[ovirt-users] Re: New failure Gluster deploy: Set granual-entry-heal on --> Bricks down
We ran in to this issue as well when trying to install Ovirt Hyperconverged. The root issue is that kmod-kvdo in Centos 8 (and probably upstream) is built for a specific kernel and if you don't run that kernel it is not found. This is a major issue even if you match the kernel version then if the kernel is updated then your volume will fail because a rpm package for kmod-kvdo for that specific kernel would have to be built. It doesn't even declare a rpm dependency on the kernel version it works with. I cloned the git repo from https://github.com/dm-vdo/kvdo and built a rpm from there, it uses dkms so it will build a module for the running kernel and if the kernel is updated then a new module for that version will be built. Works like a charm every time, but I haven't yet tried to run the hyperconverged wizard again. /Sverker Den 2021-01-11 kl. 21:32, skrev Charles Lam: Dear Strahil and Ritesh, Thank you both. I am back where I started with: "One or more bricks could be down. Please execute the command again after bringing all bricks online and finishing any pending heals\nVolume heal failed.", "stdout_lines": ["One or more bricks could be down. Please execute the command again after bringing all bricks online and finishing any pending heals", "Volume heal failed."] Regarding my most recent issue: "vdo: ERROR - Kernel module kvdo not installed\nvdo: ERROR - modprobe: FATAL: Module kvdo not found in directory /lib/modules/4.18.0-240.1.1.el8_3.x86_64\n" Per Strahil's note, I checked for kvdo: [r...@host1.tld.com conf.d]# rpm -qa | grep vdo libblockdev-vdo-2.24-1.el8.x86_64 vdo-6.2.3.114-14.el8.x86_64 kmod-kvdo-6.2.2.117-65.el8.x86_64 [r...@host1.tld.com conf.d]# [r...@host2.tld.com conf.d]# rpm -qa | grep vdo libblockdev-vdo-2.24-1.el8.x86_64 vdo-6.2.3.114-14.el8.x86_64 kmod-kvdo-6.2.2.117-65.el8.x86_64 [r...@host2.tld.com conf.d]# [r...@host3.tld.com ~]# rpm -qa | grep vdo libblockdev-vdo-2.24-1.el8.x86_64 vdo-6.2.3.114-14.el8.x86_64 kmod-kvdo-6.2.2.117-65.el8.x86_64 [r...@host3.tld.com ~]# I found https://unix.stackexchange.com/questions/624011/problem-on-centos-8-with-creating-vdo-kernel-module-kvdo-not-installed which pointed to https://bugs.centos.org/view.php?id=17928. As suggested on the CentOS bug tracker I attempted to manually install vdo-support-6.2.4.14-14.el8.x86_64 vdo-6.2.4.14-14.el8.x86_64 kmod-kvdo-6.2.3.91-73.el8.x86_64 but there was a dependency that kernel-core be greater than what I was installed, so I manually upgraded kernel-core to kernel-core-4.18.0-259.el8.x86_64.rpm then upgraded vdo and kmod-kvdo to vdo-6.2.4.14-14.el8.x86_64.rpm kmod-kvdo-6.2.4.26-76.el8.x86_64.rpm and installed vdo-support-6.2.4.14-14.el8.x86_64.rpm. Upon clean-up and redeploy I am now back at Gluster deploy failing at TASK [gluster.features/roles/gluster_hci : Set granual-entry-heal on] ** task path: /etc/ansible/roles/gluster.features/roles/gluster_hci/tasks/hci_volumes.yml:67 failed: [fmov1n1.sn.dtcorp.com] (item={'volname': 'engine', 'brick': '/gluster_bricks/engine/engine', 'arbiter': 0}) => {"ansible_loop_var": "item", "changed": true, "cmd": ["gluster", "volume", "heal", "engine", "granular-entry-heal", "enable"], "delta": "0:00:10.098573", "end": "2021-01-11 19:27:05.333720", "item": {"arbiter": 0, "brick": "/gluster_bricks/engine/engine", "volname": "engine"}, "msg": "non-zero return code", "rc": 107, "start": "2021-01-11 19:26:55.235147", "stderr": "", "stderr_lines": [], "stdout": "One or more bricks could be down. Please execute the command again after bringing all bricks online and finishing any pending heals\nVolume heal failed.", "stdout_lines": ["One or more bricks could be down. Please execute the command again after bringing all bricks online and finishing any pending heals", "Volume heal failed."]} failed: [fmov1n1.sn.dtcorp.com] (item={'volname': 'data', 'brick': '/gluster_bricks/data/data', 'arbiter': 0}) => {"ansible_loop_var": "item", "changed": true, "cmd": ["gluster", "volume", "heal", "data", "granular-entry-heal", "enable"], "delta": "0:00:10.099670", "end": "2021-01-11 19:27:20.564554", "item": {"arbiter": 0, "brick": "/gluster_bricks/data/data", "volname": "data"}, "msg": "non-zero return code", "rc": 107, "start": "2021-01-11 19:27:10.464884", "stderr": "", "stderr_lines": [], "stdout": "One or more bricks could be down. Please execute the command again after bringing all bricks online and finishing any pending heals\nVolume heal failed.", "stdout_lines": ["One or more bricks could be down. Please execute the command again after bringing all bricks online and finishing any pending heals", "Volume heal failed."]} failed: [fmov1n1.sn.dtcorp.com] (item={'volname': 'vmstore', 'brick': '/gluster_bricks/vmstore/vmstore', 'arbiter': 0}) => {"ansible_loop_var": "item", "changed": true, "cmd": ["gluster", "volume", "heal", "vmstore", "granular-entry-heal", "enable"], "delta": "0:00:10.104624", "end": "2021-01-11 19:27:35.774230", "ite
[ovirt-users] Re: Hosted engine install failure: ipv6.gateway: gateway cannot be set if there are no addresses configured
Hi Ales, this is a CentOS 8 so my impression was that you always have NetworkManager then? At least my attempt to remove it failed miserably. The enp4s0 config was created by the install, so it should be controlled by NetworkManager. /Sverker Den 2020-09-03 kl. 12:29, skrev Ales Musil: On Thu, Sep 3, 2020 at 12:21 PM Sverker Abrahamsson mailto:sver...@abrahamsson.com>> wrote: Hi Ales, right now I have a manually created ovirtmgmt bridge (virbr0 and vnet0 seems to be created during the failed attempt to deploy hosted engine): [root@h1-mgmt ~]# nmcli con show NAME UUID TYPE DEVICE enp4s0 af7ccb53-011b-4c36-998a-1878b4ae7100 ethernet enp4s0 Bridge ovirtmgmt 9a0b07c0-2983-fe97-ec7f-ad2b51c3a3f0 bridge ovirtmgmt virbr0 aa593151-2c12-4cf7-985b-f105b3575d09 bridge virbr0 enp4s0.4000 ecc8064d-18c1-99b7-3fe4-9c5a593ece6f vlan enp4s0.4000 vnet0 a6db45bd-93c8-4c37-85fc-0c58ba3e9d00 tun vnet0 [root@h1-mgmt ~]# nmstatectl show --- dns-resolver: config: search: [] server: - 213.133.98.98 running: search: [] server: - 213.133.98.98 route-rules: config: [] routes: config: - destination: 0.0.0.0/0 <http://0.0.0.0/0> metric: -1 next-hop-address: 144.76.84.65 next-hop-interface: enp4s0 table-id: 0 - destination: ::/0 metric: -1 next-hop-address: fe80::1 next-hop-interface: enp4s0 table-id: 0 running: - destination: 0.0.0.0/0 <http://0.0.0.0/0> metric: 100 next-hop-address: 144.76.84.65 next-hop-interface: enp4s0 table-id: 254 - destination: 144.76.84.65/32 <http://144.76.84.65/32> metric: 100 next-hop-address: '' next-hop-interface: enp4s0 table-id: 254 - destination: 172.27.1.0/24 <http://172.27.1.0/24> metric: 425 next-hop-address: '' next-hop-interface: ovirtmgmt table-id: 254 - destination: 192.168.1.0/24 <http://192.168.1.0/24> metric: 0 next-hop-address: '' next-hop-interface: virbr0 table-id: 254 - destination: 2a01:4f8:192:1148::/64 metric: 100 next-hop-address: '' next-hop-interface: enp4s0 table-id: 254 - destination: ::/0 metric: 100 next-hop-address: fe80::1 next-hop-interface: enp4s0 table-id: 254 - destination: fe80::/64 metric: 100 next-hop-address: '' next-hop-interface: enp4s0 table-id: 254 - destination: ff00::/8 metric: 256 next-hop-address: '' next-hop-interface: enp4s0 table-id: 255 interfaces: - name: ;vdsmdummy; type: linux-bridge state: down ipv4: enabled: false ipv6: enabled: false mac-address: DE:D3:A8:24:27:F6 mtu: 1500 - name: br-int type: unknown state: down ipv4: enabled: false ipv6: enabled: false mac-address: 6E:37:94:63:E0:4B mtu: 1500 - name: enp4s0 type: ethernet state: up ethernet: auto-negotiation: true duplex: full speed: 1000 ipv4: address: - ip: 144.76.84.73 prefix-length: 32 dhcp: false enabled: true ipv6: address: - ip: 2a01:4f8:192:1148::2 prefix-length: 64 - ip: fe80::62a4:4cff:fee9:4ac prefix-length: 64 auto-dns: true auto-gateway: true auto-routes: true autoconf: true dhcp: true enabled: true mac-address: 60:A4:4C:E9:04:AC mtu: 1500 - name: enp4s0.4000 type: vlan state: up ipv4: dhcp: false enabled: false ipv6: enabled: false mac-address: 60:A4:4C:E9:04:AC mtu: 1500 vlan: base-iface: enp4s0 id: 4000 - name: lo type: unknown state: down ipv4: enabled: false ipv6: enabled: false mtu: 65536 - name: ovirtmgmt type: linux-bridge state: up bridge: options: group-forward-mask: 0 mac-ageing-time: 300 multicast-snooping: true stp: enabled: false forward-delay: 15 hello-time: 2 max-age: 20 priority: 32768 port: - name: enp4s0.4000 stp-hairpin-mode: false stp-path-cost: 100 stp-priority: 32 ipv4: address: - ip: 172.27.1.1 prefix-length: 24 dhcp: false enabled: true ipv6: autoconf: false dhcp: false enabled: false
[ovirt-users] Re: Hosted engine install failure: ipv6.gateway: gateway cannot be set if there are no addresses configured
back to not have the static bridge and a reboot I get this: [root@h1-mgmt ~]# nmcli con show NAME UUID TYPE DEVICE enp4s0 af7ccb53-011b-4c36-998a-1878b4ae7100 ethernet enp4s0 enp4s0.4000 ecc8064d-18c1-99b7-3fe4-9c5a593ece6f vlan enp4s0.4000 [root@h1-mgmt ~]# nmstatectl show --- dns-resolver: config: search: [] server: - 213.133.98.98 running: search: [] server: - 213.133.98.98 route-rules: config: [] routes: config: - destination: 0.0.0.0/0 metric: -1 next-hop-address: 144.76.84.65 next-hop-interface: enp4s0 table-id: 0 - destination: ::/0 metric: -1 next-hop-address: fe80::1 next-hop-interface: enp4s0 table-id: 0 running: - destination: 0.0.0.0/0 metric: 100 next-hop-address: 144.76.84.65 next-hop-interface: enp4s0 table-id: 254 - destination: 144.76.84.65/32 metric: 100 next-hop-address: '' next-hop-interface: enp4s0 table-id: 254 - destination: 172.27.1.0/24 metric: 400 next-hop-address: '' next-hop-interface: enp4s0.4000 table-id: 254 - destination: 2a01:4f8:192:1148::/64 metric: 100 next-hop-address: '' next-hop-interface: enp4s0 table-id: 254 - destination: ::/0 metric: 100 next-hop-address: fe80::1 next-hop-interface: enp4s0 table-id: 254 - destination: fe80::/64 metric: 100 next-hop-address: '' next-hop-interface: enp4s0 table-id: 254 - destination: ff00::/8 metric: 256 next-hop-address: '' next-hop-interface: enp4s0 table-id: 255 interfaces: - name: ;vdsmdummy; type: linux-bridge state: down ipv4: enabled: false ipv6: enabled: false mac-address: B2:9E:E0:61:71:88 mtu: 1500 - name: br-int type: unknown state: down ipv4: enabled: false ipv6: enabled: false mac-address: 6E:37:94:63:E0:4B mtu: 1500 - name: enp4s0 type: ethernet state: up ethernet: auto-negotiation: true duplex: full speed: 1000 ipv4: address: - ip: 144.76.84.73 prefix-length: 32 dhcp: false enabled: true ipv6: address: - ip: 2a01:4f8:192:1148::2 prefix-length: 64 - ip: fe80::62a4:4cff:fee9:4ac prefix-length: 64 auto-dns: true auto-gateway: true auto-routes: true autoconf: true dhcp: true enabled: true mac-address: 60:A4:4C:E9:04:AC mtu: 1500 - name: enp4s0.4000 type: vlan state: up ipv4: address: - ip: 172.27.1.1 prefix-length: 24 dhcp: false enabled: true ipv6: autoconf: false dhcp: false enabled: false mac-address: 60:A4:4C:E9:04:AC mtu: 1500 vlan: base-iface: enp4s0 id: 4000 - name: lo type: unknown state: down ipv4: enabled: false ipv6: enabled: false mtu: 65536 - name: ovs-system type: unknown state: down ipv4: enabled: false ipv6: enabled: false mac-address: A2:35:7A:6C:B7:EF mtu: 1500 /Sverker Den 2020-09-03 kl. 11:54, skrev Ales Musil: On Thu, Sep 3, 2020 at 11:51 AM Sverker Abrahamsson via Users mailto:users@ovirt.org>> wrote: Hi Dominik That is my issue, I don't get to where I can get the ovirtmgmt bridge established because vdsm insists on creating it. It used to be possible to create that bridge statically and vdsm would just skip it but seems to be broken now. If it would be possible to use OVN for the management network that would solve my issue and would be the preferable solution, but as you write that isn't possible which was what I suspected. Do you have any other suggestion on how to solve this issue? That I get the external interface untagged and the internal network tagged is not possible to change. /Sverker Hello Sverker, can you please share output from "nmcli con show" and "nmstatectl show"? Thank you. Regards, Ales Den 2020-09-03 kl. 10:52, skrev Dominik Holler: On Wed, Sep 2, 2020 at 10:38 PM Sverker Abrahamsson via Users mailto:users@ovirt.org>> wrote: Well, unforturnatly I don't have a choise since it is out of my control. I only have one physical network port where the external traffic is untagged and the internal vlan is tagged. If I could run with OVN OVN is for VM traffic only, not usable for the management network. instead I wouldn't need that tagged vlan, but I haven't been able to get that to work neither. Please let us know if OVN does not work for VM traffic for you. It's perfectly possible to have both tagged and untagged traffic on the same switch port, issue is that vdsm tries to take control over the network without being able to be flexible enough.. I'm attempting now to have ovirtmgmt bridge created before, that us
[ovirt-users] Re: Hosted engine install failure: ipv6.gateway: gateway cannot be set if there are no addresses configured
Hi Dominik That is my issue, I don't get to where I can get the ovirtmgmt bridge established because vdsm insists on creating it. It used to be possible to create that bridge statically and vdsm would just skip it but seems to be broken now. If it would be possible to use OVN for the management network that would solve my issue and would be the preferable solution, but as you write that isn't possible which was what I suspected. Do you have any other suggestion on how to solve this issue? That I get the external interface untagged and the internal network tagged is not possible to change. /Sverker Den 2020-09-03 kl. 10:52, skrev Dominik Holler: On Wed, Sep 2, 2020 at 10:38 PM Sverker Abrahamsson via Users mailto:users@ovirt.org>> wrote: Well, unforturnatly I don't have a choise since it is out of my control. I only have one physical network port where the external traffic is untagged and the internal vlan is tagged. If I could run with OVN OVN is for VM traffic only, not usable for the management network. instead I wouldn't need that tagged vlan, but I haven't been able to get that to work neither. Please let us know if OVN does not work for VM traffic for you. It's perfectly possible to have both tagged and untagged traffic on the same switch port, issue is that vdsm tries to take control over the network without being able to be flexible enough.. I'm attempting now to have ovirtmgmt bridge created before, that used to be possible but according to previous mails on the list it went broken somewhere at 4.x. /Sverker Den 2020-09-02 kl. 21:39, skrev Strahil Nikolov: > Switchports can either be tagged or untagged. > I'm not sure that your setup is supported at all. > > Best Regards, > Strahil Nikolov > > > > > > > В сряда, 2 септември 2020 г., 20:41:57 Гринуич+3, Sverker Abrahamsson via Users mailto:users@ovirt.org>> написа: > > > > > > Pretty formatting the "desired state" it seems that vdsm tries to remove > the ip of my underlying interface, that is enp4s0: > > { > 'interfaces': [{ > 'name': 'enp4s0', > 'state': 'up', > 'mtu': 1500 > }, { > 'vlan': { > 'id': 4000, > 'base-iface': 'enp4s0' > }, > 'name': 'enp4s0.4000', > 'type': 'vlan', > 'state': 'up', > 'mtu': 1500, > 'ipv4': { > 'enabled': False > }, > 'ipv6': { > 'enabled': False > } > }, { > 'name': 'ovirtmgmt', > 'type': 'linux-bridge', > 'state': 'up', > 'mtu': 1500, > 'bridge': { > 'port': [{ > 'name': 'enp4s0.4000' > } > ], > 'options': { > 'stp': { > 'enabled': False > } > } > }, > 'ipv4': { > 'enabled': True, > 'address': [{ > 'ip': '172.27.1.1', > 'prefix-length': 24 > } > ], > 'dhcp': False > }, > 'ipv6': { > 'enabled': False > } > } > ], > 'dns-resolver': { > 'config' > : { > 'server': ['213.133.98.98'] > } > } > } > Thanks, this is helpful information. Can you please share the getCapabilities result sent from vdsm to Engine directly before the setupNetworks request, and the parameters of the setupNetworks request from Engine to vdsm? Both are in the vdsm.log during adding the host.
