Hello Luca, The osn provider basically only connects the vnics to the osn networks, items like dhcp must be handled on the osn networks itself. Have you tried connecting any nics manually and checking connectivity between them? No connectivity with static IP's could hint at some configuration problems. Are osn/ovs set up correctly? Firewall blocking traffic? Can you please send us a more detailed descirption of your env (ip addr, brctl show, ovs-vsctl show)? Thanks, Marcin

Hello Luca I assume Networking Plugin: Open vSwitch Do you have any security groups defined for the vnic profile on your external network? Looking at the output you provided is seems you do.The qpb bridge and the qvb/qvo veth pair are created when the security groups are present. Can you try without the security groups? This would connect your vm vnic right into the ovs integration bridge (br-int). Yes, this is what I needed. This is quite worrying. Is ovs on the host working properly? Can you please check: service openvswitch status If not active: service openvswitch start Add 2 ports in openstack directly Add 2 nics on vm's, specifying the libvirt xml for the interfaces: <interface type="bridge"> <model type="virtio"/> <source bridge="br-int"/> <virtualport type="openvswitch"> <parameters interfaceid="<neturon port id>"/> </virtualport> </interface> If the openstack neutron plugin works correctly, the ports should be connected to the osn network. On Thu, Mar 9, 2017 at 4:26 PM, Luca 'remix_tj' Lorenzetto < lorenzetto.luca(a)gmail.com&gt; wrote:

On Fri, Mar 10, 2017 at 12:11 PM, Luca 'remix_tj' Lorenzetto <lorenzetto.luca(a)gmail.com&gt; wrote: Hello, i started ovs on the node where vm is running, but now i see only this: [root@fltiovirt002 ~]# ovs-vsctl show 53bfa8a8-80c5-4372-abb1-633c29eb1e60 ovs_version: "2.6.90 even restarting the vm on the same host changes nothing. I see instead q* interfaces appearing and disappearing from ip a s command output. -- "E' assurdo impiegare gli uomini di intelligenza eccellente per fare calcoli che potrebbero essere affidati a chiunque se si usassero delle macchine" Gottfried Wilhelm von Leibnitz, Filosofo e Matematico (1646-1716) "Internet è la più grande biblioteca del mondo. Ma il problema è che i libri sono tutti sparsi sul pavimento" John Allen Paulos, Matematico (1945-vivente) Luca 'remix_tj' Lorenzetto, http://www.remixtj.net , <lorenzetto.luca(a)gmail.com&gt;

