='vhost'/> has been specified | |
| 12:15:57 | mriedem | and looking at https://review.openstack.org/#/c/595592/1/nova/virt/libvirt/designer.py | |
| 12:16:06 | mriedem | i guess in this case the net_type was 'direct'? | |
| 12:16:11 | mriedem | so L120 | |
| 12:17:17 | sahid | mriedem: are you sure vfio is a valid driver? | |
| 12:17:22 | sahid | moshele: | |
| 12:17:55 | ykarel | jroll, hi | |
| 12:17:56 | moshele | shaid: https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/7/html/virtualization_deployment_and_administration_guide/sect-pci_devices-pci_passthrough | |
| 12:18:19 | moshele | sahid: https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/7/html/virtualization_deployment_and_administration_guide/sect-pci_devices-pci_passthrough | |
| 12:19:12 | ykarel | jroll, can u check https://bugs.launchpad.net/tripleo/+bug/1787910/comments/13 | |
| 12:19:12 | openstack | Launchpad bug 1787910 in tripleo "OVB overcloud deploy fails on nova placement errors" [Critical,Triaged] - Assigned to Marios Andreou (marios-b) | |
| 12:19:16 | sahid | ok thanks, because i remeber of a bug in libvirt for vhostuser interface interface, when you basically set it to vhostuser (which is the driver used) libvirt is reaisng an error | |
| 12:19:18 | moshele | mriedem: no the net_type 'direct' is for macvtap port for port direct which is hostdev we don't set the driver name | |
| 12:20:04 | mriedem | moshele: yeah i figured it out, | |
| 12:20:09 | mriedem | the comment is busted in the code, | |
| 12:20:13 | mriedem | i left a comment about fixing it :) | |
| 12:21:16 | moshele | mriedem: the problem is with this https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/virt/libvirt/vif.py#L166-L176 | |
| 12:21:40 | moshele | mriedem: driver is not specified it put vhost | |
| 12:21:45 | ykarel | mriedem, or u can check https://bugs.launchpad.net/tripleo/+bug/1787910/comments/13 | |
| 12:21:45 | openstack | Launchpad bug 1787910 in tripleo "OVB overcloud deploy fails on nova placement errors" [Critical,Triaged] - Assigned to Marios Andreou (marios-b) | |
| 12:22:24 | mriedem | aha | |
| 12:22:26 | mriedem | moshele: ok | |
| 12:23:04 | moshele | mriedem: I send the mail because I am not sure about the fix. I know that my fix solve direct port | |
| 12:23:57 | mriedem | i'd ask jangutter but he's not around | |
| 12:24:23 | sahid | moshele: perhaps it's better to use None, lbvirt might have some kind of fallback, example using vfio but if available on host fallback to uio | |
| 12:24:24 | mriedem | moshele: vnic_type = 'direct' right? | |
| 12:24:31 | zigo | Hi there ! | |
| 12:24:45 | zigo | Any idea why I'm getting this when building the Nova Debian package? https://stretch-rocky.infomaniak.ch/job/nova/9/console | |
| 12:24:50 | zigo | (see the end of unit test run...) | |
| 12:25:03 | sahid | it's what is done for vhost, norammy by default, libvirt use vhost but fallback to qemu if vhost module is not loaded on host | |
| 12:25:06 | mriedem | zigo: you'll need to put that in a paste | |
| 12:25:06 | moshele | mriedem: yes | |
| 12:25:09 | mriedem | or give out creds | |
| 12:25:30 | zigo | Sure. | |
| 12:25:55 | zigo | mriedem: http://paste.openstack.org/show/728679/ | |
| 12:26:50 | zigo | From where is comming the MIMEAccept object? Is that nova's internal? | |
| 12:27:01 | zigo | A grep in nova shows nothing ... | |
| 12:27:03 | moshele | sahid: but this port will set driver name to vhost if it None https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/virt/libvirt/vif.py#L166-L176 | |
| 12:27:17 | mriedem | ykarel: i'll let jroll look at that b/c https://review.openstack.org/#/c/565841 was master only atm which is stein, and today is rocky rc3 so i'm focused on what we need to get done for rc3 | |
| 12:27:21 | moshele | sahid: s/port/part | |
| 12:27:36 | ykarel | mriedem, no prob, Thanks | |
| 12:27:47 | ykarel | rocky is more priority now | |
| 12:28:00 | mriedem | zigo: i'd guess the wsgi-intercept library | |
| 12:28:02 | mriedem | cdent: ^ | |