[ovirt-users] Re: Hosted engine install failure: ipv6.gateway: gateway cannot be set if there are no addresses configured
Hi Strahil, there is nothing wrong with the switch other than that I don't have any control over it so the network setup is as it is. The issue is that vdsm breaks the network setup because it's not flexible enough. /Sverker Den 2020-09-03 kl. 00:47, skrev Strahil Nikolov: What is you switch brand & model ? Maybe someone more experienced in networking can help. Best Regards, Strahil Nikolov В сряда, 2 септември 2020 г., 23:39:57 Гринуич+3, Sverker Abrahamsson via Users написа: Well, unforturnatly I don't have a choise since it is out of my control. I only have one physical network port where the external traffic is untagged and the internal vlan is tagged. If I could run with OVN instead I wouldn't need that tagged vlan, but I haven't been able to get that to work neither. It's perfectly possible to have both tagged and untagged traffic on the same switch port, issue is that vdsm tries to take control over the network without being able to be flexible enough.. I'm attempting now to have ovirtmgmt bridge created before, that used to be possible but according to previous mails on the list it went broken somewhere at 4.x. /Sverker Den 2020-09-02 kl. 21:39, skrev Strahil Nikolov: Switchports can either be tagged or untagged. I'm not sure that your setup is supported at all. Best Regards, Strahil Nikolov В сряда, 2 септември 2020 г., 20:41:57 Гринуич+3, Sverker Abrahamsson via Users написа: Pretty formatting the "desired state" it seems that vdsm tries to remove the ip of my underlying interface, that is enp4s0: { 'interfaces': [{ 'name': 'enp4s0', 'state': 'up', 'mtu': 1500 }, { 'vlan': { 'id': 4000, 'base-iface': 'enp4s0' }, 'name': 'enp4s0.4000', 'type': 'vlan', 'state': 'up', 'mtu': 1500, 'ipv4': { 'enabled': False }, 'ipv6': { 'enabled': False } }, { 'name': 'ovirtmgmt', 'type': 'linux-bridge', 'state': 'up', 'mtu': 1500, 'bridge': { 'port': [{ 'name': 'enp4s0.4000' } ], 'options': { 'stp': { 'enabled': False } } }, 'ipv4': { 'enabled': True, 'address': [{ 'ip': '172.27.1.1', 'prefix-length': 24 } ], 'dhcp': False }, 'ipv6': { 'enabled': False } } ], 'dns-resolver': { 'config' : { 'server': ['213.133.98.98'] } } } This is my interfaces before vdsm attemtpts to change the config: enp4s0: flags=4163 mtu 1500 inet 144.76.84.73 netmask 255.255.255.255 broadcast 0.0.0.0 inet6 fe80::62a4:4cff:fee9:4ac prefixlen 64 scopeid 0x20 inet6 2a01:4f8:192:1148::2 prefixlen 64 scopeid 0x0 ether 60:a4:4c:e9:04:ac txqueuelen 1000 (Ethernet) RX packets 293442 bytes 385541799 (367.6 MiB) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 91095 bytes 31160348 (29.7 MiB) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 device interrupt 17 memory 0xf7d0-f7d2 enp4s0.4000: flags=4163 mtu 1500 inet 172.27.1.1 netmask 255.255.255.0 broadcast 172.27.1.255 inet6 fe80::62a4:4cff:fee9:4ac prefixlen 64 scopeid 0x20 ether 60:a4:4c:e9:04:ac txqueuelen 1000 (Ethernet) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 13 bytes 938 (938.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 I.e. enp4s0 is the external interface that must not be changed, bridge must be created on the vlan interface. I would prefer to create the bridge manually and not through vdsm if that is possible. /Sverker Den 2020-09-02 kl. 19:14, skrev Sverk
[ovirt-users] Re: Hosted engine install failure: ipv6.gateway: gateway cannot be set if there are no addresses configured
Well, unforturnatly I don't have a choise since it is out of my control. I only have one physical network port where the external traffic is untagged and the internal vlan is tagged. If I could run with OVN instead I wouldn't need that tagged vlan, but I haven't been able to get that to work neither. It's perfectly possible to have both tagged and untagged traffic on the same switch port, issue is that vdsm tries to take control over the network without being able to be flexible enough.. I'm attempting now to have ovirtmgmt bridge created before, that used to be possible but according to previous mails on the list it went broken somewhere at 4.x. /Sverker Den 2020-09-02 kl. 21:39, skrev Strahil Nikolov: Switchports can either be tagged or untagged. I'm not sure that your setup is supported at all. Best Regards, Strahil Nikolov В сряда, 2 септември 2020 г., 20:41:57 Гринуич+3, Sverker Abrahamsson via Users написа: Pretty formatting the "desired state" it seems that vdsm tries to remove the ip of my underlying interface, that is enp4s0: { 'interfaces': [{ 'name': 'enp4s0', 'state': 'up', 'mtu': 1500 }, { 'vlan': { 'id': 4000, 'base-iface': 'enp4s0' }, 'name': 'enp4s0.4000', 'type': 'vlan', 'state': 'up', 'mtu': 1500, 'ipv4': { 'enabled': False }, 'ipv6': { 'enabled': False } }, { 'name': 'ovirtmgmt', 'type': 'linux-bridge', 'state': 'up', 'mtu': 1500, 'bridge': { 'port': [{ 'name': 'enp4s0.4000' } ], 'options': { 'stp': { 'enabled': False } } }, 'ipv4': { 'enabled': True, 'address': [{ 'ip': '172.27.1.1', 'prefix-length': 24 } ], 'dhcp': False }, 'ipv6': { 'enabled': False } } ], 'dns-resolver': { 'config' : { 'server': ['213.133.98.98'] } } } This is my interfaces before vdsm attemtpts to change the config: enp4s0: flags=4163 mtu 1500 inet 144.76.84.73 netmask 255.255.255.255 broadcast 0.0.0.0 inet6 fe80::62a4:4cff:fee9:4ac prefixlen 64 scopeid 0x20 inet6 2a01:4f8:192:1148::2 prefixlen 64 scopeid 0x0 ether 60:a4:4c:e9:04:ac txqueuelen 1000 (Ethernet) RX packets 293442 bytes 385541799 (367.6 MiB) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 91095 bytes 31160348 (29.7 MiB) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 device interrupt 17 memory 0xf7d0-f7d2 enp4s0.4000: flags=4163 mtu 1500 inet 172.27.1.1 netmask 255.255.255.0 broadcast 172.27.1.255 inet6 fe80::62a4:4cff:fee9:4ac prefixlen 64 scopeid 0x20 ether 60:a4:4c:e9:04:ac txqueuelen 1000 (Ethernet) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 13 bytes 938 (938.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 I.e. enp4s0 is the external interface that must not be changed, bridge must be created on the vlan interface. I would prefer to create the bridge manually and not through vdsm if that is possible. /Sverker Den 2020-09-02 kl. 19:14, skrev Sverker Abrahamsson via Users: Hi, I'm attempting to install hosted engine but getting this failure: [ ERROR ] fatal: [localhost]: FAILED! => {"changed": false, "msg": "The host has been set in non_operational status, deployment errors: code 505: Host h1-mgmt.limetransit.com installation failed. Failed to configure management network on the host., code 1120: Failed to configure management network on host h1-mgmt.limetransit.com due to setup networks failure., code 9000: Failed to verify Power Management configuration for Host h1-mgmt.limetransit.com., code 10802: VDSM h1-mgmt.limetransit.com command HostSe
[ovirt-users] Hosted engine + OVN
Was it ever solved to install hosted engine with ovn? I tried a few years ago, got it almost to work but then gave up. ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/6KWV2MKV6E3ABQL6DGKVJ7FZTGNG3JPK/
[ovirt-users] Re: Hosted engine install failure: ipv6.gateway: gateway cannot be set if there are no addresses configured
Pretty formatting the "desired state" it seems that vdsm tries to remove the ip of my underlying interface, that is enp4s0: { 'interfaces': [{ 'name': 'enp4s0', 'state': 'up', 'mtu': 1500 }, { 'vlan': { 'id': 4000, 'base-iface': 'enp4s0' }, 'name': 'enp4s0.4000', 'type': 'vlan', 'state': 'up', 'mtu': 1500, 'ipv4': { 'enabled': False }, 'ipv6': { 'enabled': False } }, { 'name': 'ovirtmgmt', 'type': 'linux-bridge', 'state': 'up', 'mtu': 1500, 'bridge': { 'port': [{ 'name': 'enp4s0.4000' } ], 'options': { 'stp': { 'enabled': False } } }, 'ipv4': { 'enabled': True, 'address': [{ 'ip': '172.27.1.1', 'prefix-length': 24 } ], 'dhcp': False }, 'ipv6': { 'enabled': False } } ], 'dns-resolver': { 'config' : { 'server': ['213.133.98.98'] } } } This is my interfaces before vdsm attemtpts to change the config: enp4s0: flags=4163 mtu 1500 inet 144.76.84.73 netmask 255.255.255.255 broadcast 0.0.0.0 inet6 fe80::62a4:4cff:fee9:4ac prefixlen 64 scopeid 0x20 inet6 2a01:4f8:192:1148::2 prefixlen 64 scopeid 0x0 ether 60:a4:4c:e9:04:ac txqueuelen 1000 (Ethernet) RX packets 293442 bytes 385541799 (367.6 MiB) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 91095 bytes 31160348 (29.7 MiB) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 device interrupt 17 memory 0xf7d0-f7d2 enp4s0.4000: flags=4163 mtu 1500 inet 172.27.1.1 netmask 255.255.255.0 broadcast 172.27.1.255 inet6 fe80::62a4:4cff:fee9:4ac prefixlen 64 scopeid 0x20 ether 60:a4:4c:e9:04:ac txqueuelen 1000 (Ethernet) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 13 bytes 938 (938.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 I.e. enp4s0 is the external interface that must not be changed, bridge must be created on the vlan interface. I would prefer to create the bridge manually and not through vdsm if that is possible. /Sverker Den 2020-09-02 kl. 19:14, skrev Sverker Abrahamsson via Users: Hi, I'm attempting to install hosted engine but getting this failure: [ ERROR ] fatal: [localhost]: FAILED! => {"changed": false, "msg": "The host has been set in non_operational status, deployment errors: code 505: Host h1-mgmt.limetransit.com installation failed. Failed to configure management network on the host., code 1120: Failed to configure management network on host h1-mgmt.limetransit.com due to setup networks failure., code 9000: Failed to verify Power Management configuration for Host h1-mgmt.limetransit.com., code 10802: VDSM h1-mgmt.limetransit.com command HostSetupNetworksVDS failed: Internal JSON-RPC error: {'reason': 'Unexpected failure of libnm when running the mainloop: run execution'}, fix accordingly and re-deploy."} Looking in vdsm.log I find this which I believe is the root cause: MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:25,897::supervdsm_server::93::SuperVdsm.ServerCallback::(wrapper) call setupNetworks with ({'ovirtmgmt': {'vlan': '4000', 'netmask': '255.255.255.0', 'ipv6autoconf': False, 'nic': 'enp4s0', 'bridged': 'true', 'ipaddr': '172.27.1.1', 'defaultRoute': True, 'dhcpv6': False, 'STP': 'no', 'mtu': 1500, 'switch': 'legacy'}}, {}, {'connectivityTimeout': 120, 'commitOnSuccess': True, 'connectivityCh eck': 'true'}) {} MainProcess|jsonrpc/0::INFO::2020-09-02 16:38:25,897::api::220::root::(setupNe
[ovirt-users] Hosted engine install failure: ipv6.gateway: gateway cannot be set if there are no addresses configured
Hi, I'm attempting to install hosted engine but getting this failure: [ ERROR ] fatal: [localhost]: FAILED! => {"changed": false, "msg": "The host has been set in non_operational status, deployment errors: code 505: Host h1-mgmt.limetransit.com installation failed. Failed to configure management network on the host., code 1120: Failed to configure management network on host h1-mgmt.limetransit.com due to setup networks failure., code 9000: Failed to verify Power Management configuration for Host h1-mgmt.limetransit.com., code 10802: VDSM h1-mgmt.limetransit.com command HostSetupNetworksVDS failed: Internal JSON-RPC error: {'reason': 'Unexpected failure of libnm when running the mainloop: run execution'}, fix accordingly and re-deploy."} Looking in vdsm.log I find this which I believe is the root cause: MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:25,897::supervdsm_server::93::SuperVdsm.ServerCallback::(wrapper) call setupNetworks with ({'ovirtmgmt': {'vlan': '4000', 'netmask': '255.255.255.0', 'ipv6autoconf': False, 'nic': 'enp4s0', 'bridged': 'true', 'ipaddr': '172.27.1.1', 'defaultRoute': True, 'dhcpv6': False, 'STP': 'no', 'mtu': 1500, 'switch': 'legacy'}}, {}, {'connectivityTimeout': 120, 'commitOnSuccess': True, 'connectivityCh eck': 'true'}) {} MainProcess|jsonrpc/0::INFO::2020-09-02 16:38:25,897::api::220::root::(setupNetworks) Setting up network according to configuration: networks:{'ovirtmgmt': {'vlan': '4000', 'netmask': '255.255.255.0', 'ipv6autoconf': Fal se, 'nic': 'enp4s0', 'bridged': 'true', 'ipaddr': '172.27.1.1', 'defaultRoute': True, 'dhcpv6': False, 'STP': 'no', 'mtu': 1500, 'switch': 'legacy'}}, bondings:{}, options:{'connectivityTimeout': 120, 'commitOnSuccess': True, 'connectivityCheck': 'true'} MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:25,902::cmdutils::130::root::(exec_cmd) /sbin/tc qdisc show (cwd None) MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:25,906::cmdutils::138::root::(exec_cmd) SUCCESS: = b''; = 0 MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:25,945::vsctl::74::root::(commit) Executing commands: /usr/bin/ovs-vsctl --timeout=5 --oneline --format=json -- list Bridge -- list Port -- list Interface MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:25,945::cmdutils::130::root::(exec_cmd) /usr/bin/ovs-vsctl --timeout=5 --oneline --format=json -- list Bridge -- list Port -- list Interface (cwd None) MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:25,952::cmdutils::138::root::(exec_cmd) SUCCESS: = b''; = 0 MainProcess|jsonrpc/0::INFO::2020-09-02 16:38:25,957::netconfpersistence::58::root::(setNetwork) Adding network ovirtmgmt({'vlan': 4000, 'netmask': '255.255.255.0', 'ipv6autoconf': False, 'nic': 'enp4s0', 'bridged': True , 'ipaddr': '172.27.1.1', 'defaultRoute': True, 'dhcpv6': False, 'mtu': 1500, 'switch': 'legacy', 'stp': False, 'bootproto': 'none', 'nameservers': ['213.133.98.98']}) MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:25,958::commands::153::common.commands::(start) /usr/bin/taskset --cpu-list 0-7 /usr/libexec/vdsm/hooks/before_network_setup/50_fcoe (cwd None) MainProcess|jsonrpc/0::INFO::2020-09-02 16:38:26,154::hooks::122::root::(_runHooksDir) /usr/libexec/vdsm/hooks/before_network_setup/50_fcoe: rc=0 err=b'' MainProcess|jsonrpc/0::INFO::2020-09-02 16:38:26,155::configurator::195::root::(_setup_nmstate) Processing setup through nmstate MainProcess|jsonrpc/0::INFO::2020-09-02 16:38:26,175::configurator::197::root::(_setup_nmstate) Desired state: {'interfaces': [{'name': 'enp4s0', 'state': 'up', 'mtu': 1500}, {'vlan': {'id': 4000, 'base-iface': 'enp4s0'} , 'name': 'enp4s0.4000', 'type': 'vlan', 'state': 'up', 'mtu': 1500, 'ipv4': {'enabled': False}, 'ipv6': {'enabled': False}}, {'name': 'ovirtmgmt', 'type': 'linux-bridge', 'state': 'up', 'mtu': 1500, 'bridge': {'port': [ {'name': 'enp4s0.4000'}], 'options': {'stp': {'enabled': False}}}, 'ipv4': {'enabled': True, 'address': [{'ip': '172.27.1.1', 'prefix-length': 24}], 'dhcp': False}, 'ipv6': {'enabled': False}}], 'dns-resolver': {'config' : {'server': ['213.133.98.98']}}} MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:26,217::checkpoint::121::root::(create) Checkpoint /org/freedesktop/NetworkManager/Checkpoint/1 created for all devices: 60 MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:26,218::netapplier::239::root::(_add_interfaces) Adding new interfaces: ['ovirtmgmt'] MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:26,221::netapplier::251::root::(_edit_interfaces) Editing interfaces: ['enp4s0.4000', 'enp4s0'] MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:26,224::nmclient::136::root::(execute_next_action) Executing NM action: func=add_connection_async MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:26,231::connection::329::root::(_add_connection_callback) Connection adding succeeded: dev=ovirtmgmt MainProcess|jsonrpc/0::DEBUG::2020-09-02 16:38:26,232::nmclient::136::root::(execute_next_action) Executing NM action: func=commit_c
[ovirt-users] Re: Unable to start VM on host with OVS networking
In my setup the reason why I attempted to use OVS/OVN is that I (currently) have two hosts in a data center (Hetzner) which does not have a common subnet. Each of these hosts have one public IPv4 address. Because of that I want the ovirt networks to be on private ip ranges and they do not correspond to any physical network interfaces. I've been able to work around that by making dummy interfaces visible for vdsm to which it can bind the ovirtmgmt bridge. It also means that I need to connect the obirtmgmt bridges on the two hosts with tunnels. With OVS I can set that up manually and it works great, issue is that vdsm wants to control the network interfaces and creates an OVS bridge with random name to attach ovirtmgmt port to, hence I have no convenient way to connect these bridges. I could put a hook in after_network_setup to create tunnels, but I suppose the OVS bridges needs to have the same name on both sides. I want to use ovirt for a test environment with quite complex network, hence why OVN is desired to be able to set it up virtual. With OVN + physnet as you describe it I assume I should get the desired functionality as ovirtmgmt needs to be present on all hosts. Since you disabled changing switch type in engine gui I changed it in db, don't know if anything more is needed, and in the vdsm persistent file. On vdsm level that works fine, the OVS bridge with random name is created with ovirtmgmt port, but as you write since 4.2.3 it no longer creates the port for hosted engine on the OVS bridge. I worked around that with a hook script that add the missing elements to interface section of the vm xml. My understanding is that for OVN+physnet then OVS switch is needed but doesn't then the ovirtmgmt port has to be on an OVS bridge? If so then hosted engine must be able to connect it's port to the same bridge. Next concern is where should the OVN databases be? With hosted-engine setup they are created on the vm which seems like a chicken and the egg issue to me as how could the hosts where ovn-controller runs be able to connect to db before the vm has started? In my setup the vm only have private address which means the other host is not able to reach ovn database until the virtual network has established. Therefore I created OVN databases on one of the hosts, and configured with ssl cert/keys so that they are able to communicate fine. I configured ovirt-provider-ovn running on hosted engine vm to connect to the db host but even when ovn-remote is set to the correct address on the host the requests from ovirt-provider-ovn towards port 6641 still goes to the engine vm. Are there any additional steps that is needed to take? /Sverker Den 2018-05-25 kl. 17:41, skrev Dan Kenigsberg: I believe the user community deserves a little background for this decision. OVS has been "experimental" since ovirt-4.0.z with migration disabled by default. We were not aware of huge benefits it had over the default Linux bridge, and did not expect people to be using it in important deployments. I would love to hear your experience regarding our OVS support, and why you have chosen it. In ovirt-4.2.0, the way in which VM libvirt definition is built has changed considerably, and takes place in ovirt-engine, not in vdsm. The vdsm code that supports OVS connectivity was disabled in ovirt-4.2.3 which means that indeed, the experimental OVS feature is no longer available for direct usage (unless you still use cluster compatibility level 4.1) However, as Thomas Davis explains, with OVN + physnet, ovirt-4.2 gives you a matching functionality, including live migration out-of-the-box. OVS switchtype was upgraded from "experimental" to "tech-preview". I'd like to drop the advisory altogether, but we keep it because we still have bugs and missing features comparing to Linux bridge clusters. We've blocked changing the switchtype of existing clusters because this functionality is buggy (particularly on the SPM host), and as of ovirt-4.2, we do not have code to support live migration from a Linux bridge host to an OVS one. Only cold migration is possible. We kept it open over REST to allow testing and bugfixes to that flow, as well as usage by careful users. Thanks for using oVirt and its new features, and for engaging with the community. Regards, Dan. On Tue, May 22, 2018 at 9:20 PM, wrote: The answer is.. OVN replaced OVS as the networking technology. You cannot switch back to legacy, they disabled switching between ovs and legacy in the default (1st) datacenter using the gui. You can however, use Ansible to switch it. Remove the VDSM ovs setting, it will just mess you up, and it's not supported in 4.2 To able to migrate a VM in 4.2, you have use OVN with OVS. I did this a few months back, on a 4.2.2 hosted-engine setup: 0) To setup a node in a cluster, make sure the cluster is in OVS, not legacy. 1) Make sure you have an OVN controller setup somewhere. Default appears
[ovirt-users] Re: Unable to start VM on host with OVS networking