defined through horizon without any special option. This is strange, as the basic flow should only connect the vnic to the ovs integration bridge. Please give me some time to look at the code. This is in the libvirt vm definition. It can be edited using the "virsh" tools: virsh edit <vm name> You could also attempt to check this by attaching some an interface (like a veth pair) directly to the ovs bridge and setting the ovs interfaceid parameter to mark is as osn port. Let me know if this is clear, if not I will try to decribe the procedure in more detail. On Fri, Mar 10, 2017 at 12:11 PM, Luca 'remix_tj' Lorenzetto < lorenzetto.luca(a)gmail.com&gt; wrote: > > > Il 10 mar 2017 11:48 AM, "Marcin Mirecki" <mmirecki(a)redhat.com&gt; ha > scritto: > > Hello Luca > > >Name: openstack-networks > >Type: OpenStack Networking > >Description: > >Provider URL: http://openstack.example.com:9696 > > I assume Networking Plugin: Open vSwitch > > > Yes, i confirm. > > > > Do you have any security groups defined for the vnic profile on your > external network? > Looking at the output you provided is seems you do.The qpb bridge and the > qvb/qvo veth pair are created when the security groups are present. > Can you try without the security groups? This would connect your vm vnic > right into the ovs integration bridge (br-int). > > > defined through horizon without any special option. > > > >This are the output of the commands you asked from a node where a vm > >that is attached to a neutron network is running: > Yes, this is what I needed. > > >[root@ovirt002 ~]# ovs-vsctl show > >ovs-vsctl: unix:/var/run/openvswitch/db.sock: database connection > >failed (No such file or directory) > > This is quite worrying. Is ovs on the host working properly? > Can you please check: service openvswitch status > If not active: service openvswitch start > > > Who should enable it, a cluster confoguration or whatl else? I'm using > ovirt-node-ng, i suppose that openvswitch is already installed (didn't > check). > > > > > >> Have you tried connecting any nics manually and checking connectivity > >> between them? > >What do you mean? > > Add 2 ports in openstack directly > Add 2 nics on vm's, specifying the libvirt xml for the interfaces: > > <interface type="bridge"> > <model type="virtio"/> > <source bridge="br-int"/> > <virtualport type="openvswitch"> > <parameters interfaceid="<neturon port id>"/> > </virtualport> > </interface> > > > > > > If the openstack neutron plugin works correctly, the ports should be > connected to the osn network. > > > > > > > On Thu, Mar 9, 2017 at 4:26 PM, Luca 'remix_tj' Lorenzetto < > lorenzetto.luca(a)gmail.com&gt; wrote: > >> On Thu, Mar 9, 2017 at 2:24 PM, Marcin Mirecki <mmirecki(a)redhat.com&gt; >> wrote: >> > Hello Luca, >> >> Hello Marcin, >> >> > The osn provider basically only connects the vnics to the osn networks, >> > items like dhcp must be handled on the osn networks itself. >> >> Yes, i know. The network is defined by neutron, which has it's own dhcp >> server. >> >> > Have you tried connecting any nics manually and checking connectivity >> > between them? >> >> What do you mean? >> >> > No connectivity with static IP's could hint at some configuration >> problems. >> > Are osn/ovs set up correctly? Firewall blocking traffic? >> >> there is no firewall between openstack controllers and ovirt >> engine/hosts. My doubt is about configuration, i've configured in this >> way: >> >> Name: openstack-networks >> Type: OpenStack Networking >> Description: >> Provider URL: http://openstack.example.com:9696 >> >> Flagged read-only and requires authentication >> >> Set the authentication and tested, reports everything ok. >> >> Nothing else has been configured. I didn't found any documentation >> that clarified if is enough. >> >> After powering on i see on openstack this: >> >> [stack@opstrio1101 ~]$ openstack port list | grep 00:1a:4a:16:01:51 >> <-- this is mac address of oVirt VM >> | 86c46fed-dddf-4776-a765-27d4e52e861c | nic1 >> | 00:1a:4a:16:01:51 | ip_address='172.25.7.4', >> subnet_id='280a98ad-0fd5-4961-a307-d1bfea8355cd' | >> >> >> >> > Can you please send us a more detailed descirption of your env (ip addr, >> > brctl show, ovs-vsctl show)? >> >> This are the output of the commands you asked from a node where a vm >> that is attached to a neutron network is running: >> >> [root@ovirt002 ~]# ip addr >> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1 >> link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 >> inet 127.0.0.1/8 scope host lo >> valid_lft forever preferred_lft forever >> inet6 ::1/128 scope host >> valid_lft forever preferred_lft forever >> 2: enp2s0f0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq >> master bond0 state UP qlen 1000 >> link/ether 00:21:5a:9b:b7:93 brd ff:ff:ff:ff:ff:ff >> 3: enp2s0f1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq >> master bond0 state UP qlen 1000 >> link/ether 00:21:5a:9b:b7:93 brd ff:ff:ff:ff:ff:ff >> 4: enp2s0f2: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq >> master bond1 state UP qlen 1000 >> link/ether 00:21:5a:9b:b7:97 brd ff:ff:ff:ff:ff:ff >> 5: enp2s0f3: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq >> master bond1 state UP qlen 1000 >> link/ether 00:21:5a:9b:b7:97 brd ff:ff:ff:ff:ff:ff >> 6: enp2s0f4: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq >> state DOWN qlen 1000 >> link/ether 00:21:5a:9b:b7:9b brd ff:ff:ff:ff:ff:ff >> 7: enp2s0f5: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq >> state DOWN qlen 1000 >> link/ether 00:21:5a:9b:b7:9d brd ff:ff:ff:ff:ff:ff >> 8: enp2s0f6: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq >> state DOWN qlen 1000 >> link/ether b4:b5:2f:55:bc:eb brd ff:ff:ff:ff:ff:ff >> 9: enp2s0f7: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq >> state DOWN qlen 1000 >> link/ether b4:b5:2f:55:bc:ef brd ff:ff:ff:ff:ff:ff >> 10: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc >> noqueue master ovirtmgmt state UP qlen 1000 >> link/ether 00:21:5a:9b:b7:93 brd ff:ff:ff:ff:ff:ff >> 11: ovirtmgmt: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc >> noqueue state UP qlen 1000 >> link/ether 00:21:5a:9b:b7:93 brd ff:ff:ff:ff:ff:ff >> inet 10.5.40.192/22 brd 10.5.43.255 scope global ovirtmgmt >> valid_lft forever preferred_lft forever >> inet6 fe80::221:5aff:fe9b:b793/64 scope link >> valid_lft forever preferred_lft forever >> 12: bond1: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc >> noqueue master NFS state UP qlen 1000 >> link/ether 00:21:5a:9b:b7:97 brd ff:ff:ff:ff:ff:ff >> 13: NFS: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue >> state UP qlen 1000 >> link/ether 00:21:5a:9b:b7:97 brd ff:ff:ff:ff:ff:ff >> inet 10.5.160.46/22 brd 10.5.163.255 scope global NFS >> valid_lft forever preferred_lft forever >> inet6 fe80::221:5aff:fe9b:b797/64 scope link >> valid_lft forever preferred_lft forever >> 14: ;vdsmdummy;: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN >> qlen 1000 >> link/ether be:ae:6d:b7:0f:f5 brd ff:ff:ff:ff:ff:ff >> 16: qvo86c46fed-dd@qvb86c46fed-dd: >> <BROADCAST,MULTICAST,PROMISC,UP,LOWER_UP> mtu 1500 qdisc noqueue state >> UP qlen 1000 >> link/ether 5e:d1:06:97:f0:19 brd ff:ff:ff:ff:ff:ff >> inet6 fe80::5cd1:6ff:fe97:f019/64 scope link >> valid_lft forever preferred_lft forever >> 17: qvb86c46fed-dd@qvo86c46fed-dd: >> <BROADCAST,MULTICAST,PROMISC,UP,LOWER_UP> mtu 1500 qdisc noqueue state >> UP qlen 1000 >> link/ether e6:73:ee:8e:03:09 brd ff:ff:ff:ff:ff:ff >> inet6 fe80::e473:eeff:fe8e:309/64 scope link >> valid_lft forever preferred_lft forever >> 18: qbr86c46fed-dd: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state >> DOWN qlen 1000 >> link/ether fe:1a:4a:16:01:51 brd ff:ff:ff:ff:ff:ff >> 21: tap86c46fed-dd: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc >> pfifo_fast master qbr86c46fed-dd state UNKNOWN qlen 1000 >> link/ether fe:1a:4a:16:01:51 brd ff:ff:ff:ff:ff:ff >> >> >> [root@ovirt002 ~]# brctl show >> bridge name bridge id STP enabled interfaces >> ;vdsmdummy; 8000.000000000000 no >> NFS 8000.00215a9bb797 no bond1 >> ovirtmgmt 8000.00215a9bb793 no bond0 >> qbr86c46fed-dd 8000.fe1a4a160151 no tap86c46fed-dd >> >> >> [root@ovirt002 ~]# ovs-vsctl show >> ovs-vsctl: unix:/var/run/openvswitch/db.sock: database connection >> failed (No such file or directory) >> >> >> If you need other clarifications i'll dig without problem in this env. >> >> Luca >> >> >> >> >> -- >> "E' assurdo impiegare gli uomini di intelligenza eccellente per fare >> calcoli che potrebbero essere affidati a chiunque se si usassero delle >> macchine" >> Gottfried Wilhelm von Leibnitz, Filosofo e Matematico (1646-1716) >> >> "Internet è la più grande biblioteca del mondo. >> Ma il problema è che i libri sono tutti sparsi sul pavimento" >> John Allen Paulos, Matematico (1945-vivente) >> >> Luca 'remix_tj' Lorenzetto, http://www.remixtj.net , < >> lorenzetto.luca(a)gmail.com&gt; >> > > >