| 12:28:05 | mriedem | http://paste.openstack.org/show/728679/ | |
| 12:29:14 | mriedem | well best_matches = req.accept.acceptable_offers(types) is clearly in placement code | |
| 12:29:21 | cdent | zigo, looks webob | |
| 12:29:30 | sahid | moshele: hum... it's complicated, because if you set it to None, the LibvirtConfig will not generate the tx/rx args | |
| 12:29:30 | mriedem | yeah | |
| 12:29:42 | cdent | acceptable_offer is new in webob 1.8.x I think | |
| 12:30:02 | mriedem | sahid: but it looks like we default to vhost if rx/tx is configured https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/virt/libvirt/vif.py#L166-L176 | |
| 12:30:10 | mriedem | so driver_name will always be at least vhost? | |
| 12:30:25 | mriedem | which doesn't work for net_type == 'hostdev'? | |
| 12:30:42 | cdent | zigo: my guess is that the webob is not up to date with minimum requirements? | |
| 12:30:53 | zigo | cdent: Yeah, that should be it, thanks! | |
| 12:31:05 | mriedem | https://github.com/openstack/nova/blob/master/lower-constraints.txt#L170 fwiw | |
| 12:31:08 | mriedem | 1.8.2 | |
| 12:31:09 | zigo | Debian version is 1:1.7.3-2 | |
| 12:31:11 | cdent | zigo: I'm glad it was that easy :) | |
| 12:31:12 | zigo | Thanks guys. | |
| 12:31:14 | zigo | Will update it. | |
| 12:31:22 | sahid | yes yes, it's just i don't think we should specically set it to vfio, i think we should not set it at all for hostdev | |
| 12:31:40 | sahid | but if we do that tx/rx are then ignored | |
| 12:31:54 | jroll | mriedem: ykarel: I'm looking at this, I'm confused what is special about tripleo jobs here, as this works fine in our ironic jobs | |
| 12:32:08 | sahid | https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/virt/libvirt/config.py#L1390 | |
| 12:32:17 | jroll | oh wait, it's probably the flavors | |
| 12:32:24 | mriedem | sahid: yup | |
| 12:33:13 | mriedem | but if rx is configured, | |
| 12:33:15 | mriedem | and driver is None, | |
| 12:33:20 | mriedem | we hard-code to vhost | |
| 12:33:21 | mriedem | https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/virt/libvirt/vif.py#L169 | |
| 12:33:39 | mriedem | but apparently that doesn't work | |
| 12:33:39 | mriedem | : libvirtError: unsupported configuration: Unknown PCI device <driver name='vhost'/> has been specified | |
| 12:33:45 | mriedem | http://paste.openstack.org/show/728666/ | |
| 12:34:14 | sahid | can't we do something like that? https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/virt/libvirt/vif.py#L432 | |
| 12:34:53 | mriedem | heh, that TODO is overwriting the thing done in get_base_config which is based on another TODO | |
| 12:35:28 | mriedem | and here https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/virt/libvirt/vif.py#L519 | |
| 12:35:42 | moshele | sahid.mriedem : we can skip it for vnic_type direct in https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/virt/libvirt/vif.py#L166-L176 | |
| 12:35:48 | mriedem | sahid: which vnic_type did you test the rx queue stuff with? | |
| 12:36:04 | jroll | ykarel: posted a comment on the bug | |
| 12:36:07 | moshele | sahid, mriedem ^ | |
| 12:36:13 | ykarel | jroll, checking | |
| 12:36:27 | mriedem | moshele: but what about the other sriov types? https://github.com/openstack/nova/blob/34956bea4beb8e5ba474b42ba777eb88a5eadd76/nova/network/model.py#L111 | |
| 12:36:35 | sahid | get_config_vhostuser and _set_config_VIFHostUser are the same thing, one is native Nova the other for os-vif | |
| 12:36:45 | mriedem | i know | |
| 12:37:01 | sahid | ah sorry i did not have undersdoot the qustion so | |
| 12:37:11 | mriedem | so are direct and macvtap the only supported vnic types for rx/tx queues? | |
| 12:37:29 | moshele | mriedem: they are got beacuse the macvtap is also virtio and also the virtio forwarder | |
| 12:37:40 | moshele | mriedem: s/got/good | |
| 12:37:46 | sahid | all vnic that are using virtio | |
| 12:37:56 | sahid | can have rc/tx queues | |
| 12:38:12 | moshele | sahid: SR-IOV passthrough is not virtio | |