Hi Jonathan, it seems somewhat similar to what I've been struggeling with. I've worked around it with a hook script until someone can explain from where it should discover that the interface should be created on an ovs bridge. Look at the thread with topic "hosted engine with openvswitch". /Sverker Den 2018-05-17 kl. 22:09, skrev Jonathan Dieter: I have a production ovirt setup that's gone through multiple updates over the years. At some point when 4.0 or 4.1 came out, I switched from legacy networking to OVS, and everything worked perfectly until I upgraded to 4.2. Since I upgraded to 4.2, I've been getting messages that the networks were all out of sync, but everything continued working properly. Today I tracked down the network sync problem, fixed it on one of my three hosts, and then attempted to start a VM on the host. It refused to start with the error message: "Unable to add bridge ovirtmgmt port vnet0: Operation not supported". From what I can tell, the xml being generated is still for the old legacy network. I completely reinstalled the node, using the latest 4.2.3 node ISO image, and it still doesn't work. In the cluster, the switch type is "OVS (Experimental)" (and this option can't be changed, apparently), the compatibility version is 4.2, the firewall type is firewalld and there's no "Default Network Provider". I suspect that my upgrades have somehow left my system in half OVS/half legacy mode, but I'm not sure how to move it all the way to OVS mode and I don't want to mess with the other two hosts until I'm sure I've got it figured out. My (compressed) vdsm.log is at https://www.lesbg.com/jdieter/vdsm.log.xz and my (compressed) supervdsm.log is at https://www.lesbg.com/jdieter/supervdsm.log.xz. If anyone could point me in the right direction to get this fixed, I'd sure appreciate it. Jonathan ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org
[ovirt-users] Re: hosted engine with openvswitch
Hi I've been able to work around this issue by adding a hook which checks if the ovirtmgmt interface is a port on a ovs switch and if so add the correct data to interface element. I added this to before_device_create and before_nic_hotplug folders, it works as far as bringing the hosted engine up, but resyncing host network from engine side makes it loose the connection. /Sverker Den 2018-05-14 kl. 10:12, skrev Sverker Abrahamsson: Apparently the logs were too big for the mailing list, but should have been received by Petr and Dominik. Let me know if not. What it boils down to, I think, is that the xml sent to libvirt does not have the correct data for openvswitch in interface element, see snippet below. The question is, where should that be set? What I actually want to reach is to have ovirtmgmt network on OVN according to this description "https://gist.github.com/phoracek/ebbd9a6bcfcb14d88f6b463c8dccb27b"; so it might be an unnecessary step to get it to work with OVS switch but the content of interface element in xml sent to libvirt is similar so I think an understanding of where it comes from is necessary to troubleshoot that (if needed). /Sverker Den 2018-05-13 kl. 00:23, skrev Sverker Abrahamsson: Sure, they are attached. The attempt to start the hosted engine vm occurs at 14:29:42 Best regards Sverker Den 2018-05-11 kl. 13:39, skrev Petr Horacek: Hi, would you mind sharing the whole vdsm.log and supervdsm.log? Thanks, Petr 2018-05-10 22:48 GMT+02:00 Sverker Abrahamsson mailto:sver...@abrahamsson.com>>: Hi Dominik Den 2018-05-10 kl. 22:08, skrev Dominik Holler: On Thu, 10 May 2018 19:15:07 +0200 Sverker Abrahamsson mailto:sver...@abrahamsson.com>> wrote: Hi I have a problem with running hosted engine with openvswitch. I have Which version of oVirt? Ovirt version is 4.2.3.5, vdsm version 4.20.27 How do you configured hosted engine to use openvswitch? Did you follow the steps in https://lists.ovirt.org/pipermail/users/2017-March/080748.html <https://lists.ovirt.org/pipermail/users/2017-March/080748.html> ? I've followed the same steps as outlined in that mail just that I found it when I started to troubleshoot. one cluster where the ovirt engine runs on the host, there it works What is the switch type of this cluster? OVS and when starting a vm the interface definition looks like this: Where do you pick up this xml? From vdsm.log when the vm is created The xml for that vm as fetched from vdsm does not contain virtualport Where do you pick up this xml, maybe from a logfile or virsh? vdsm-client VM getInfo vmID=3efc0421-5ad5-4a04-9a0e-30bd8822b299 tag nor does it use the correct bridge, it looks like this: I.e. somewhere the definition is modified to contain the correct data to work with openvswitch On the other cluster where I try to run hosted engine I don't get the What is the switch type of this cluster? In the engine it's set to OVS, but at this point it hasn't started yet. Does the host have that setting somewhere? above behaviour. When the engine vm starts the interface settings are not modified to use the bridge in openvswitch, with the result that the vm fails to start: Can you share the error message from engine.log and vdsm.log, and maybe supervdsm.log? These are the error messages when vm fails to start, but it's caused by the interface element does not contain virtualport nor the correct bridge. vdsm.log: 2018-05-10 22:45:30,562+0200 ERROR (vm/f222c92f) [virt.vm] (vmId='f222c92f-91af-4d1c-860d-c2b406211496') The vm start process failed (vm:943) Traceback (most recent call last): File "/usr/lib/python2.7/site-packages/vdsm/virt/vm.py", line 872, in _startUnderlyingVm self._run() File "/usr/lib/python2.7/site-packages/vdsm/virt/vm.py", line 2872, in _run dom.createWithFlags(flags) File "/usr/lib/python2.7/site-packages/vdsm/common/libvirtconnection.py", line 130, in wrapper ret = f(*args, **kwargs) File "/usr/lib/python2.7/site-packages/vdsm/common/function.py", line
[ovirt-users] Re: Fwd: Re: Re: hosted engine with openvswitch
Apparently the logs were too big for the mailing list, but should have been received by Petr and Dominik. Let me know if not. What it boils down to, I think, is that the xml sent to libvirt does not have the correct data for openvswitch in interface element, see snippet below. The question is, where should that be set? What I actually want to reach is to have ovirtmgmt network on OVN according to this description "https://gist.github.com/phoracek/ebbd9a6bcfcb14d88f6b463c8dccb27b"; so it might be an unnecessary step to get it to work with OVS switch but the content of interface element in xml sent to libvirt is similar so I think an understanding of where it comes from is necessary to troubleshoot that (if needed). /Sverker Den 2018-05-13 kl. 00:23, skrev Sverker Abrahamsson: Sure, they are attached. The attempt to start the hosted engine vm occurs at 14:29:42 Best regards Sverker Den 2018-05-11 kl. 13:39, skrev Petr Horacek: Hi, would you mind sharing the whole vdsm.log and supervdsm.log? Thanks, Petr 2018-05-10 22:48 GMT+02:00 Sverker Abrahamsson mailto:sver...@abrahamsson.com>>: Hi Dominik Den 2018-05-10 kl. 22:08, skrev Dominik Holler: On Thu, 10 May 2018 19:15:07 +0200 Sverker Abrahamsson mailto:sver...@abrahamsson.com>> wrote: Hi I have a problem with running hosted engine with openvswitch. I have Which version of oVirt? Ovirt version is 4.2.3.5, vdsm version 4.20.27 How do you configured hosted engine to use openvswitch? Did you follow the steps in https://lists.ovirt.org/pipermail/users/2017-March/080748.html <https://lists.ovirt.org/pipermail/users/2017-March/080748.html> ? I've followed the same steps as outlined in that mail just that I found it when I started to troubleshoot. one cluster where the ovirt engine runs on the host, there it works What is the switch type of this cluster? OVS and when starting a vm the interface definition looks like this: Where do you pick up this xml? From vdsm.log when the vm is created The xml for that vm as fetched from vdsm does not contain virtualport Where do you pick up this xml, maybe from a logfile or virsh? vdsm-client VM getInfo vmID=3efc0421-5ad5-4a04-9a0e-30bd8822b299 tag nor does it use the correct bridge, it looks like this: I.e. somewhere the definition is modified to contain the correct data to work with openvswitch On the other cluster where I try to run hosted engine I don't get the What is the switch type of this cluster? In the engine it's set to OVS, but at this point it hasn't started yet. Does the host have that setting somewhere? above behaviour. When the engine vm starts the interface settings are not modified to use the bridge in openvswitch, with the result that the vm fails to start: Can you share the error message from engine.log and vdsm.log, and maybe supervdsm.log? These are the error messages when vm fails to start, but it's caused by the interface element does not contain virtualport nor the correct bridge. vdsm.log: 2018-05-10 22:45:30,562+0200 ERROR (vm/f222c92f) [virt.vm] (vmId='f222c92f-91af-4d1c-860d-c2b406211496') The vm start process failed (vm:943) Traceback (most recent call last): File "/usr/lib/python2.7/site-packages/vdsm/virt/vm.py", line 872, in _startUnderlyingVm self._run() File "/usr/lib/python2.7/site-packages/vdsm/virt/vm.py", line 2872, in _run dom.createWithFlags(flags) File "/usr/lib/python2.7/site-packages/vdsm/common/libvirtconnection.py", line 130, in wrapper ret = f(*args, **kwargs) File "/usr/lib/python2.7/site-packages/vdsm/common/function.py", line 92, in wrapper return func(inst, *args, **kwargs) File "/usr/lib64/python2.7/site-packages/libvirt.py", line 1099, in createWithFlags if ret == -1: raise libvirtError ('virDomainCreateWithFlags() failed', dom=self) libvirtError: Unable to add bridge ovirtmgmt port vnet0: Operation not supported 2018-05-10 22:45:30,562+0200 INFO (vm/f222c92f) [virt.vm] (vmId='f222c92f-91af-4d
[ovirt-users] Re: hosted engine with openvswitch
Hi Dominik Den 2018-05-10 kl. 22:08, skrev Dominik Holler: On Thu, 10 May 2018 19:15:07 +0200 Sverker Abrahamsson wrote: Hi I have a problem with running hosted engine with openvswitch. I have Which version of oVirt? Ovirt version is 4.2.3.5, vdsm version 4.20.27 How do you configured hosted engine to use openvswitch? Did you follow the steps in https://lists.ovirt.org/pipermail/users/2017-March/080748.html ? I've followed the same steps as outlined in that mail just that I found it when I started to troubleshoot. one cluster where the ovirt engine runs on the host, there it works What is the switch type of this cluster? OVS and when starting a vm the interface definition looks like this: Where do you pick up this xml? From vdsm.log when the vm is created The xml for that vm as fetched from vdsm does not contain virtualport Where do you pick up this xml, maybe from a logfile or virsh? vdsm-client VM getInfo vmID=3efc0421-5ad5-4a04-9a0e-30bd8822b299 tag nor does it use the correct bridge, it looks like this: I.e. somewhere the definition is modified to contain the correct data to work with openvswitch On the other cluster where I try to run hosted engine I don't get the What is the switch type of this cluster? In the engine it's set to OVS, but at this point it hasn't started yet. Does the host have that setting somewhere? above behaviour. When the engine vm starts the interface settings are not modified to use the bridge in openvswitch, with the result that the vm fails to start: Can you share the error message from engine.log and vdsm.log, and maybe supervdsm.log? These are the error messages when vm fails to start, but it's caused by the interface element does not contain virtualport nor the correct bridge. vdsm.log: 2018-05-10 22:45:30,562+0200 ERROR (vm/f222c92f) [virt.vm] (vmId='f222c92f-91af-4d1c-860d-c2b406211496') The vm start process failed (vm:943) Traceback (most recent call last): File "/usr/lib/python2.7/site-packages/vdsm/virt/vm.py", line 872, in _startUnderlyingVm self._run() File "/usr/lib/python2.7/site-packages/vdsm/virt/vm.py", line 2872, in _run dom.createWithFlags(flags) File "/usr/lib/python2.7/site-packages/vdsm/common/libvirtconnection.py", line 130, in wrapper ret = f(*args, **kwargs) File "/usr/lib/python2.7/site-packages/vdsm/common/function.py", line 92, in wrapper return func(inst, *args, **kwargs) File "/usr/lib64/python2.7/site-packages/libvirt.py", line 1099, in createWithFlags if ret == -1: raise libvirtError ('virDomainCreateWithFlags() failed', dom=self) libvirtError: Unable to add bridge ovirtmgmt port vnet0: Operation not supported 2018-05-10 22:45:30,562+0200 INFO (vm/f222c92f) [virt.vm] (vmId='f222c92f-91af-4d1c-860d-c2b406211496') Changed state to Down: Unable to add bridge ovirtmgmt port vnet0: Operation not supported (code=1) (vm:1683) 2018-05-10 22:45:30,604+0200 INFO (vm/f222c92f) [virt.vm] (vmId='f222c92f-91af-4d1c-860d-c2b406211496') Stopping connection (guestagent:438) It's the hosted engine vm that fails to start, hence I don't have any engine.log and supervdsm.log doesn't contain anything about this Last login: Thu May 10 16:23:48 2018 from 172.27.1.32 [root@h2 ~]# ovs-vsctl show dfcf7463-ce51-4115-9a3a-ecab9efa8146 Bridge "vdsmbr_H91hH5sG" Port "vdsmbr_H91hH5sG" Interface "vdsmbr_H91hH5sG" type: internal Port ovirtmgmt Interface ovirtmgmt type: internal Port "dummy0" Interface "dummy0" ovs_version: "2.9.0" I assumed first there is a hook that make the needed change, but the only hooks I can find that mentions openvswitch are ovirt_provider_ovn_hook and 50_openstacknet but both those would set the source bridge to br-int and not look up the dynamic name of the bridge as created by vdsm. One special thing about the host where I try to run hosted engine is that the there is a dummy port since otherwise I couldn't get vdsm to create the bridge, but that shouldn't affect changing the interface definition for the vm. Where should I look next? ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org
[ovirt-users] hosted engine with openvswitch
Hi I have a problem with running hosted engine with openvswitch. I have one cluster where the ovirt engine runs on the host, there it works and when starting a vm the interface definition looks like this: slot="0x03" type="pci" /> The xml for that vm as fetched from vdsm does not contain virtualport tag nor does it use the correct bridge, it looks like this: type="pci"/> I.e. somewhere the definition is modified to contain the correct data to work with openvswitch On the other cluster where I try to run hosted engine I don't get the above behaviour. When the engine vm starts the interface settings are not modified to use the bridge in openvswitch, with the result that the vm fails to start: slot="0x03" type="pci"/> Last login: Thu May 10 16:23:48 2018 from 172.27.1.32 [root@h2 ~]# ovs-vsctl show dfcf7463-ce51-4115-9a3a-ecab9efa8146 Bridge "vdsmbr_H91hH5sG" Port "vdsmbr_H91hH5sG" Interface "vdsmbr_H91hH5sG" type: internal Port ovirtmgmt Interface ovirtmgmt type: internal Port "dummy0" Interface "dummy0" ovs_version: "2.9.0" I assumed first there is a hook that make the needed change, but the only hooks I can find that mentions openvswitch are ovirt_provider_ovn_hook and 50_openstacknet but both those would set the source bridge to br-int and not look up the dynamic name of the bridge as created by vdsm. One special thing about the host where I try to run hosted engine is that the there is a dummy port since otherwise I couldn't get vdsm to create the bridge, but that shouldn't affect changing the interface definition for the vm. Where should I look next? ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org
Re: [ovirt-users] Network interface persistence
This discussion seems relevant: https://access.redhat.com/discussions/916973 For my CentOS machines there are various alternatives described here which should work, but what to do about my debian-based appliance where I can't get access enough to edit such things? I then need to be able to set up a consistent naming from the virtual bios. If it matters the switch type used is OVS. /Sverker Den 2018-03-26 kl. 13:41, skrev Sverker Abrahamsson: I have a number of vm's running under Ovirt 4.2, most are CentOS 7 and one Debian. Issue is that when multiple network interfaces are assigned they don't persist but it variates at boot which interface will be which. I've tried various methods, setting UUID and HWADDRESS in the ifcfg file and a file /etc/udev/rules.d/70-persistent-net.rules file like the below: SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1a:4a:16:01:63", KERNEL=="eth*", NAME="eth0" SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1a:4a:16:01:6d", KERNEL=="eth*", NAME="eth1" I've also tried in the gui to change the network profile each virtual card is attached, but even then the interfaces in vm will be oposide of what I had intended. How to accomplish with Ovirt to get network interface persistence? The Debian vm is an appliance so I can't log in to it to change network interfaces. /Sverker ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[ovirt-users] Network interface persistence
I have a number of vm's running under Ovirt 4.2, most are CentOS 7 and one Debian. Issue is that when multiple network interfaces are assigned they don't persist but it variates at boot which interface will be which. I've tried various methods, setting UUID and HWADDRESS in the ifcfg file and a file /etc/udev/rules.d/70-persistent-net.rules file like the below: SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1a:4a:16:01:63", KERNEL=="eth*", NAME="eth0" SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1a:4a:16:01:6d", KERNEL=="eth*", NAME="eth1" I've also tried in the gui to change the network profile each virtual card is attached, but even then the interfaces in vm will be oposide of what I had intended. How to accomplish with Ovirt to get network interface persistence? The Debian vm is an appliance so I can't log in to it to change network interfaces. /Sverker ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[ovirt-users] Error during SSO authentication Cannot authenticate user 'admin@internal'
Since upgrading my test lab to ovirt 4.2 I can't get ovirt-provider-ovn to work. From ovirt-provider-ovn.log: 2017-11-14 00:40:15,795 Request: POST : /v2.0///tokens 2017-11-14 00:40:15,795 Request body: { "auth" : { "passwordCredentials" : { "username" : "admin@internal", "password" : "x" } } } 2017-11-14 00:40:15,819 Starting new HTTPS connection (1): h2-int 2017-11-14 00:40:20,829 "POST /ovirt-engine/sso/oauth/token HTTP/1.1" 400 118 2017-11-14 00:40:20,830 Error during SSO authentication Cannot authenticate user 'admin@internal': The username or password is incorrect.. : access_deniedNone Traceback (most recent call last): File "/usr/share/ovirt-provider-ovn/handlers/base_handler.py", line 119, in _handle_request method, path_parts, content) File "/usr/share/ovirt-provider-ovn/handlers/selecting_handler.py", line 177, in handle_request handler, content, parameters File "/usr/share/ovirt-provider-ovn/handlers/keystone.py", line 28, in call_response_handler return response_handler(content, parameters) File "/usr/share/ovirt-provider-ovn/handlers/keystone_responses.py", line 58, in post_tokens user_password=user_password) File "/usr/share/ovirt-provider-ovn/auth/plugin_facade.py", line 26, in create_token return auth.core.plugin.create_token(user_at_domain, user_password) File "/usr/share/ovirt-provider-ovn/auth/plugins/ovirt/plugin.py", line 48, in create_token timeout=self._timeout()) File "/usr/share/ovirt-provider-ovn/auth/plugins/ovirt/sso.py", line 62, in create_token username, password, engine_url, ca_file, timeout) File "/usr/share/ovirt-provider-ovn/auth/plugins/ovirt/sso.py", line 54, in wrapper _check_for_error(response) File "/usr/share/ovirt-provider-ovn/auth/plugins/ovirt/sso.py", line 168, in _check_for_error result['error'], details)) Unauthorized: Error during SSO authentication Cannot authenticate user 'admin@internal': The username or password is incorrect.. : access_deniedNone And in engine.log: 2017-11-14 00:40:20,828+01 ERROR [org.ovirt.engine.core.sso.utils.SsoUtils] (default task-16) [] OAuthException access_denied: Cannot authenticate user 'admin@internal': The username or password is incorrect.. The password in the request is the same as used to log in to the admin portal and works fine there. /Sverker ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [ovirt-users] engine upgrade 3.5 -> 4.1.1