On Fri, Mar 10, 2017 at 2:41 PM, Marcin Mirecki <mmirecki(a)redhat.com&gt; wrote: You mean when i power on a VM that is using that external network? -- "E' assurdo impiegare gli uomini di intelligenza eccellente per fare calcoli che potrebbero essere affidati a chiunque se si usassero delle macchine" Gottfried Wilhelm von Leibnitz, Filosofo e Matematico (1646-1716) "Internet è la più grande biblioteca del mondo. Ma il problema è che i libri sono tutti sparsi sul pavimento" John Allen Paulos, Matematico (1945-vivente) Luca 'remix_tj' Lorenzetto, http://www.remixtj.net , <lorenzetto.luca(a)gmail.com&gt;

Hello, now, didn't understand why, the VM isn't starting. This is the VDSM log when i try to poweron while attached to an external network. If i attach to ovirtmgmt or other standard network, vm boots. Luca -- "E' assurdo impiegare gli uomini di intelligenza eccellente per fare calcoli che potrebbero essere affidati a chiunque se si usassero delle macchine" Gottfried Wilhelm von Leibnitz, Filosofo e Matematico (1646-1716) "Internet è la più grande biblioteca del mondo. Ma il problema è che i libri sono tutti sparsi sul pavimento" John Allen Paulos, Matematico (1945-vivente) Luca 'remix_tj' Lorenzetto, http://www.remixtj.net , <lorenzetto.luca(a)gmail.com&gt;

PROVIDER_FAILURE indicates a problem with the external network provider. The detailed cause of the error should be present in the vdsm logs. On Mon, Mar 13, 2017 at 11:37 AM, Luca 'remix_tj' Lorenzetto < lorenzetto.luca(a)gmail.com&gt; wrote:

On Tue, Mar 21, 2017 at 12:59 PM, Luca 'remix_tj' Lorenzetto <lorenzetto.luca(a)gmail.com&gt; wrote: [cut] Hello all, after some time i've been able to complete the integration between neutron (deployed with TripleO) and oVirt 4.1 What did i have to do? First, to have the configuration done directly by ovirt-manager i had to remove the host and re-add it specifying the external network provider in the related tab. The external provider should be configured also to point to the rabbitmq service listening into the internal_api network of Openstack. Additionally, all hosts that should use the openstack networks need to have access to the tenant networks directly or through router (in my case the tenant network is not routed, so i had to put an interface of the host in that network). Then the new host procedure should install and configure all requirements but: - no openstack packages are installed in ovirt-node-ng image - no repos available with openstack packages So i had to install manually this packages: openstack-neutron openstack-neutron-openvswitch openstack-utils openstack-selinux from RDO repos (and required to enable CentOS repos for dependencies) Then install also vsdm openstacknet hook (which is not available in enabled repos). Once did this also removing a default iptables policy: -A INPUT -j REJECT --reject-with icmp-host-prohibited that denies VXLAN to be established. Maybe a more detailed rule can be added allowing only the VXLAN port, but since i'm lazy and is only a lab, i tried with the ANY-ANY-ALLOW method. Then if the user adds a new host using as network provider a neutron provider, the configuration of neutron-openvswitch service should go on directly and configure all the requirements. After that the integration is working as i expect. If you want to do with already existing hosts you need to configure by hand openstack-neutron. Not a big deal, but if someone does this for you is better. I'll post my configuration in another email (need to check if is ok and has no useless entries). Had to open this two bugs for asking integration in the setup procedure and in ovirt-host-ng. https://bugzilla.redhat.com/show_bug.cgi?id=1446264 https://bugzilla.redhat.com/show_bug.cgi?id=1446306 Luca -- "E' assurdo impiegare gli uomini di intelligenza eccellente per fare calcoli che potrebbero essere affidati a chiunque se si usassero delle macchine" Gottfried Wilhelm von Leibnitz, Filosofo e Matematico (1646-1716) "Internet è la più grande biblioteca del mondo. Ma il problema è che i libri sono tutti sparsi sul pavimento" John Allen Paulos, Matematico (1945-vivente) Luca 'remix_tj' Lorenzetto, http://www.remixtj.net , <lorenzetto.luca(a)gmail.com&gt;