Hi Simone. I get the following result: # sudo -u postgres psql engine -c "SELECT vm_guid, run_on_vds FROM vms WHERE vm_name in (select 'HostedEngine' union select option_value from vdc_options where option_name='HostedEngineVmName')" could not change directory to "/root" vm_guid| run_on_vds --+ 4e4df6ea-ea01-465b-9424-4436aa24cec0 | (1 row) /Sverker Den 2017-04-28 kl. 10:32, skrev Simone Tiraboschi: On Thu, Apr 27, 2017 at 9:27 PM, Sverker Abrahamsson mailto:sver...@abrahamsson.com>> wrote: Hi I upgraded an old ovirt installation from 3.5 to latest version, but end up with this error message when running engine-setup: [ INFO ] Stage: Setup validation [ ERROR ] It seems that you are running your engine inside of the hosted-engine VM and are not in "Global Maintenance" mode. In that case you should put the system into the "Global Maintenance" mode before running engine-setup, or the hosted-engine HA agent might kill the machine, which might corrupt your data. [ ERROR ] Failed to execute stage 'Setup validation': Hosted Engine setup detected, but Global Maintenance is not set. This is not a hosted engine, so why does it think so? How can I get it past this step? Hi Sverker, could you please share the output of: sudo -u postgres psql engine -c "SELECT vm_guid, run_on_vds FROM vms WHERE vm_name in (select 'HostedEngine' union select option_value from vdc_options where option_name='HostedEngineVmName')" thanks Best regards /Sverker ___ Users mailing list Users@ovirt.org <mailto:Users@ovirt.org> http://lists.ovirt.org/mailman/listinfo/users <http://lists.ovirt.org/mailman/listinfo/users> ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[ovirt-users] engine upgrade 3.5 -> 4.1.1
Hi I upgraded an old ovirt installation from 3.5 to latest version, but end up with this error message when running engine-setup: [ INFO ] Stage: Setup validation [ ERROR ] It seems that you are running your engine inside of the hosted-engine VM and are not in "Global Maintenance" mode. In that case you should put the system into the "Global Maintenance" mode before running engine-setup, or the hosted-engine HA agent might kill the machine, which might corrupt your data. [ ERROR ] Failed to execute stage 'Setup validation': Hosted Engine setup detected, but Global Maintenance is not set. This is not a hosted engine, so why does it think so? How can I get it past this step? Best regards /Sverker ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
I've followed the instructions to best effort, so hopefully it's right.. Den 2017-01-13 kl. 10:31, skrev Marcin Mirecki: Please push the patch into: https://gerrit.ovirt.org/ovirt-provider-ovn (let me know if you need some directions) - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Monday, January 9, 2017 1:45:37 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Ok, found it. The issue is right here: There are two elements for virtualport, the first without id and the second with. On h2 I had fixed this which was the patch I posted earlier although I switched back to use br-int after understanding that was the correct way. When that hook was copied to h1 the port gets attached fine. Patch with updated testcase attached. /Sverker Den 2017-01-09 kl. 10:41, skrev Sverker Abrahamsson: This is the content of vdsm.log on h1 at this time: 2017-01-06 20:54:12,636 INFO (jsonrpc/7) [jsonrpc.JsonRpcServer] RPC call VM.create succeeded in 0.01 seconds (__init__:515) 2017-01-06 20:54:12,636 INFO (vm/6dd5291e) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') VM wrapper has started (vm:1901) 2017-01-06 20:54:12,636 INFO (vm/6dd5291e) [vds] prepared volume path: /rhev/data-center/mnt/h2-int.limetransit.com:_var_lib_exports_iso/1d49c4bc-0fec-4503-a583-d476fa3a370d/images/----/CentOS-7-x86_64-NetInstall-1611.iso (clientIF:374) 2017-01-06 20:54:12,743 INFO (vm/6dd5291e) [root] (hooks:108) 2017-01-06 20:54:12,847 INFO (vm/6dd5291e) [root] (hooks:108) 2017-01-06 20:54:12,863 INFO (vm/6dd5291e) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') http://ovirt.org/vm/tune/1.0"; type="kvm"> CentOS7_3 6dd5291e-6556-4d29-8b4e-ea896e627645 1048576 1048576 4294967296 16 hvm oVirt oVirt Node 7-3.1611.el7.centos 62f1adff-b29e-4a7c-abba-c2c4c73248c6 6dd5291e-6556-4d29-8b4e-ea896e627645 SandyBridge (vm:1988) 2017-01-06 20:54:13,046 INFO (libvirt/events) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') CPU running: onResume (vm:4863) 2017-01-06 20:54:13,058 INFO (vm/6dd5291e) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') Starting connection (guestagent:245) 2017-01-06 20:54:13,060 INFO (vm/6dd5291e) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') CPU running: domain initialization (vm:4863) 2017-01-06 20:54:15,154 INFO (jsonrpc/6) [jsonrpc.JsonRpcServer] RPC call Host.getVMFullList succeeded in 0.01 seconds (__init__:515) 2017-01-06 20:54:17,571 INFO (periodic/2) [dispatcher] Run and protect: getVolumeSize(sdUUID=u'2ee54fb8-48f2-4576-8cff-f2346504b08b', spUUID=u'584ebd64-0268-0193-025b-038e', imgUUID=u'5a3aae57-ffe0-4a3b-aa87-8461669db7f9', volUUID=u'b6a88789-fcb1-4d3e-911b-2a4d3b6c69c7', options=None) (logUtils:49) 2017-01-06 20:54:17,573 INFO (periodic/2) [dispatcher] Run and protect: getVolumeSize, Return response: {'truesize': '1859723264', 'apparentsize': '21474836480'} (logUtils:52) 2017-01-06 20:54:21,211 INFO (periodic/2) [dispatcher] Run and protect: repoStats(options=None) (logUtils:49) 2017-01-06 20:54:21,212 INFO (periodic/2) [dispatcher] Run and protect: repoStats, Return response: {u'2ee54fb8-48f2-4576-8cff-f2346504b08b': {'code': 0, 'actual': True, 'version': 3, 'acquired': True, 'delay': '0.000936552', 'lastCheck': '1.4', 'valid': True}, u'1d49c4bc-0fec-4503-a583-d476fa3a370d': {'code': 0, 'actual': True, 'version': 0, 'acquired': True, 'delay': '0.000960248', 'lastCheck': '1.4&
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Ok, found it. The issue is right here: interfaceid="912cba79-982e-4a87-868e-241fedccb59a" /> There are two elements for virtualport, the first without id and the second with. On h2 I had fixed this which was the patch I posted earlier although I switched back to use br-int after understanding that was the correct way. When that hook was copied to h1 the port gets attached fine. Patch with updated testcase attached. /Sverker Den 2017-01-09 kl. 10:41, skrev Sverker Abrahamsson: This is the content of vdsm.log on h1 at this time: 2017-01-06 20:54:12,636 INFO (jsonrpc/7) [jsonrpc.JsonRpcServer] RPC call VM.create succeeded in 0.01 seconds (__init__:515) 2017-01-06 20:54:12,636 INFO (vm/6dd5291e) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') VM wrapper has started (vm:1901) 2017-01-06 20:54:12,636 INFO (vm/6dd5291e) [vds] prepared volume path: /rhev/data-center/mnt/h2-int.limetransit.com:_var_lib_exports_iso/1d49c4bc-0fec-4503-a583-d476fa3a370d/images/----/CentOS-7-x86_64-NetInstall-1611.iso (clientIF:374) 2017-01-06 20:54:12,743 INFO (vm/6dd5291e) [root] (hooks:108) 2017-01-06 20:54:12,847 INFO (vm/6dd5291e) [root] (hooks:108) 2017-01-06 20:54:12,863 INFO (vm/6dd5291e) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') encoding='UTF-8'?> http://ovirt.org/vm/tune/1.0"; type="kvm"> CentOS7_3 6dd5291e-6556-4d29-8b4e-ea896e627645 1048576 1048576 4294967296 16 path="/var/lib/libvirt/qemu/channels/6dd5291e-6556-4d29-8b4e-ea896e627645.com.redhat.rhevm.vdsm" /> path="/var/lib/libvirt/qemu/channels/6dd5291e-6556-4d29-8b4e-ea896e627645.org.qemu.guest_agent.0" /> vram="32768" /> passwdValidTo="1970-01-01T00:00:01" port="-1" tlsPort="-1" type="spice"> interfaceid="912cba79-982e-4a87-868e-241fedccb59a" /> file="/rhev/data-center/mnt/h2-int.limetransit.com:_var_lib_exports_iso/1d49c4bc-0fec-4503-a583-d476fa3a370d/images/----/CentOS-7-x86_64-NetInstall-1611.iso" startupPolicy="optional" /> hvm oVirt oVirt Node 7-3.1611.el7.centos name="serial">62f1adff-b29e-4a7c-abba-c2c4c73248c6 name="uuid">6dd5291e-6556-4d29-8b4e-ea896e627645 SandyBridge (vm:1988) 2017-01-06 20:54:13,046 INFO (libvirt/events) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') CPU running: onResume (vm:4863) 2017-01-06 20:54:13,058 INFO (vm/6dd5291e) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') Starting connection (guestagent:245) 2017-01-06 20:54:13,060 INFO (vm/6dd5291e) [virt.vm] (vmId='6dd5291e-6556-4d29-8b4e-ea896e627645') CPU running: domain initialization (vm:4863) 2017-01-06 20:54:15,154 INFO (jsonrpc/6) [jsonrpc.JsonRpcServer] RPC call Host.getVMFullList succeeded in 0.01 seconds (__init__:515) 2017-01-06 20:54:17,571 INFO (periodic/2) [dispatcher] Run and protect: getVolumeSize(sdUUID=u'2ee54fb8-48f2-4576-8cff-f2346504b08b', spUUID=u'584ebd64-0268-0193-025b-038e', imgUUID=u'5a3aae57-ffe0-4a3b-aa87-8461669db7f9', volUUID=u'b6a88789-fcb1-4d3e-911b-2a4d3b6c69c7', options=None) (logUtils:49) 2017-01-06 20:54:17,573 INFO (periodic/2) [dispatcher] Run and protect: getVolumeSize, Return response: {'truesize': '1859723264', 'apparentsize': '21474836480'} (logUtils:52) 2017-01-06 20:54:21,211 INFO (periodic/2) [dispatcher] Run and protect: repoStats(options=None) (logUtils:49) 2017-01-06 20:54:21,212 INFO (periodic/2) [dispatcher] Run and protect: repoStats, Return response: {u'2ee54fb8-48f2-4576-8cff-f2346504b08b': {'code': 0, 'actual': True, 'version': 3, 'acquired': True, 'delay': '0.000936552', 'lastCheck': '1.4
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
tils:49) 2017-01-06 20:54:24,918 INFO (jsonrpc/0) [dispatcher] Run and protect: repoStats, Return response: {u'2ee54fb8-48f2-4576-8cff-f2346504b08b': {'code': 0, 'actual': True, 'version': 3, 'acquired': True, 'delay': '0.000936552', 'lastCheck': '5.1', 'valid': True}, u'1d49c4bc-0fec-4503-a583-d476fa3a370d': {'code': 0, 'actual': True, 'version': 0, 'acquired': True, 'delay': '0.000960248', 'lastCheck': '2.1', 'valid': True}} (logUtils:52) 2017-01-06 20:54:24,924 INFO (jsonrpc/0) [jsonrpc.JsonRpcServer] RPC call Host.getStats succeeded in 0.01 seconds (__init__:515) Vdsm and the OVN driver must have been called as the port IS created, but with the wrong id. I don't find the faulty id in vdsm.log neither, the xml above have the correct id. /Sverker Den 2017-01-09 kl. 10:06, skrev Marcin Mirecki: The port is set up on the host by the ovirt-provider-ovn-driver. The driver is invoked by the vdsm hook whenever any operation on the port is done. Please ensure that this is installed properly. You can check the vdsm log (/var/log/vdsm/vdsm.log) to see if the hook was executed properly. - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Friday, January 6, 2017 9:00:26 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network I created a new VM in the ui and assigned it to host h1. In /var/log/ovirt-provider-ovn.log I get the following: 2017-01-06 20:54:11,940 Request: GET : /v2.0/ports 2017-01-06 20:54:11,940 Connecting to remote ovn database: tcp:127.0.0.1:6641 2017-01-06 20:54:12,157 Connected (number of retries: 2) 2017-01-06 20:54:12,158 Response code: 200 2017-01-06 20:54:12,158 Response body: {"ports": [{"name": "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873", "network_id": "e53554cf-e553-40a1-8d22-9c8d95ec0601", "device_owner": "oVirt", "mac_address": "00:1a:4a:16:01:51", "id": "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873", "device_id": "40cd7328-d575-4c3d-b656-9ef9bacc0078"}, {"name": "92f6d3c8-68b3-4986-9c09-60bee04644b5", "network_id": "e53554cf-e553-40a1-8d22-9c8d95ec0601", "device_owner": "oVirt", "mac_address": "00:1a:4a:16:01:52", "id": "92f6d3c8-68b3-4986-9c09-60bee04644b5", "device_id": "4baefa8c-3822-4de0-9cd0-1d025bab7844"}]} 2017-01-06 20:54:12,160 Request: SHOW : /v2.0/networks/e53554cf-e553-40a1-8d22-9c8d95ec0601 2017-01-06 20:54:12,160 Connecting to remote ovn database: tcp:127.0.0.1:6641 2017-01-06 20:54:12,377 Connected (number of retries: 2) 2017-01-06 20:54:12,378 Response code: 200 2017-01-06 20:54:12,378 Response body: {"network": {"id": "e53554cf-e553-40a1-8d22-9c8d95ec0601", "name": "ovirtbridge"}} 2017-01-06 20:54:12,380 Request: POST : /v2.0/ports 2017-01-06 20:54:12,380 Request body: { "port" : { "name" : "nic1", "binding:host_id" : "h1.limetransit.com", "admin_state_up" : true, "device_id" : "e8553a88-05f0-401d-8b9b-5fff77f7bbbe", "device_owner" : "oVirt", "mac_address" : "00:1a:4a:16:01:54", "network_id" : "e53554cf-e553-40a1-8d22-9c8d95ec0601" } } 2017-01-06 20:54:12,380 Connecting to remote ovn database: tcp:127.0.0.1:6641 2017-01-06 20:54:12,610 Connected (number of retries: 2) 2017-01-06 20:54:12,614 Response code: 200 2017-01-06 20:54:12,614 Response body: {"port": {"name": "912cba79-982e-4a87-868e-241fedccb59a", "network_id": "e53554cf-e553-40a1-8d22-9c8d95ec0601", "device_owner": "oVirt", "mac_address": "00:1a:4a:16:01:54", "id": "912cba79-982e-4a87-868e-241fedccb59a", "device_id": "e8553a88-05f0-401d-8b9b-5fff77f7bbbe"}} h1:/var/log/messages Jan 6 20:54:12 h1 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl --timeout=5 -- --if-exists del-port vnet1 -- add-port br-int vnet1 -- set Interface vnet1 "external-ids:attached-mac=\"00:1a:4a:16:01:54\"" -- set Interface vnet1 "external-ids:iface-id=\"20388407-0f76-41d8-97aa-8e2b5978f908\"" -- set Interface vnet1 "external-ids:vm-id=\"6dd5291e-6556-4d29-8b4e-ea896e627645\"" -- set Interface vnet1 external-ids:iface-status=active [root@h2 ~]# ovn-nbctl show switch e53554cf-e553-40a1-8d22-9c8d95ec0601 (ovirtbridge)
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
I created a new VM in the ui and assigned it to host h1. In /var/log/ovirt-provider-ovn.log I get the following: 2017-01-06 20:54:11,940 Request: GET : /v2.0/ports 2017-01-06 20:54:11,940 Connecting to remote ovn database: tcp:127.0.0.1:6641 2017-01-06 20:54:12,157 Connected (number of retries: 2) 2017-01-06 20:54:12,158 Response code: 200 2017-01-06 20:54:12,158 Response body: {"ports": [{"name": "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873", "network_id": "e53554cf-e553-40a1-8d22-9c8d95ec0601", "device_owner": "oVirt", "mac_address": "00:1a:4a:16:01:51", "id": "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873", "device_id": "40cd7328-d575-4c3d-b656-9ef9bacc0078"}, {"name": "92f6d3c8-68b3-4986-9c09-60bee04644b5", "network_id": "e53554cf-e553-40a1-8d22-9c8d95ec0601", "device_owner": "oVirt", "mac_address": "00:1a:4a:16:01:52", "id": "92f6d3c8-68b3-4986-9c09-60bee04644b5", "device_id": "4baefa8c-3822-4de0-9cd0-1d025bab7844"}]} 2017-01-06 20:54:12,160 Request: SHOW : /v2.0/networks/e53554cf-e553-40a1-8d22-9c8d95ec0601 2017-01-06 20:54:12,160 Connecting to remote ovn database: tcp:127.0.0.1:6641 2017-01-06 20:54:12,377 Connected (number of retries: 2) 2017-01-06 20:54:12,378 Response code: 200 2017-01-06 20:54:12,378 Response body: {"network": {"id": "e53554cf-e553-40a1-8d22-9c8d95ec0601", "name": "ovirtbridge"}} 2017-01-06 20:54:12,380 Request: POST : /v2.0/ports 2017-01-06 20:54:12,380 Request body: { "port" : { "name" : "nic1", "binding:host_id" : "h1.limetransit.com", "admin_state_up" : true, "device_id" : "e8553a88-05f0-401d-8b9b-5fff77f7bbbe", "device_owner" : "oVirt", "mac_address" : "00:1a:4a:16:01:54", "network_id" : "e53554cf-e553-40a1-8d22-9c8d95ec0601" } } 2017-01-06 20:54:12,380 Connecting to remote ovn database: tcp:127.0.0.1:6641 2017-01-06 20:54:12,610 Connected (number of retries: 2) 2017-01-06 20:54:12,614 Response code: 200 2017-01-06 20:54:12,614 Response body: {"port": {"name": "912cba79-982e-4a87-868e-241fedccb59a", "network_id": "e53554cf-e553-40a1-8d22-9c8d95ec0601", "device_owner": "oVirt", "mac_address": "00:1a:4a:16:01:54", "id": "912cba79-982e-4a87-868e-241fedccb59a", "device_id": "e8553a88-05f0-401d-8b9b-5fff77f7bbbe"}} h1:/var/log/messages Jan 6 20:54:12 h1 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl --timeout=5 -- --if-exists del-port vnet1 -- add-port br-int vnet1 -- set Interface vnet1 "external-ids:attached-mac=\"00:1a:4a:16:01:54\"" -- set Interface vnet1 "external-ids:iface-id=\"20388407-0f76-41d8-97aa-8e2b5978f908\"" -- set Interface vnet1 "external-ids:vm-id=\"6dd5291e-6556-4d29-8b4e-ea896e627645\"" -- set Interface vnet1 external-ids:iface-status=active [root@h2 ~]# ovn-nbctl show switch e53554cf-e553-40a1-8d22-9c8d95ec0601 (ovirtbridge) port 4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873 addresses: ["00:1a:4a:16:01:51"] port 912cba79-982e-4a87-868e-241fedccb59a addresses: ["00:1a:4a:16:01:54"] port 92f6d3c8-68b3-4986-9c09-60bee04644b5 addresses: ["00:1a:4a:16:01:52"] port ovirtbridge-port2 addresses: ["unknown"] port ovirtbridge-port1 addresses: ["unknown"] [root@h2 ~]# ovn-sbctl show Chassis "6e4dd29f-7607-48d7-8e5a-eef4c6aeefb5" hostname: "h2.limetransit.com" Encap geneve ip: "148.251.126.50" options: {csum="true"} Port_Binding "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873" Port_Binding "ovirtbridge-port1" Chassis "4f10fb04-8fb2-48d7-8a3f-ea6444c02cf9" hostname: "h1.limetransit.com" Encap geneve ip: "144.76.84.73" options: {csum="true"} Port_Binding "ovirtbridge-port2" Port_Binding "92f6d3c8-68b3-4986-9c09-60bee04644b5" I.e. same issue /Sverker Den 2017-01-06 kl. 20:49, skrev Sverker Abrahamsson: The port is created from Ovirt UI, the ovs-vsctl command below is executed when VM is started. In /var/log/ovirt-provider-ovn.log on h2 I get the following: 2017-01-06 20:19:25,452 Request: GET : /v2.0/ports 2017-01-06 20:19:25,452 Connecting to remote ovn database: tcp:127.0.0.1:6641 2017-01-06 20:19:25,670 Connected
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
6.84.73" options: {csum="true"} Port_Binding "ovirtbridge-port2" Port_Binding "92f6d3c8-68b3-4986-9c09-60bee04644b5" I don't know from where the ID 72dafda5-03c2-4bb6-bcb6-241fa5c0a1f3 comes from, doesn't show in any log other than /var/log/messages. If I do the same exercise on the same host as engine is running on then the port for the VM gets the right id and is working from beginning. /Sverker Den 2017-01-03 kl. 10:23, skrev Marcin Mirecki: How did you create this port? From the oVirt engine UI? The OVN provider creates the port when you add the port in the engine UI, it is then plugged into the ovs bridge by the VIF driver. Please attach /var/log/ovirt-provider-ovn.log - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Tuesday, January 3, 2017 2:06:22 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Found an issue with Ovirt - OVN integration. Engine and OVN central db running on host h2. Created VM to run on host h1, which is started. Ovn db state: [root@h2 env3]# ovn-nbctl show switch e53554cf-e553-40a1-8d22-9c8d95ec0601 (ovirtbridge) port 4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873 addresses: ["00:1a:4a:16:01:51"] port 92f6d3c8-68b3-4986-9c09-60bee04644b5 addresses: ["00:1a:4a:16:01:52"] port ovirtbridge-port2 addresses: ["unknown"] port ovirtbridge-port1 addresses: ["unknown"] [root@h2 env3]# ovn-sbctl show Chassis "6e4dd29f-7607-48d7-8e5a-eef4c6aeefb5" hostname: "h2.limetransit.com" Encap geneve ip: "148.251.126.50" options: {csum="true"} Port_Binding "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873" Port_Binding "ovirtbridge-port1" Chassis "4f10fb04-8fb2-48d7-8a3f-ea6444c02cf9" hostname: "h1.limetransit.com" Encap geneve ip: "144.76.84.73" options: {csum="true"} Port_Binding "ovirtbridge-port2" Port 92f6d3c8-68b3-4986-9c09-60bee04644b5 is for the new VM which is started on h1, but it is not assigned to that chassis. The reason is that on h1 the port on br-int is created like this: ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:52\"" -- set Interface vnet0 "external-ids:iface-id=\"35bcbe31-2c7e-4d97-add9-ce150eeb2f11\"" -- set Interface vnet0 "external-ids:vm-id=\"4d0c134a-11a0-40f4-b2fb-c13c17c7251c\"" -- set Interface vnet0 external-ids:iface-status=active I.e. the extrernal id of interface is wrong. When I manually change to the right id like this the port works fine: ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:52\"" -- set Interface vnet0 "external-ids:iface-id=\"92f6d3c8-68b3-4986-9c09-60bee04644b5\"" -- set Interface vnet0 "external-ids:vm-id=\"4d0c134a-11a0-40f4-b2fb-c13c17c7251c\"" -- set Interface vnet0 external-ids:iface-status=active sb db after correcting the port: Chassis "6e4dd29f-7607-48d7-8e5a-eef4c6aeefb5" hostname: "h2.limetransit.com" Encap geneve ip: "148.251.126.50" options: {csum="true"} Port_Binding "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873" Port_Binding "ovirtbridge-port1" Chassis "4f10fb04-8fb2-48d7-8a3f-ea6444c02cf9" hostname: "h1.limetransit.com" Encap geneve ip: "144.76.84.73" options: {csum="true"} Port_Binding "ovirtbridge-port2" Port_Binding "92f6d3c8-68b3-4986-9c09-60bee04644b5" I don't know from where the faulty id comes from, it's not in any logs. In the domain xml as printed in vdsm.log the id is correct: Where is the ovs-vsctl command line built for this call? /Sverker Den 2017-01-02 kl. 13:40, skrev Sverker Abrahamsson: Got it to work now by following the env8 example in OVN tutorial, where a port is added with type l2gateway. Not sure how that is different from the localnet variant, but didn't suceed in getting that one working. Now I'm able to ping and telnet over the tunnel, but not ssh even when the port is answering on telnet. Neither does nfs traffic work even though mount did. Suspecting MTU issue. I did notice that ovn-controller starts too
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Found an issue with Ovirt - OVN integration. Engine and OVN central db running on host h2. Created VM to run on host h1, which is started. Ovn db state: [root@h2 env3]# ovn-nbctl show switch e53554cf-e553-40a1-8d22-9c8d95ec0601 (ovirtbridge) port 4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873 addresses: ["00:1a:4a:16:01:51"] port 92f6d3c8-68b3-4986-9c09-60bee04644b5 addresses: ["00:1a:4a:16:01:52"] port ovirtbridge-port2 addresses: ["unknown"] port ovirtbridge-port1 addresses: ["unknown"] [root@h2 env3]# ovn-sbctl show Chassis "6e4dd29f-7607-48d7-8e5a-eef4c6aeefb5" hostname: "h2.limetransit.com" Encap geneve ip: "148.251.126.50" options: {csum="true"} Port_Binding "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873" Port_Binding "ovirtbridge-port1" Chassis "4f10fb04-8fb2-48d7-8a3f-ea6444c02cf9" hostname: "h1.limetransit.com" Encap geneve ip: "144.76.84.73" options: {csum="true"} Port_Binding "ovirtbridge-port2" Port 92f6d3c8-68b3-4986-9c09-60bee04644b5 is for the new VM which is started on h1, but it is not assigned to that chassis. The reason is that on h1 the port on br-int is created like this: ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:52\"" -- set Interface vnet0 "external-ids:iface-id=\"35bcbe31-2c7e-4d97-add9-ce150eeb2f11\"" -- set Interface vnet0 "external-ids:vm-id=\"4d0c134a-11a0-40f4-b2fb-c13c17c7251c\"" -- set Interface vnet0 external-ids:iface-status=active I.e. the extrernal id of interface is wrong. When I manually change to the right id like this the port works fine: ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:52\"" -- set Interface vnet0 "external-ids:iface-id=\"92f6d3c8-68b3-4986-9c09-60bee04644b5\"" -- set Interface vnet0 "external-ids:vm-id=\"4d0c134a-11a0-40f4-b2fb-c13c17c7251c\"" -- set Interface vnet0 external-ids:iface-status=active sb db after correcting the port: Chassis "6e4dd29f-7607-48d7-8e5a-eef4c6aeefb5" hostname: "h2.limetransit.com" Encap geneve ip: "148.251.126.50" options: {csum="true"} Port_Binding "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873" Port_Binding "ovirtbridge-port1" Chassis "4f10fb04-8fb2-48d7-8a3f-ea6444c02cf9" hostname: "h1.limetransit.com" Encap geneve ip: "144.76.84.73" options: {csum="true"} Port_Binding "ovirtbridge-port2" Port_Binding "92f6d3c8-68b3-4986-9c09-60bee04644b5" I don't know from where the faulty id comes from, it's not in any logs. In the domain xml as printed in vdsm.log the id is correct: interfaceid="92f6d3c8-68b3-4986-9c09-60bee04644b5" /> Where is the ovs-vsctl command line built for this call? /Sverker Den 2017-01-02 kl. 13:40, skrev Sverker Abrahamsson: Got it to work now by following the env8 example in OVN tutorial, where a port is added with type l2gateway. Not sure how that is different from the localnet variant, but didn't suceed in getting that one working. Now I'm able to ping and telnet over the tunnel, but not ssh even when the port is answering on telnet. Neither does nfs traffic work even though mount did. Suspecting MTU issue. I did notice that ovn-controller starts too early, before network interfaces are established and hence can't reach the db. As these is a purely OVS/OVN issue I'll ask about it on their mailing list. Getting back to the original issue with Ovirt, I've now added the second host h1 to ovirt-engine. Had to do the same as with h2 to create a dummy ovirtmgmt network but configured access via the public IP. My firewall settings was replaced with iptables config and vdsm.conf was overwritten when engine was set up, so those had to be manually restored. It would be preferable if it would be possible to configure ovirt-engine that it does not "own" the host and instead comply with the settings it has instead of enforcing it's own view.. Apart from that it seems the second host works, although I need to resolve the traffic issue over the OVS tunnel. /Sverker Den 2017-01-02 kl. 01:13, skrev Sverker Abrahamsson: 1. That is not possible as ovirt (or vdsm) will rewrite the network configu
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Got it to work now by following the env8 example in OVN tutorial, where a port is added with type l2gateway. Not sure how that is different from the localnet variant, but didn't suceed in getting that one working. Now I'm able to ping and telnet over the tunnel, but not ssh even when the port is answering on telnet. Neither does nfs traffic work even though mount did. Suspecting MTU issue. I did notice that ovn-controller starts too early, before network interfaces are established and hence can't reach the db. As these is a purely OVS/OVN issue I'll ask about it on their mailing list. Getting back to the original issue with Ovirt, I've now added the second host h1 to ovirt-engine. Had to do the same as with h2 to create a dummy ovirtmgmt network but configured access via the public IP. My firewall settings was replaced with iptables config and vdsm.conf was overwritten when engine was set up, so those had to be manually restored. It would be preferable if it would be possible to configure ovirt-engine that it does not "own" the host and instead comply with the settings it has instead of enforcing it's own view.. Apart from that it seems the second host works, although I need to resolve the traffic issue over the OVS tunnel. /Sverker Den 2017-01-02 kl. 01:13, skrev Sverker Abrahamsson: 1. That is not possible as ovirt (or vdsm) will rewrite the network configuration to a non-working state. That is why I've set that if as hidden to vdsm and is why I'm keen on getting OVS/OVN to work 2. I've been reading the doc for OVN and starting to connect the dots, which is not trivial as it is complex. Some insights reached: First step is the OVN database, installed by openvswitch-ovn-central, which I currently have running on h2 host. The 'ovn-nbctl' and 'ovn-sbctl' commands are only possible to execute on a database node. Two ip's are given to 'vdsm-tool ovn-config ip>' as arguments, where is how this OVN node reaches the database and is the ip to which other OVN nodes sets up a tunnel to this node. I.e. it is not for creating a tunnel to the database which I thought first from the description in blog post. The tunnel between OVN nodes is of type geneve which is a UDP based protocol but I have not been able to find anywhere which port is used so that I can open it in firewalld. I have added OVN on another host, called h1, and connected it to the db. I see there is traffic to the db port, but I don't see any geneve traffic between the nodes. Ovirt is now able to create it's vnet0 interface on the br-int ovs bridge, but then I run into the next issue. How do I create a connection from the logical switch to the physical host? I need that to a) get a connection out to the internet through a masqueraded if or ipv6 and b) be able to run a dhcp server to give ip's to the VM's. /Sverker Den 2016-12-30 kl. 18:05, skrev Marcin Mirecki: 1. Why not use your physical nic for ovirtmgmt then? 2. "ovn-nbctl ls-add" does not add a bridge, but a logical switch. br-int is an internal OVN implementation detail, which the user should not care about. What you see in the ovirt UI are logical networks. They are implemented as OVN logical switches in case of the OVN provider. Please look at: http://www.ovirt.org/blog/2016/11/ovirt-provider-ovn/ You can get the latest rpms from here: http://resources.ovirt.org/repos/ovirt/experimental/master/ovirt-provider-ovn_fc24_46/rpm/fc24/noarch/ - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Friday, December 30, 2016 4:25:58 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network 1. No, I did not want to put the ovirtmgmt bridge on my physical nic as it always messed up the network config making the host unreachable. I have put a ovs bridge on this nic which I will use to make tunnels when I add other hosts. Maybe br-int will be used for that instead, will see when I get that far. As it is now I have a dummy if for ovirtmgmt bridge but this will probably not work when I add other hosts as that bridge cannot connect to the other hosts. I'm considering keeping this just as a dummy to keep ovirt engine satisfied while the actual communication will happen over OVN/OVS bridges and tunnels. 2. On https://www.ovirt.org//develop/release-management/features/ovirt-ovn-provider/ there is instructions how to add an OVS bridge to OVN with |ovn-nbctl ls-add |. If you want to use br-int then it makes sense to make that bridge visible in ovirt webui under networks so that it can be selected for VM's. It quite doesn't make sense to me that I can select other network for my VM but then that setting is not used when setting up the network. /Sverker Den 2016-12-30 kl. 15:34, skrev
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
1. That is not possible as ovirt (or vdsm) will rewrite the network configuration to a non-working state. That is why I've set that if as hidden to vdsm and is why I'm keen on getting OVS/OVN to work 2. I've been reading the doc for OVN and starting to connect the dots, which is not trivial as it is complex. Some insights reached: First step is the OVN database, installed by openvswitch-ovn-central, which I currently have running on h2 host. The 'ovn-nbctl' and 'ovn-sbctl' commands are only possible to execute on a database node. Two ip's are given to 'vdsm-tool ovn-config ip>' as arguments, where is how this OVN node reaches the database and is the ip to which other OVN nodes sets up a tunnel to this node. I.e. it is not for creating a tunnel to the database which I thought first from the description in blog post. The tunnel between OVN nodes is of type geneve which is a UDP based protocol but I have not been able to find anywhere which port is used so that I can open it in firewalld. I have added OVN on another host, called h1, and connected it to the db. I see there is traffic to the db port, but I don't see any geneve traffic between the nodes. Ovirt is now able to create it's vnet0 interface on the br-int ovs bridge, but then I run into the next issue. How do I create a connection from the logical switch to the physical host? I need that to a) get a connection out to the internet through a masqueraded if or ipv6 and b) be able to run a dhcp server to give ip's to the VM's. /Sverker Den 2016-12-30 kl. 18:05, skrev Marcin Mirecki: 1. Why not use your physical nic for ovirtmgmt then? 2. "ovn-nbctl ls-add" does not add a bridge, but a logical switch. br-int is an internal OVN implementation detail, which the user should not care about. What you see in the ovirt UI are logical networks. They are implemented as OVN logical switches in case of the OVN provider. Please look at: http://www.ovirt.org/blog/2016/11/ovirt-provider-ovn/ You can get the latest rpms from here: http://resources.ovirt.org/repos/ovirt/experimental/master/ovirt-provider-ovn_fc24_46/rpm/fc24/noarch/ - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Friday, December 30, 2016 4:25:58 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network 1. No, I did not want to put the ovirtmgmt bridge on my physical nic as it always messed up the network config making the host unreachable. I have put a ovs bridge on this nic which I will use to make tunnels when I add other hosts. Maybe br-int will be used for that instead, will see when I get that far. As it is now I have a dummy if for ovirtmgmt bridge but this will probably not work when I add other hosts as that bridge cannot connect to the other hosts. I'm considering keeping this just as a dummy to keep ovirt engine satisfied while the actual communication will happen over OVN/OVS bridges and tunnels. 2. On https://www.ovirt.org//develop/release-management/features/ovirt-ovn-provider/ there is instructions how to add an OVS bridge to OVN with |ovn-nbctl ls-add |. If you want to use br-int then it makes sense to make that bridge visible in ovirt webui under networks so that it can be selected for VM's. It quite doesn't make sense to me that I can select other network for my VM but then that setting is not used when setting up the network. /Sverker Den 2016-12-30 kl. 15:34, skrev Marcin Mirecki: Hi, The OVN provider does not require you to add any bridges manually. As I understand we were dealing with two problems: 1. You only had one physical nic and wanted to put a bridge on it, attaching the management network to the bridge. This was the reason for creating the bridge (the recommended setup would be to used a separate physical nic for the management network). This bridge has nothing to do with the OVN bridge. 2. OVN - you want to use OVN on this system. For this you have to install OVN on your hosts. This should create the br-int bridge, which are then used by the OVN provider. This br-int bridge must be configured to connect to other hosts using the geneve tunnels. In both cases the systems will not be aware of any bridges you create. They need a nic (be it physical or virtual) to connect to other system. Usually this is the physical nic. In your case you decided to put a bridge on the physical nic, and give oVirt a virtual nic attached to this bridge. This works, but keep in mind that the bridge you have introduced is outside of oVirt's (and OVN) control (and as such is not supported). What is the purpose of adding my bridges to Ovirt through the external provider and configure them on my VM I am not quite sure I understand. The external provider (OVN provider to be specific), does
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
1. No, I did not want to put the ovirtmgmt bridge on my physical nic as it always messed up the network config making the host unreachable. I have put a ovs bridge on this nic which I will use to make tunnels when I add other hosts. Maybe br-int will be used for that instead, will see when I get that far. As it is now I have a dummy if for ovirtmgmt bridge but this will probably not work when I add other hosts as that bridge cannot connect to the other hosts. I'm considering keeping this just as a dummy to keep ovirt engine satisfied while the actual communication will happen over OVN/OVS bridges and tunnels. 2. On https://www.ovirt.org//develop/release-management/features/ovirt-ovn-provider/ there is instructions how to add an OVS bridge to OVN with |ovn-nbctl ls-add |. If you want to use br-int then it makes sense to make that bridge visible in ovirt webui under networks so that it can be selected for VM's. It quite doesn't make sense to me that I can select other network for my VM but then that setting is not used when setting up the network. /Sverker Den 2016-12-30 kl. 15:34, skrev Marcin Mirecki: Hi, The OVN provider does not require you to add any bridges manually. As I understand we were dealing with two problems: 1. You only had one physical nic and wanted to put a bridge on it, attaching the management network to the bridge. This was the reason for creating the bridge (the recommended setup would be to used a separate physical nic for the management network). This bridge has nothing to do with the OVN bridge. 2. OVN - you want to use OVN on this system. For this you have to install OVN on your hosts. This should create the br-int bridge, which are then used by the OVN provider. This br-int bridge must be configured to connect to other hosts using the geneve tunnels. In both cases the systems will not be aware of any bridges you create. They need a nic (be it physical or virtual) to connect to other system. Usually this is the physical nic. In your case you decided to put a bridge on the physical nic, and give oVirt a virtual nic attached to this bridge. This works, but keep in mind that the bridge you have introduced is outside of oVirt's (and OVN) control (and as such is not supported). What is the purpose of adding my bridges to Ovirt through the external provider and configure them on my VM I am not quite sure I understand. The external provider (OVN provider to be specific), does not add any bridges to the system. It is using the br-int bridge created by OVN. The networks created by the OVN provider are purely logical entities, implemented using the OVN br-int bridge. Marcin - Original Message ----- From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Friday, December 30, 2016 12:15:43 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi That is the logic I quite don't understand. What is the purpose of adding my bridges to Ovirt through the external provider and configure them on my VM if you are disregarding that and using br-int anyway? /Sverker Den 2016-12-30 kl. 10:53, skrev Marcin Mirecki: Sverker, br-int is the integration bridge created by default in OVN. This is the bridge we use for the OVN provider. As OVN is required to be installed, we assume that this bridge is present. Using any other ovs bridge is not supported, and will require custom code changes (such as the ones you created). The proper setup in your case would probably be to create br-int and connect this to your ovirtbridge, although I don't know the details of your env, so this is just my best guess. Marcin - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" , "Numan Siddique" Sent: Friday, December 30, 2016 1:14:50 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Even better, if the value is not hardcoded then the configured value is used. Might be that I'm missunderstanding something but this is the behaviour I expected instead of that it is using br-int. Attached is a patch which properly sets up the xml, in case there is already a virtual port there + testcode of some variants /Sverker Den 2016-12-29 kl. 22:55, skrev Sverker Abrahamsson: When I change /usr/libexec/vdsm/hooks/before_device_create/ovirt_provider_ovn_hook to instead of hardcoded to br-int use BRIDGE_NAME = 'ovirtbridge' then I get the expected behaviour and I get a working network connectivity in my VM with IP provided by dhcp. /Sverker Den 2016-12-29 kl. 22:07, skrev Sverker Abrahamsson: By default the vNic profile of my OVN bridge ovirtbridge gets a Network filter named vdsm-no-mac-spoofing. If I instead set No filter then I don't get those ebtables / iptables messages. It seems that there is some issue between ovirt/v
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Hi That is the logic I quite don't understand. What is the purpose of adding my bridges to Ovirt through the external provider and configure them on my VM if you are disregarding that and using br-int anyway? /Sverker Den 2016-12-30 kl. 10:53, skrev Marcin Mirecki: Sverker, br-int is the integration bridge created by default in OVN. This is the bridge we use for the OVN provider. As OVN is required to be installed, we assume that this bridge is present. Using any other ovs bridge is not supported, and will require custom code changes (such as the ones you created). The proper setup in your case would probably be to create br-int and connect this to your ovirtbridge, although I don't know the details of your env, so this is just my best guess. Marcin - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" , "Numan Siddique" Sent: Friday, December 30, 2016 1:14:50 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Even better, if the value is not hardcoded then the configured value is used. Might be that I'm missunderstanding something but this is the behaviour I expected instead of that it is using br-int. Attached is a patch which properly sets up the xml, in case there is already a virtual port there + testcode of some variants /Sverker Den 2016-12-29 kl. 22:55, skrev Sverker Abrahamsson: When I change /usr/libexec/vdsm/hooks/before_device_create/ovirt_provider_ovn_hook to instead of hardcoded to br-int use BRIDGE_NAME = 'ovirtbridge' then I get the expected behaviour and I get a working network connectivity in my VM with IP provided by dhcp. /Sverker Den 2016-12-29 kl. 22:07, skrev Sverker Abrahamsson: By default the vNic profile of my OVN bridge ovirtbridge gets a Network filter named vdsm-no-mac-spoofing. If I instead set No filter then I don't get those ebtables / iptables messages. It seems that there is some issue between ovirt/vdsm and firewalld, which we can put to the side for now. It is not clear for me why the port is added on br-int instead of the bridge I've assigned to the VM, which is ovirtbridge?? /Sverker Den 2016-12-29 kl. 14:20, skrev Sverker Abrahamsson: The specific command most likely fails because there is no chain named libvirt-J-vnet0, but when should that have been created? /Sverker Vidarebefordrat meddelande Ämne: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Datum: Thu, 29 Dec 2016 08:06:29 -0500 (EST) Från: Marcin Mirecki Till: Sverker Abrahamsson Kopia: Ovirt Users , Lance Richardson , Numan Siddique Let me add the OVN team. Lance, Numan, Can you please look at this? Trying to plug a vNIC results in: Dec 28 23:31:35 h2 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:51\"" -- set Interface vnet0 "external-ids:iface-id=\"e8853aac-8a75-41b0-8010-e630017dcdd8\"" -- set Interface vnet0 "external-ids:vm-id=\"b9440d60-ef5a-4e2b-83cf-081df7c09e6f\"" -- set Interface vnet0 external-ids:iface-status=active Dec 28 23:31:35 h2 kernel: device vnet0 entered promiscuous mode Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: More details below - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Thursday, December 29, 2016 1:42:11 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi Same problem still.. /Sverker Den 2016-12-29 kl. 13:34, skrev Marcin Mirecki: Hi, The tunnels are created to connect multiple OVN controllers. If there is only one, there is no need for the tunnels, so none will be created, this is the correct behavior. Does the problem still occur after setting configuring the OVN-controller? Marcin - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Thursday, December 29, 2016 11:44:32 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi The rpm packages you listed in the other mail are installed but I had not run vdsm-tool ovn-config to create tunnel as the OVN controller is on the same host. [root@h2 ~]# rpm -q openvswitch-ovn-common openvswitch-ovn-common-2.6.90-1.el7.centos.x86_64 [root@h2 ~]# rpm -q openvswitch-ovn-host openvswitch-ovn-host-2.6.90-1.el7.centos.x86_64 [root@h2 ~]# rpm -q python-openvswitch python-openvswitch-2.6.90-1.el7.centos.noarch After removing my manually created br-int and run vdsm-tool ovn-config 127.0.0.1 172.27.1.1 then I have the
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Even better, if the value is not hardcoded then the configured value is used. Might be that I'm missunderstanding something but this is the behaviour I expected instead of that it is using br-int. Attached is a patch which properly sets up the xml, in case there is already a virtual port there + testcode of some variants /Sverker Den 2016-12-29 kl. 22:55, skrev Sverker Abrahamsson: When I change /usr/libexec/vdsm/hooks/before_device_create/ovirt_provider_ovn_hook to instead of hardcoded to br-int use BRIDGE_NAME = 'ovirtbridge' then I get the expected behaviour and I get a working network connectivity in my VM with IP provided by dhcp. /Sverker Den 2016-12-29 kl. 22:07, skrev Sverker Abrahamsson: By default the vNic profile of my OVN bridge ovirtbridge gets a Network filter named vdsm-no-mac-spoofing. If I instead set No filter then I don't get those ebtables / iptables messages. It seems that there is some issue between ovirt/vdsm and firewalld, which we can put to the side for now. It is not clear for me why the port is added on br-int instead of the bridge I've assigned to the VM, which is ovirtbridge?? /Sverker Den 2016-12-29 kl. 14:20, skrev Sverker Abrahamsson: The specific command most likely fails because there is no chain named libvirt-J-vnet0, but when should that have been created? /Sverker Vidarebefordrat meddelande Ämne: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Datum: Thu, 29 Dec 2016 08:06:29 -0500 (EST) Från: Marcin Mirecki Till: Sverker Abrahamsson Kopia: Ovirt Users , Lance Richardson , Numan Siddique Let me add the OVN team. Lance, Numan, Can you please look at this? Trying to plug a vNIC results in: > >>>>>> Dec 28 23:31:35 h2 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl > >>>>>> --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- > >>>>>> set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:51\"" > >>>>>> -- set Interface vnet0 > >>>>>> "external-ids:iface-id=\"e8853aac-8a75-41b0-8010-e630017dcdd8\"" -- > >>>>>> set Interface vnet0 > >>>>>> "external-ids:vm-id=\"b9440d60-ef5a-4e2b-83cf-081df7c09e6f\"" -- set > >>>>>> Interface vnet0 external-ids:iface-status=active > >>>>>> Dec 28 23:31:35 h2 kernel: device vnet0 entered promiscuous mode > >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: > >>>>>> '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j > >>>>>> libvirt-J-vnet0' failed: > >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: More details below - Original Message - > From: "Sverker Abrahamsson" > To: "Marcin Mirecki" > Cc: "Ovirt Users" > Sent: Thursday, December 29, 2016 1:42:11 PM > Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network > > Hi > Same problem still.. > /Sverker > > Den 2016-12-29 kl. 13:34, skrev Marcin Mirecki: > > Hi, > > > > The tunnels are created to connect multiple OVN controllers. > > If there is only one, there is no need for the tunnels, so none > > will be created, this is the correct behavior. > > > > Does the problem still occur after setting configuring the OVN-controller? > > > > Marcin > > > > - Original Message - > >> From: "Sverker Abrahamsson" > >> To: "Marcin Mirecki" > >> Cc: "Ovirt Users" > >> Sent: Thursday, December 29, 2016 11:44:32 AM > >> Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt > >> network > >> > >> Hi > >> The rpm packages you listed in the other mail are installed but I had > >> not run vdsm-tool ovn-config to create tunnel as the OVN controller is > >> on the same host. > >> > >> [root@h2 ~]# rpm -q openvswitch-ovn-common > >> openvswitch-ovn-common-2.6.90-1.el7.centos.x86_64 > >> [root@h2 ~]# rpm -q openvswitch-ovn-host > >> openvswitch-ovn-host-2.6.90-1.el7.centos.x86_64 > >> [root@h2 ~]# rpm -q python-openvswitch > >> python-openvswitch-2.6.90-1.el7.centos.noarch > >> > >> After removing my manually created br-int and run > >> > >> vdsm-tool ovn-config 127.0.0.1 172.27.1.1 > >> > >> then I have the br-int but 'ip link show' does not show any link > >> 'genev_sys_' nor does 'ovs-vsctl show' any port for ovn. I ass
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
When I change /usr/libexec/vdsm/hooks/before_device_create/ovirt_provider_ovn_hook to instead of hardcoded to br-int use BRIDGE_NAME = 'ovirtbridge' then I get the expected behaviour and I get a working network connectivity in my VM with IP provided by dhcp. /Sverker Den 2016-12-29 kl. 22:07, skrev Sverker Abrahamsson: By default the vNic profile of my OVN bridge ovirtbridge gets a Network filter named vdsm-no-mac-spoofing. If I instead set No filter then I don't get those ebtables / iptables messages. It seems that there is some issue between ovirt/vdsm and firewalld, which we can put to the side for now. It is not clear for me why the port is added on br-int instead of the bridge I've assigned to the VM, which is ovirtbridge?? /Sverker Den 2016-12-29 kl. 14:20, skrev Sverker Abrahamsson: The specific command most likely fails because there is no chain named libvirt-J-vnet0, but when should that have been created? /Sverker Vidarebefordrat meddelande Ämne: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Datum: Thu, 29 Dec 2016 08:06:29 -0500 (EST) Från: Marcin Mirecki Till: Sverker Abrahamsson Kopia: Ovirt Users , Lance Richardson , Numan Siddique Let me add the OVN team. Lance, Numan, Can you please look at this? Trying to plug a vNIC results in: > >>>>>> Dec 28 23:31:35 h2 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl > >>>>>> --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- > >>>>>> set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:51\"" > >>>>>> -- set Interface vnet0 > >>>>>> "external-ids:iface-id=\"e8853aac-8a75-41b0-8010-e630017dcdd8\"" -- > >>>>>> set Interface vnet0 > >>>>>> "external-ids:vm-id=\"b9440d60-ef5a-4e2b-83cf-081df7c09e6f\"" -- set > >>>>>> Interface vnet0 external-ids:iface-status=active > >>>>>> Dec 28 23:31:35 h2 kernel: device vnet0 entered promiscuous mode > >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: > >>>>>> '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j > >>>>>> libvirt-J-vnet0' failed: > >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: More details below - Original Message - > From: "Sverker Abrahamsson" > To: "Marcin Mirecki" > Cc: "Ovirt Users" > Sent: Thursday, December 29, 2016 1:42:11 PM > Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network > > Hi > Same problem still.. > /Sverker > > Den 2016-12-29 kl. 13:34, skrev Marcin Mirecki: > > Hi, > > > > The tunnels are created to connect multiple OVN controllers. > > If there is only one, there is no need for the tunnels, so none > > will be created, this is the correct behavior. > > > > Does the problem still occur after setting configuring the OVN-controller? > > > > Marcin > > > > - Original Message - > >> From: "Sverker Abrahamsson" > >> To: "Marcin Mirecki" > >> Cc: "Ovirt Users" > >> Sent: Thursday, December 29, 2016 11:44:32 AM > >> Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt > >> network > >> > >> Hi > >> The rpm packages you listed in the other mail are installed but I had > >> not run vdsm-tool ovn-config to create tunnel as the OVN controller is > >> on the same host. > >> > >> [root@h2 ~]# rpm -q openvswitch-ovn-common > >> openvswitch-ovn-common-2.6.90-1.el7.centos.x86_64 > >> [root@h2 ~]# rpm -q openvswitch-ovn-host > >> openvswitch-ovn-host-2.6.90-1.el7.centos.x86_64 > >> [root@h2 ~]# rpm -q python-openvswitch > >> python-openvswitch-2.6.90-1.el7.centos.noarch > >> > >> After removing my manually created br-int and run > >> > >> vdsm-tool ovn-config 127.0.0.1 172.27.1.1 > >> > >> then I have the br-int but 'ip link show' does not show any link > >> 'genev_sys_' nor does 'ovs-vsctl show' any port for ovn. I assume these > >> are when there is an actual tunnel? > >> > >> [root@h2 ~]# ovs-vsctl show > >> ebb6aede-cbbc-4f4f-a88a-a9cd72b2bd23 > >> Bridge br-int > >> fail_mode: secure > >> Port br-int > >> Interface br-int > >> type: internal > >> Bridge ovirtbri
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
By default the vNic profile of my OVN bridge ovirtbridge gets a Network filter named vdsm-no-mac-spoofing. If I instead set No filter then I don't get those ebtables / iptables messages. It seems that there is some issue between ovirt/vdsm and firewalld, which we can put to the side for now. It is not clear for me why the port is added on br-int instead of the bridge I've assigned to the VM, which is ovirtbridge?? /Sverker Den 2016-12-29 kl. 14:20, skrev Sverker Abrahamsson: The specific command most likely fails because there is no chain named libvirt-J-vnet0, but when should that have been created? /Sverker Vidarebefordrat meddelande Ämne: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Datum: Thu, 29 Dec 2016 08:06:29 -0500 (EST) Från: Marcin Mirecki Till: Sverker Abrahamsson Kopia: Ovirt Users , Lance Richardson , Numan Siddique Let me add the OVN team. Lance, Numan, Can you please look at this? Trying to plug a vNIC results in: > >>>>>> Dec 28 23:31:35 h2 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl > >>>>>> --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- > >>>>>> set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:51\"" > >>>>>> -- set Interface vnet0 > >>>>>> "external-ids:iface-id=\"e8853aac-8a75-41b0-8010-e630017dcdd8\"" -- > >>>>>> set Interface vnet0 > >>>>>> "external-ids:vm-id=\"b9440d60-ef5a-4e2b-83cf-081df7c09e6f\"" -- set > >>>>>> Interface vnet0 external-ids:iface-status=active > >>>>>> Dec 28 23:31:35 h2 kernel: device vnet0 entered promiscuous mode > >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: > >>>>>> '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j > >>>>>> libvirt-J-vnet0' failed: > >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: More details below - Original Message - > From: "Sverker Abrahamsson" > To: "Marcin Mirecki" > Cc: "Ovirt Users" > Sent: Thursday, December 29, 2016 1:42:11 PM > Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network > > Hi > Same problem still.. > /Sverker > > Den 2016-12-29 kl. 13:34, skrev Marcin Mirecki: > > Hi, > > > > The tunnels are created to connect multiple OVN controllers. > > If there is only one, there is no need for the tunnels, so none > > will be created, this is the correct behavior. > > > > Does the problem still occur after setting configuring the OVN-controller? > > > > Marcin > > > > - Original Message - > >> From: "Sverker Abrahamsson" > >> To: "Marcin Mirecki" > >> Cc: "Ovirt Users" > >> Sent: Thursday, December 29, 2016 11:44:32 AM > >> Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt > >> network > >> > >> Hi > >> The rpm packages you listed in the other mail are installed but I had > >> not run vdsm-tool ovn-config to create tunnel as the OVN controller is > >> on the same host. > >> > >> [root@h2 ~]# rpm -q openvswitch-ovn-common > >> openvswitch-ovn-common-2.6.90-1.el7.centos.x86_64 > >> [root@h2 ~]# rpm -q openvswitch-ovn-host > >> openvswitch-ovn-host-2.6.90-1.el7.centos.x86_64 > >> [root@h2 ~]# rpm -q python-openvswitch > >> python-openvswitch-2.6.90-1.el7.centos.noarch > >> > >> After removing my manually created br-int and run > >> > >> vdsm-tool ovn-config 127.0.0.1 172.27.1.1 > >> > >> then I have the br-int but 'ip link show' does not show any link > >> 'genev_sys_' nor does 'ovs-vsctl show' any port for ovn. I assume these > >> are when there is an actual tunnel? > >> > >> [root@h2 ~]# ovs-vsctl show > >> ebb6aede-cbbc-4f4f-a88a-a9cd72b2bd23 > >> Bridge br-int > >> fail_mode: secure > >> Port br-int > >> Interface br-int > >> type: internal > >> Bridge ovirtbridge > >> Port ovirtbridge > >> Interface ovirtbridge > >> type: internal > >> Bridge "ovsbridge0" > >> Port "ovsbridge0" > >> Interface "ovsbridge0" > >> type:
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
e "/usr/bin/vdsm-tool", line 219, in main Dec 29 21:08:09 h2.limetransit.com vdsm-tool[3736]: return tool_command[cmd]["command"](*args) Dec 29 21:08:09 h2.limetransit.com vdsm-tool[3736]: File "/usr/lib/python2.7/site-packages/vdsm/tool/restore_nets.py", line 41, in restore_command Dec 29 21:08:09 h2.limetransit.com vdsm-tool[3736]: exec_restore(cmd) Dec 29 21:08:09 h2.limetransit.com vdsm-tool[3736]: File "/usr/lib/python2.7/site-packages/vdsm/tool/restore_nets.py", line 54, in exec_restore Dec 29 21:08:09 h2.limetransit.com vdsm-tool[3736]: raise EnvironmentError('Failed to restore the persisted networks') Dec 29 21:08:09 h2.limetransit.com vdsm-tool[3736]: EnvironmentError: Failed to restore the persisted networks So I remove the nameservers section as well. At first I could not get it to work, as vdsm-network would find the backup file /var/lib/vdsm/persistence/netconf/nets/ovirtmgmt~ but after removing that it worked fine. Not sure when nameservers and defaultroute was set in vdsm as it worked on reboot before. /Sverker Den 2016-12-29 kl. 14:20, skrev Sverker Abrahamsson: The specific command most likely fails because there is no chain named libvirt-J-vnet0, but when should that have been created? /Sverker Vidarebefordrat meddelande Ämne: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Datum: Thu, 29 Dec 2016 08:06:29 -0500 (EST) Från: Marcin Mirecki Till: Sverker Abrahamsson Kopia: Ovirt Users , Lance Richardson , Numan Siddique Let me add the OVN team. Lance, Numan, Can you please look at this? Trying to plug a vNIC results in: > >>>>>> Dec 28 23:31:35 h2 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl > >>>>>> --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- > >>>>>> set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:51\"" > >>>>>> -- set Interface vnet0 > >>>>>> "external-ids:iface-id=\"e8853aac-8a75-41b0-8010-e630017dcdd8\"" -- > >>>>>> set Interface vnet0 > >>>>>> "external-ids:vm-id=\"b9440d60-ef5a-4e2b-83cf-081df7c09e6f\"" -- set > >>>>>> Interface vnet0 external-ids:iface-status=active > >>>>>> Dec 28 23:31:35 h2 kernel: device vnet0 entered promiscuous mode > >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: > >>>>>> '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j > >>>>>> libvirt-J-vnet0' failed: > >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: More details below - Original Message - > From: "Sverker Abrahamsson" > To: "Marcin Mirecki" > Cc: "Ovirt Users" > Sent: Thursday, December 29, 2016 1:42:11 PM > Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network > > Hi > Same problem still.. > /Sverker > > Den 2016-12-29 kl. 13:34, skrev Marcin Mirecki: > > Hi, > > > > The tunnels are created to connect multiple OVN controllers. > > If there is only one, there is no need for the tunnels, so none > > will be created, this is the correct behavior. > > > > Does the problem still occur after setting configuring the OVN-controller? > > > > Marcin > > > > - Original Message - > >> From: "Sverker Abrahamsson" > >> To: "Marcin Mirecki" > >> Cc: "Ovirt Users" > >> Sent: Thursday, December 29, 2016 11:44:32 AM > >> Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt > >> network > >> > >> Hi > >> The rpm packages you listed in the other mail are installed but I had > >> not run vdsm-tool ovn-config to create tunnel as the OVN controller is > >> on the same host. > >> > >> [root@h2 ~]# rpm -q openvswitch-ovn-common > >> openvswitch-ovn-common-2.6.90-1.el7.centos.x86_64 > >> [root@h2 ~]# rpm -q openvswitch-ovn-host > >> openvswitch-ovn-host-2.6.90-1.el7.centos.x86_64 > >> [root@h2 ~]# rpm -q python-openvswitch > >> python-openvswitch-2.6.90-1.el7.centos.noarch > >> > >> After removing my manually created br-int and run > >> > >> vdsm-tool ovn-config 127.0.0.1 172.27.1.1 > >> > >> then I have the br-int but 'ip link show' does not show any link > >> 'genev_sys_' nor does 'ovs-vsctl show' any port for ovn. I assume these > >> are when there is an actual tunnel? > >> > >> [root@h2 ~]#
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
The specific command most likely fails because there is no chain named libvirt-J-vnet0, but when should that have been created? /Sverker Vidarebefordrat meddelande Ämne: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Datum: Thu, 29 Dec 2016 08:06:29 -0500 (EST) Från: Marcin Mirecki Till: Sverker Abrahamsson Kopia: Ovirt Users , Lance Richardson , Numan Siddique Let me add the OVN team. Lance, Numan, Can you please look at this? Trying to plug a vNIC results in: >>>>>> Dec 28 23:31:35 h2 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl >>>>>> --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- >>>>>> set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:51\"" >>>>>> -- set Interface vnet0 >>>>>> "external-ids:iface-id=\"e8853aac-8a75-41b0-8010-e630017dcdd8\"" -- >>>>>> set Interface vnet0 >>>>>> "external-ids:vm-id=\"b9440d60-ef5a-4e2b-83cf-081df7c09e6f\"" -- set >>>>>> Interface vnet0 external-ids:iface-status=active >>>>>> Dec 28 23:31:35 h2 kernel: device vnet0 entered promiscuous mode >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: >>>>>> '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j >>>>>> libvirt-J-vnet0' failed: >>>>>> Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: More details below - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Thursday, December 29, 2016 1:42:11 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi Same problem still.. /Sverker Den 2016-12-29 kl. 13:34, skrev Marcin Mirecki: > Hi, > > The tunnels are created to connect multiple OVN controllers. > If there is only one, there is no need for the tunnels, so none > will be created, this is the correct behavior. > > Does the problem still occur after setting configuring the OVN-controller? > > Marcin > > - Original Message - >> From: "Sverker Abrahamsson" >> To: "Marcin Mirecki" >> Cc: "Ovirt Users" >> Sent: Thursday, December 29, 2016 11:44:32 AM >> Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt >> network >> >> Hi >> The rpm packages you listed in the other mail are installed but I had >> not run vdsm-tool ovn-config to create tunnel as the OVN controller is >> on the same host. >> >> [root@h2 ~]# rpm -q openvswitch-ovn-common >> openvswitch-ovn-common-2.6.90-1.el7.centos.x86_64 >> [root@h2 ~]# rpm -q openvswitch-ovn-host >> openvswitch-ovn-host-2.6.90-1.el7.centos.x86_64 >> [root@h2 ~]# rpm -q python-openvswitch >> python-openvswitch-2.6.90-1.el7.centos.noarch >> >> After removing my manually created br-int and run >> >> vdsm-tool ovn-config 127.0.0.1 172.27.1.1 >> >> then I have the br-int but 'ip link show' does not show any link >> 'genev_sys_' nor does 'ovs-vsctl show' any port for ovn. I assume these >> are when there is an actual tunnel? >> >> [root@h2 ~]# ovs-vsctl show >> ebb6aede-cbbc-4f4f-a88a-a9cd72b2bd23 >> Bridge br-int >> fail_mode: secure >> Port br-int >> Interface br-int >> type: internal >> Bridge ovirtbridge >> Port ovirtbridge >> Interface ovirtbridge >> type: internal >> Bridge "ovsbridge0" >> Port "ovsbridge0" >> Interface "ovsbridge0" >> type: internal >> Port "eth0" >> Interface "eth0" >> ovs_version: "2.6.90" >> >> [root@h2 ~]# ip link show >> 1: lo: mtu 65536 qdisc noqueue state UNKNOWN mode >> DEFAULT qlen 1 >> link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 >> 2: eth0: mtu 1500 qdisc pfifo_fast >> master ovs-system state UP mode DEFAULT qlen 1000 >> link/ether 44:8a:5b:84:7d:b3 brd ff:ff:ff:ff:ff:ff >> 3: ovs-system: mtu 1500 qdisc noop state DOWN mode >> DEFAULT qlen 1000 >> link/ether 5a:14:cf:28:47:e2 brd ff:ff:ff:ff:ff:ff >> 4: ovsbridge0: mtu 1500 qdisc noqueue >> state UNKNOWN mode DEFAULT qlen 1000 >> link/ether 44:8a:5b:84:7d:b3 brd ff:ff:ff:ff:ff:ff >> 5: br-int: mtu 1500 qdisc noop st
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Hi Same problem still.. /Sverker Den 2016-12-29 kl. 13:34, skrev Marcin Mirecki: Hi, The tunnels are created to connect multiple OVN controllers. If there is only one, there is no need for the tunnels, so none will be created, this is the correct behavior. Does the problem still occur after setting configuring the OVN-controller? Marcin - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Thursday, December 29, 2016 11:44:32 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi The rpm packages you listed in the other mail are installed but I had not run vdsm-tool ovn-config to create tunnel as the OVN controller is on the same host. [root@h2 ~]# rpm -q openvswitch-ovn-common openvswitch-ovn-common-2.6.90-1.el7.centos.x86_64 [root@h2 ~]# rpm -q openvswitch-ovn-host openvswitch-ovn-host-2.6.90-1.el7.centos.x86_64 [root@h2 ~]# rpm -q python-openvswitch python-openvswitch-2.6.90-1.el7.centos.noarch After removing my manually created br-int and run vdsm-tool ovn-config 127.0.0.1 172.27.1.1 then I have the br-int but 'ip link show' does not show any link 'genev_sys_' nor does 'ovs-vsctl show' any port for ovn. I assume these are when there is an actual tunnel? [root@h2 ~]# ovs-vsctl show ebb6aede-cbbc-4f4f-a88a-a9cd72b2bd23 Bridge br-int fail_mode: secure Port br-int Interface br-int type: internal Bridge ovirtbridge Port ovirtbridge Interface ovirtbridge type: internal Bridge "ovsbridge0" Port "ovsbridge0" Interface "ovsbridge0" type: internal Port "eth0" Interface "eth0" ovs_version: "2.6.90" [root@h2 ~]# ip link show 1: lo: mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT qlen 1 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: eth0: mtu 1500 qdisc pfifo_fast master ovs-system state UP mode DEFAULT qlen 1000 link/ether 44:8a:5b:84:7d:b3 brd ff:ff:ff:ff:ff:ff 3: ovs-system: mtu 1500 qdisc noop state DOWN mode DEFAULT qlen 1000 link/ether 5a:14:cf:28:47:e2 brd ff:ff:ff:ff:ff:ff 4: ovsbridge0: mtu 1500 qdisc noqueue state UNKNOWN mode DEFAULT qlen 1000 link/ether 44:8a:5b:84:7d:b3 brd ff:ff:ff:ff:ff:ff 5: br-int: mtu 1500 qdisc noop state DOWN mode DEFAULT qlen 1000 link/ether 9e:b0:3a:9d:f2:4b brd ff:ff:ff:ff:ff:ff 6: ovirtbridge: mtu 1500 qdisc noqueue state UNKNOWN mode DEFAULT qlen 1000 link/ether a6:f6:e5:a4:5b:45 brd ff:ff:ff:ff:ff:ff 7: dummy0: mtu 1500 qdisc noqueue master ovirtmgmt state UNKNOWN mode DEFAULT qlen 1000 link/ether 66:e0:1c:c3:a9:d8 brd ff:ff:ff:ff:ff:ff 8: ovirtmgmt: mtu 1500 qdisc noqueue state UP mode DEFAULT qlen 1000 link/ether 66:e0:1c:c3:a9:d8 brd ff:ff:ff:ff:ff:ff Firewall settings: [root@h2 ~]# firewall-cmd --list-all-zones work target: default icmp-block-inversion: no interfaces: sources: services: dhcpv6-client ssh ports: protocols: masquerade: no forward-ports: sourceports: icmp-blocks: rich rules: drop target: DROP icmp-block-inversion: no interfaces: sources: services: ports: protocols: masquerade: no forward-ports: sourceports: icmp-blocks: rich rules: internal target: default icmp-block-inversion: no interfaces: sources: services: dhcpv6-client mdns samba-client ssh ports: protocols: masquerade: no forward-ports: sourceports: icmp-blocks: rich rules: external target: default icmp-block-inversion: no interfaces: sources: services: ssh ports: protocols: masquerade: yes forward-ports: sourceports: icmp-blocks: rich rules: trusted target: ACCEPT icmp-block-inversion: no interfaces: sources: services: ports: protocols: masquerade: no forward-ports: sourceports: icmp-blocks: rich rules: home target: default icmp-block-inversion: no interfaces: sources: services: dhcpv6-client mdns samba-client ssh ports: protocols: masquerade: no forward-ports: sourceports: icmp-blocks: rich rules: dmz target: default icmp-block-inversion: no interfaces: sources: services: ssh ports: protocols: masquerade: no forward-ports: sourceports: icmp-blocks: rich rules: public (active) target: default icmp-block-inversion: no interfaces: eth0 ovsbridge0 sources: services: dhcpv6-client ssh ports: protocols: masquerade: no forward-ports: sourceports: icmp-blocks: rich rules: block target: %%REJECT%% icmp-block-inversion: no interfaces:
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
nt dump" on the host and send me the output? Have you configured the ovn controller to connect to the OVN north? You can do it using "vdsm-tool ovn-config" or using the OVN tools directly. Please check out: https://www.ovirt.org/blog/2016/11/ovirt-provider-ovn/ for details. Also please note that the OVN provider is completely different from the neutron-openvswitch plugin. Please don't mix the two. Marcin - Original Message - From: "Marcin Mirecki" To: "Sverker Abrahamsson" Cc: "Ovirt Users" Sent: Thursday, December 29, 2016 9:27:19 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi, br-int is the OVN integration bridge, it should have been created when installing OVN. I assume you have the following packages installed on the host: openvswitch-ovn-common openvswitch-ovn-host python-openvswitch Please give me some time to look at the connectivity problem. Marcin - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Thursday, December 29, 2016 12:47:04 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network From /usr/libexec/vdsm/hooks/before_device_create/ovirt_provider_ovn_hook (installed by ovirt-provider-ovn-driver rpm): BRIDGE_NAME = 'br-int' Den 2016-12-28 kl. 23:56, skrev Sverker Abrahamsson: Googling on the message about br-int suggested adding that bridge to ovs: ovs-vsctl add-br br-int Then the VM is able to boot, but it fails to get network connectivity. Output in /var/log/messages: Dec 28 23:31:35 h2 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:51\"" -- set Interface vnet0 "external-ids:iface-id=\"e8853aac-8a75-41b0-8010-e630017dcdd8\"" -- set Interface vnet0 "external-ids:vm-id=\"b9440d60-ef5a-4e2b-83cf-081df7c09e6f\"" -- set Interface vnet0 external-ids:iface-status=active Dec 28 23:31:35 h2 kernel: device vnet0 entered promiscuous mode Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -D POSTROUTING -o vnet0 -j libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -L libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -L libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F J-vnet0-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X J-vnet0-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F J-vnet0-arp-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X J-vnet0-arp-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -D libvirt-out -m physdev --physdev-is-bridged --physdev-out vnet0 -g FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -D libvirt-out -m physdev --physdev-out vnet0 -g FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -D libvirt-in -m physdev --physdev-in vnet0 -g FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -D libvirt-host-in -m physdev --physdev-in vnet0 -g HI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -F FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -X FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -F FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -X FI-vnet0' failed: Dec 28 23:31:35 h2 firewall
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
From /usr/libexec/vdsm/hooks/before_device_create/ovirt_provider_ovn_hook (installed by ovirt-provider-ovn-driver rpm): BRIDGE_NAME = 'br-int' Den 2016-12-28 kl. 23:56, skrev Sverker Abrahamsson: Googling on the message about br-int suggested adding that bridge to ovs: ovs-vsctl add-br br-int Then the VM is able to boot, but it fails to get network connectivity. Output in /var/log/messages: Dec 28 23:31:35 h2 ovs-vsctl: ovs|1|vsctl|INFO|Called as ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int vnet0 -- set Interface vnet0 "external-ids:attached-mac=\"00:1a:4a:16:01:51\"" -- set Interface vnet0 "external-ids:iface-id=\"e8853aac-8a75-41b0-8010-e630017dcdd8\"" -- set Interface vnet0 "external-ids:vm-id=\"b9440d60-ef5a-4e2b-83cf-081df7c09e6f\"" -- set Interface vnet0 external-ids:iface-status=active Dec 28 23:31:35 h2 kernel: device vnet0 entered promiscuous mode Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -D POSTROUTING -o vnet0 -j libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -L libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -L libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X libvirt-J-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F J-vnet0-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X J-vnet0-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F J-vnet0-arp-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X J-vnet0-arp-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -D libvirt-out -m physdev --physdev-is-bridged --physdev-out vnet0 -g FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -D libvirt-out -m physdev --physdev-out vnet0 -g FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -D libvirt-in -m physdev --physdev-in vnet0 -g FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -D libvirt-host-in -m physdev --physdev-in vnet0 -g HI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -F FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -X FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -F FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -X FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -F HI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -X HI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -E FP-vnet0 FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -E FJ-vnet0 FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/iptables -w2 -w -E HJ-vnet0 HI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -D libvirt-out -m physdev --physdev-is-bridged --physdev-out vnet0 -g FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -D libvirt-out -m physdev --physdev-out vnet0 -g FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -D libvirt-in -m physdev --physdev-in vnet0 -g FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -D libvirt-host-in -m physdev --physdev-in vn
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
/sbin/ip6tables -w2 -w -X FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -F FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -X FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -F HI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -X HI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -E FP-vnet0 FO-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -E FJ-vnet0 FI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ip6tables -w2 -w -E HJ-vnet0 HI-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -D PREROUTING -i vnet0 -j libvirt-I-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -D POSTROUTING -o vnet0 -j libvirt-O-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -L libvirt-I-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -L libvirt-O-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F libvirt-I-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X libvirt-I-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F libvirt-O-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X libvirt-O-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -L libvirt-P-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -E libvirt-P-vnet0 libvirt-O-vnet0' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F I-vnet0-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X I-vnet0-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -F I-vnet0-arp-mac' failed: Dec 28 23:31:35 h2 firewalld: WARNING: COMMAND_FAILED: '/usr/sbin/ebtables --concurrent -t nat -X I-vnet0-arp-mac' failed: [root@h2 etc]# ovs-vsctl show ebb6aede-cbbc-4f4f-a88a-a9cd72b2bd23 Bridge ovirtbridge Port "ovirtport0" Interface "ovirtport0" type: internal Port ovirtbridge Interface ovirtbridge type: internal Bridge "ovsbridge0" Port "ovsbridge0" Interface "ovsbridge0" type: internal Port "eth0" Interface "eth0" Bridge br-int Port br-int Interface br-int type: internal Port "vnet0" Interface "vnet0" ovs_version: "2.6.90" Searching through the code it appears that br-int comes from neutron-openvswitch plugin ?? [root@h2 share]# rpm -qf /usr/share/otopi/plugins/ovirt-host-deploy/openstack/neutron_openvswitch.py ovirt-host-deploy-1.6.0-0.0.master.20161215101008.gitb76ad50.el7.centos.noarch /Sverker Den 2016-12-28 kl. 23:24, skrev Sverker Abrahamsson: In addition I had to add an alias to modprobe: [root@h2 modprobe.d]# cat dummy.conf alias dummy0 dummy Den 2016-12-28 kl. 23:03, skrev Sverker Abrahamsson: Hi I first tried to set device name to dummy_0, but then ifup did not succeed in creating the device unless I first did 'ip link add dummy_0 type dummy' but then it would not suceed to establish the if on reboot. Setting fake_nics = dummy0 would not work neither, but this works: fake_nics = dummy* The engine is now able to find the if and assign bridge ovirtmgmt to it. However, I then run into the next issue when starting a VM: 2016-12-28 22:28:23,897 ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (ForkJoinPool-1-worker-2) [] Correlation ID: null, Call Stack: null, Custom Event ID: -1, Message: VM CentOS7 is down with error. Exit message: Cannot get interface MTU on 'br-int': No such device. This VM has a nic on ovirtbridge, which comes from the OVN provider. /Sverker Den 2016-12-28 kl. 14:38, skrev Marcin Mirecki: Sverker, Can you try adding a vnic named veth_* or dummy_*, (or altern
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
In addition I had to add an alias to modprobe: [root@h2 modprobe.d]# cat dummy.conf alias dummy0 dummy Den 2016-12-28 kl. 23:03, skrev Sverker Abrahamsson: Hi I first tried to set device name to dummy_0, but then ifup did not succeed in creating the device unless I first did 'ip link add dummy_0 type dummy' but then it would not suceed to establish the if on reboot. Setting fake_nics = dummy0 would not work neither, but this works: fake_nics = dummy* The engine is now able to find the if and assign bridge ovirtmgmt to it. However, I then run into the next issue when starting a VM: 2016-12-28 22:28:23,897 ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (ForkJoinPool-1-worker-2) [] Correlation ID: null, Call Stack: null, Custom Event ID: -1, Message: VM CentOS7 is down with error. Exit message: Cannot get interface MTU on 'br-int': No such device. This VM has a nic on ovirtbridge, which comes from the OVN provider. /Sverker Den 2016-12-28 kl. 14:38, skrev Marcin Mirecki: Sverker, Can you try adding a vnic named veth_* or dummy_*, (or alternatively add the name of the vnic to vdsm.config fake_nics), and setup the management network using this vnic? I suppose adding the vnic you use for connecting to the engine to fake_nics should make it visible to the engine, and you should be able to use it for the setup. Marcin - Original Message - From: "Marcin Mirecki" To: "Sverker Abrahamsson" Cc: "Ovirt Users" Sent: Wednesday, December 28, 2016 12:06:26 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network I have an internal OVS bridge called ovirtbridge which has a port with IP address, but in the host network settings that port is not visible. I just verified and unfortunately the virtual ports are not visible in engine to assign a network to :( I'm afraid that the engine is not ready for such a scenario (even if it works). Please give me some time to look for a solution. - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Wednesday, December 28, 2016 11:48:24 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi Marcin Yes, that is my issue. I don't want to let ovirt/vdsm see eth0 nor ovsbridge0 since as soon as it sees them it messes up the network config so that the host will be unreachable. I have an internal OVS bridge called ovirtbridge which has a port with IP address, but in the host network settings that port is not visible. It doesn't help to name it ovirtmgmt. The engine is able to communicate with the host on the ip it has been given, it's just that it believes that it HAS to have a ovirtmgmt network which can't be on OVN. /Sverker Den 2016-12-28 kl. 10:45, skrev Marcin Mirecki: Hi Sverker, The management network is mandatory on each host. It's used by the engine to communicate with the host. Looking at your description and the exception it looks like it is missing. The error is caused by not having any network for the host (network list retrieved in InterfaceDaoImpl.getHostNetworksByCluster - which gets all the networks on nics for a host from vds_interface table in the DB). Could you maybe create a virtual nic connected to ovsbridge0 (as I understand you have no physical nic available) and use this for the management network? I then create a bridge for use with ovirt, with a private address. I'm not quite sure I understand. Is this yet another bridge connected to ovsbridge0? You could also attach the vnic for the management network here if need be. Please keep in mind that OVN has no use in setting up the management network. The OVN provider can only handle external networks, which can not be used for a management network. Marcin - Original Message - From: "Sverker Abrahamsson" To: users@ovirt.org Sent: Wednesday, December 28, 2016 12:39:59 AM Subject: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi For long time I've been looking for proper support in ovirt for Open vSwitch so I'm happy that it is moving in the right direction. However, there seems to still be a dependency on a ovirtmgmt bridge and I'm unable to move that to the OVN provider. The hosting center where I rent hw instances has a bit special network setup, so I have one physical network port with a /32 netmask and point-to-point config to router. The physical port I connect to a ovs bridge which has the public ip. Since ovirt always messes up the network config when I've tried to let it have access to the network config for the physical port, I've set eht0 and ovsbridge0 as hidden in vdsm.conf. I then create a bridge for use with ovirt, with a private address. With the OVN provider I am now able to import these into the engine and it
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Hi I first tried to set device name to dummy_0, but then ifup did not succeed in creating the device unless I first did 'ip link add dummy_0 type dummy' but then it would not suceed to establish the if on reboot. Setting fake_nics = dummy0 would not work neither, but this works: fake_nics = dummy* The engine is now able to find the if and assign bridge ovirtmgmt to it. However, I then run into the next issue when starting a VM: 2016-12-28 22:28:23,897 ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (ForkJoinPool-1-worker-2) [] Correlation ID: null, Call Stack: null, Custom Event ID: -1, Message: VM CentOS7 is down with error. Exit message: Cannot get interface MTU on 'br-int': No such device. This VM has a nic on ovirtbridge, which comes from the OVN provider. /Sverker Den 2016-12-28 kl. 14:38, skrev Marcin Mirecki: Sverker, Can you try adding a vnic named veth_* or dummy_*, (or alternatively add the name of the vnic to vdsm.config fake_nics), and setup the management network using this vnic? I suppose adding the vnic you use for connecting to the engine to fake_nics should make it visible to the engine, and you should be able to use it for the setup. Marcin - Original Message - From: "Marcin Mirecki" To: "Sverker Abrahamsson" Cc: "Ovirt Users" Sent: Wednesday, December 28, 2016 12:06:26 PM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network I have an internal OVS bridge called ovirtbridge which has a port with IP address, but in the host network settings that port is not visible. I just verified and unfortunately the virtual ports are not visible in engine to assign a network to :( I'm afraid that the engine is not ready for such a scenario (even if it works). Please give me some time to look for a solution. - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Wednesday, December 28, 2016 11:48:24 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi Marcin Yes, that is my issue. I don't want to let ovirt/vdsm see eth0 nor ovsbridge0 since as soon as it sees them it messes up the network config so that the host will be unreachable. I have an internal OVS bridge called ovirtbridge which has a port with IP address, but in the host network settings that port is not visible. It doesn't help to name it ovirtmgmt. The engine is able to communicate with the host on the ip it has been given, it's just that it believes that it HAS to have a ovirtmgmt network which can't be on OVN. /Sverker Den 2016-12-28 kl. 10:45, skrev Marcin Mirecki: Hi Sverker, The management network is mandatory on each host. It's used by the engine to communicate with the host. Looking at your description and the exception it looks like it is missing. The error is caused by not having any network for the host (network list retrieved in InterfaceDaoImpl.getHostNetworksByCluster - which gets all the networks on nics for a host from vds_interface table in the DB). Could you maybe create a virtual nic connected to ovsbridge0 (as I understand you have no physical nic available) and use this for the management network? I then create a bridge for use with ovirt, with a private address. I'm not quite sure I understand. Is this yet another bridge connected to ovsbridge0? You could also attach the vnic for the management network here if need be. Please keep in mind that OVN has no use in setting up the management network. The OVN provider can only handle external networks, which can not be used for a management network. Marcin - Original Message - From: "Sverker Abrahamsson" To: users@ovirt.org Sent: Wednesday, December 28, 2016 12:39:59 AM Subject: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi For long time I've been looking for proper support in ovirt for Open vSwitch so I'm happy that it is moving in the right direction. However, there seems to still be a dependency on a ovirtmgmt bridge and I'm unable to move that to the OVN provider. The hosting center where I rent hw instances has a bit special network setup, so I have one physical network port with a /32 netmask and point-to-point config to router. The physical port I connect to a ovs bridge which has the public ip. Since ovirt always messes up the network config when I've tried to let it have access to the network config for the physical port, I've set eht0 and ovsbridge0 as hidden in vdsm.conf. I then create a bridge for use with ovirt, with a private address. With the OVN provider I am now able to import these into the engine and it looks good. When creating a VM I can select that it will have a vNic on my OVS bridge. However, I can't start the VM as an exception is thrown in the log: 2016-12-
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
I have tried various variants to create an interface with the dummy driver, currently I have an interface dummy0 which is assigned to a legacy bridge ovirtmgmt but ovirt-engine doesn't see those neither. /Sverker Den 2016-12-28 kl. 12:06, skrev Marcin Mirecki: I have an internal OVS bridge called ovirtbridge which has a port with IP address, but in the host network settings that port is not visible. I just verified and unfortunately the virtual ports are not visible in engine to assign a network to :( I'm afraid that the engine is not ready for such a scenario (even if it works). Please give me some time to look for a solution. - Original Message - From: "Sverker Abrahamsson" To: "Marcin Mirecki" Cc: "Ovirt Users" Sent: Wednesday, December 28, 2016 11:48:24 AM Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi Marcin Yes, that is my issue. I don't want to let ovirt/vdsm see eth0 nor ovsbridge0 since as soon as it sees them it messes up the network config so that the host will be unreachable. I have an internal OVS bridge called ovirtbridge which has a port with IP address, but in the host network settings that port is not visible. It doesn't help to name it ovirtmgmt. The engine is able to communicate with the host on the ip it has been given, it's just that it believes that it HAS to have a ovirtmgmt network which can't be on OVN. /Sverker Den 2016-12-28 kl. 10:45, skrev Marcin Mirecki: Hi Sverker, The management network is mandatory on each host. It's used by the engine to communicate with the host. Looking at your description and the exception it looks like it is missing. The error is caused by not having any network for the host (network list retrieved in InterfaceDaoImpl.getHostNetworksByCluster - which gets all the networks on nics for a host from vds_interface table in the DB). Could you maybe create a virtual nic connected to ovsbridge0 (as I understand you have no physical nic available) and use this for the management network? I then create a bridge for use with ovirt, with a private address. I'm not quite sure I understand. Is this yet another bridge connected to ovsbridge0? You could also attach the vnic for the management network here if need be. Please keep in mind that OVN has no use in setting up the management network. The OVN provider can only handle external networks, which can not be used for a management network. Marcin - Original Message - From: "Sverker Abrahamsson" To: users@ovirt.org Sent: Wednesday, December 28, 2016 12:39:59 AM Subject: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi For long time I've been looking for proper support in ovirt for Open vSwitch so I'm happy that it is moving in the right direction. However, there seems to still be a dependency on a ovirtmgmt bridge and I'm unable to move that to the OVN provider. The hosting center where I rent hw instances has a bit special network setup, so I have one physical network port with a /32 netmask and point-to-point config to router. The physical port I connect to a ovs bridge which has the public ip. Since ovirt always messes up the network config when I've tried to let it have access to the network config for the physical port, I've set eht0 and ovsbridge0 as hidden in vdsm.conf. I then create a bridge for use with ovirt, with a private address. With the OVN provider I am now able to import these into the engine and it looks good. When creating a VM I can select that it will have a vNic on my OVS bridge. However, I can't start the VM as an exception is thrown in the log: 2016-12-28 00:13:33,350 ERROR [org.ovirt.engine.core.bll.RunVmCommand] (default task-5) [3c882d53] Error during ValidateFailure.: java.lang.NullPointerException at org.ovirt.engine.core.bll.scheduling.policyunits.NetworkPolicyUnit.validateRequiredNetworksAvailable(NetworkPolicyUnit.java:140) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.policyunits.NetworkPolicyUnit.filter(NetworkPolicyUnit.java:69) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.runInternalFilters(SchedulingManager.java:597) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.runFilters(SchedulingManager.java:564) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.canSchedule(SchedulingManager.java:494) [bll.jar:] at org.ovirt.engine.core.bll.validator.RunVmValidator.canRunVm(RunVmValidator.java:133) [bll.jar:] at org.ovirt.engine.core.bll.RunVmCommand.validate(RunVmCommand.java:940) [bll.jar:] at org.ovirt.engine.core.bll.CommandBase.internalValidate(CommandBase.java:886) [bll.jar:] at org.ovirt.engine.core.bll.CommandBase.validateOnly(CommandBase.java:366) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.canRunActions(PrevalidatingMultipleActionsRunner.java:113) [bll.jar:] at org.o
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Hi Marcin Yes, that is my issue. I don't want to let ovirt/vdsm see eth0 nor ovsbridge0 since as soon as it sees them it messes up the network config so that the host will be unreachable. I have an internal OVS bridge called ovirtbridge which has a port with IP address, but in the host network settings that port is not visible. It doesn't help to name it ovirtmgmt. The engine is able to communicate with the host on the ip it has been given, it's just that it believes that it HAS to have a ovirtmgmt network which can't be on OVN. /Sverker Den 2016-12-28 kl. 10:45, skrev Marcin Mirecki: Hi Sverker, The management network is mandatory on each host. It's used by the engine to communicate with the host. Looking at your description and the exception it looks like it is missing. The error is caused by not having any network for the host (network list retrieved in InterfaceDaoImpl.getHostNetworksByCluster - which gets all the networks on nics for a host from vds_interface table in the DB). Could you maybe create a virtual nic connected to ovsbridge0 (as I understand you have no physical nic available) and use this for the management network? I then create a bridge for use with ovirt, with a private address. I'm not quite sure I understand. Is this yet another bridge connected to ovsbridge0? You could also attach the vnic for the management network here if need be. Please keep in mind that OVN has no use in setting up the management network. The OVN provider can only handle external networks, which can not be used for a management network. Marcin - Original Message ----- From: "Sverker Abrahamsson" To: users@ovirt.org Sent: Wednesday, December 28, 2016 12:39:59 AM Subject: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network Hi For long time I've been looking for proper support in ovirt for Open vSwitch so I'm happy that it is moving in the right direction. However, there seems to still be a dependency on a ovirtmgmt bridge and I'm unable to move that to the OVN provider. The hosting center where I rent hw instances has a bit special network setup, so I have one physical network port with a /32 netmask and point-to-point config to router. The physical port I connect to a ovs bridge which has the public ip. Since ovirt always messes up the network config when I've tried to let it have access to the network config for the physical port, I've set eht0 and ovsbridge0 as hidden in vdsm.conf. I then create a bridge for use with ovirt, with a private address. With the OVN provider I am now able to import these into the engine and it looks good. When creating a VM I can select that it will have a vNic on my OVS bridge. However, I can't start the VM as an exception is thrown in the log: 2016-12-28 00:13:33,350 ERROR [org.ovirt.engine.core.bll.RunVmCommand] (default task-5) [3c882d53] Error during ValidateFailure.: java.lang.NullPointerException at org.ovirt.engine.core.bll.scheduling.policyunits.NetworkPolicyUnit.validateRequiredNetworksAvailable(NetworkPolicyUnit.java:140) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.policyunits.NetworkPolicyUnit.filter(NetworkPolicyUnit.java:69) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.runInternalFilters(SchedulingManager.java:597) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.runFilters(SchedulingManager.java:564) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.canSchedule(SchedulingManager.java:494) [bll.jar:] at org.ovirt.engine.core.bll.validator.RunVmValidator.canRunVm(RunVmValidator.java:133) [bll.jar:] at org.ovirt.engine.core.bll.RunVmCommand.validate(RunVmCommand.java:940) [bll.jar:] at org.ovirt.engine.core.bll.CommandBase.internalValidate(CommandBase.java:886) [bll.jar:] at org.ovirt.engine.core.bll.CommandBase.validateOnly(CommandBase.java:366) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.canRunActions(PrevalidatingMultipleActionsRunner.java:113) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.invokeCommands(PrevalidatingMultipleActionsRunner.java:99) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.execute(PrevalidatingMultipleActionsRunner.java:76) [bll.jar:] at org.ovirt.engine.core.bll.Backend.runMultipleActionsImpl(Backend.java:613) [bll.jar:] at org.ovirt.engine.core.bll.Backend.runMultipleActions(Backend.java:583) [bll.jar:] Looking at that section of code where the exception is thrown, I see that it iterates over host networks to find required networks, which I assume is ovirtmgmt. In the host network setup dialog I don't see any networks at all but it lists ovirtmgmt as required. It also list the OVN networks but these can't be statically assigned as they are added dynamically when needed, which is fine. I believe that I either need to remove ovirtmgmt network or configure
Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Hi Yaniv I have installed ovirt-provider-ovn-1.0-1.fc24.noarch.rpm and ovirt-provider-ovn-driver-1.0-1.fc24.noarch.rpm /Sverker Den 2016-12-28 kl. 09:24, skrev Yaniv Dary: Did you install the host side VIF driver? Yaniv Dary Technical Product Manager Red Hat Israel Ltd. 34 Jerusalem Road Building A, 4th floor Ra'anana, Israel 4350109 Tel : +972 (9) 7692306 8272306 Email: yd...@redhat.com <mailto:yd...@redhat.com> IRC : ydary On Wed, Dec 28, 2016 at 1:39 AM, Sverker Abrahamsson mailto:sver...@abrahamsson.com>> wrote: Hi For long time I've been looking for proper support in ovirt for Open vSwitch so I'm happy that it is moving in the right direction. However, there seems to still be a dependency on a ovirtmgmt bridge and I'm unable to move that to the OVN provider. The hosting center where I rent hw instances has a bit special network setup, so I have one physical network port with a /32 netmask and point-to-point config to router. The physical port I connect to a ovs bridge which has the public ip. Since ovirt always messes up the network config when I've tried to let it have access to the network config for the physical port, I've set eht0 and ovsbridge0 as hidden in vdsm.conf. I then create a bridge for use with ovirt, with a private address. With the OVN provider I am now able to import these into the engine and it looks good. When creating a VM I can select that it will have a vNic on my OVS bridge. However, I can't start the VM as an exception is thrown in the log: 2016-12-28 00:13:33,350 ERROR [org.ovirt.engine.core.bll.RunVmCommand] (default task-5) [3c882d53] Error during ValidateFailure.: java.lang.NullPointerException at org.ovirt.engine.core.bll.scheduling.policyunits.NetworkPolicyUnit.validateRequiredNetworksAvailable(NetworkPolicyUnit.java:140) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.policyunits.NetworkPolicyUnit.filter(NetworkPolicyUnit.java:69) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.runInternalFilters(SchedulingManager.java:597) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.runFilters(SchedulingManager.java:564) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.canSchedule(SchedulingManager.java:494) [bll.jar:] at org.ovirt.engine.core.bll.validator.RunVmValidator.canRunVm(RunVmValidator.java:133) [bll.jar:] at org.ovirt.engine.core.bll.RunVmCommand.validate(RunVmCommand.java:940) [bll.jar:] at org.ovirt.engine.core.bll.CommandBase.internalValidate(CommandBase.java:886) [bll.jar:] at org.ovirt.engine.core.bll.CommandBase.validateOnly(CommandBase.java:366) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.canRunActions(PrevalidatingMultipleActionsRunner.java:113) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.invokeCommands(PrevalidatingMultipleActionsRunner.java:99) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.execute(PrevalidatingMultipleActionsRunner.java:76) [bll.jar:] at org.ovirt.engine.core.bll.Backend.runMultipleActionsImpl(Backend.java:613) [bll.jar:] at org.ovirt.engine.core.bll.Backend.runMultipleActions(Backend.java:583) [bll.jar:] Looking at that section of code where the exception is thrown, I see that it iterates over host networks to find required networks, which I assume is ovirtmgmt. In the host network setup dialog I don't see any networks at all but it lists ovirtmgmt as required. It also list the OVN networks but these can't be statically assigned as they are added dynamically when needed, which is fine. I believe that I either need to remove ovirtmgmt network or configure that it is provided by the OVN provider, but neither is possible. Preferably it shouldn't be hardcoded which network is management and mandatory but be possible to configure. /Sverker Den 2016-12-27 kl. 17:10, skrev Marcin Mirecki: The following is a new meeting request: Subject: OVN Deep Dive for oVirt 4.1 Organizer: "Marcin Mirecki" <mailto:mmire...@redhat.com> Time: Wednesday, January 11, 2017, 4:00:00 PM - 5:00:00 PM GMT +01:00 Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna Invitees:users@ovirt.org <mailto:users@ovirt.org> *~*~*~*~*~*~*~*~*~* OVN (Open Virtual Networking) is a new network virtualization project that brings virtual networking to Open vSwitch. OVN is now availble to use as a tech preview in oVirt. It is made available using the oVirt externa
[ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network
Hi For long time I've been looking for proper support in ovirt for Open vSwitch so I'm happy that it is moving in the right direction. However, there seems to still be a dependency on a ovirtmgmt bridge and I'm unable to move that to the OVN provider. The hosting center where I rent hw instances has a bit special network setup, so I have one physical network port with a /32 netmask and point-to-point config to router. The physical port I connect to a ovs bridge which has the public ip. Since ovirt always messes up the network config when I've tried to let it have access to the network config for the physical port, I've set eht0 and ovsbridge0 as hidden in vdsm.conf. I then create a bridge for use with ovirt, with a private address. With the OVN provider I am now able to import these into the engine and it looks good. When creating a VM I can select that it will have a vNic on my OVS bridge. However, I can't start the VM as an exception is thrown in the log: 2016-12-28 00:13:33,350 ERROR [org.ovirt.engine.core.bll.RunVmCommand] (default task-5) [3c882d53] Error during ValidateFailure.: java.lang.NullPointerException at org.ovirt.engine.core.bll.scheduling.policyunits.NetworkPolicyUnit.validateRequiredNetworksAvailable(NetworkPolicyUnit.java:140) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.policyunits.NetworkPolicyUnit.filter(NetworkPolicyUnit.java:69) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.runInternalFilters(SchedulingManager.java:597) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.runFilters(SchedulingManager.java:564) [bll.jar:] at org.ovirt.engine.core.bll.scheduling.SchedulingManager.canSchedule(SchedulingManager.java:494) [bll.jar:] at org.ovirt.engine.core.bll.validator.RunVmValidator.canRunVm(RunVmValidator.java:133) [bll.jar:] at org.ovirt.engine.core.bll.RunVmCommand.validate(RunVmCommand.java:940) [bll.jar:] at org.ovirt.engine.core.bll.CommandBase.internalValidate(CommandBase.java:886) [bll.jar:] at org.ovirt.engine.core.bll.CommandBase.validateOnly(CommandBase.java:366) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.canRunActions(PrevalidatingMultipleActionsRunner.java:113) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.invokeCommands(PrevalidatingMultipleActionsRunner.java:99) [bll.jar:] at org.ovirt.engine.core.bll.PrevalidatingMultipleActionsRunner.execute(PrevalidatingMultipleActionsRunner.java:76) [bll.jar:] at org.ovirt.engine.core.bll.Backend.runMultipleActionsImpl(Backend.java:613) [bll.jar:] at org.ovirt.engine.core.bll.Backend.runMultipleActions(Backend.java:583) [bll.jar:] Looking at that section of code where the exception is thrown, I see that it iterates over host networks to find required networks, which I assume is ovirtmgmt. In the host network setup dialog I don't see any networks at all but it lists ovirtmgmt as required. It also list the OVN networks but these can't be statically assigned as they are added dynamically when needed, which is fine. I believe that I either need to remove ovirtmgmt network or configure that it is provided by the OVN provider, but neither is possible. Preferably it shouldn't be hardcoded which network is management and mandatory but be possible to configure. /Sverker Den 2016-12-27 kl. 17:10, skrev Marcin Mirecki: The following is a new meeting request: Subject: OVN Deep Dive for oVirt 4.1 Organizer: "Marcin Mirecki" Time: Wednesday, January 11, 2017, 4:00:00 PM - 5:00:00 PM GMT +01:00 Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna Invitees: users@ovirt.org *~*~*~*~*~*~*~*~*~* OVN (Open Virtual Networking) is a new network virtualization project that brings virtual networking to Open vSwitch. OVN is now availble to use as a tech preview in oVirt. It is made available using the oVirt external network providers API, which allows using external network management software inside environments managed by oVirt. The integration with OVN will allow users to take advantage of native OVS support for software defined networks. The advantages of using OVN: - ability to use large number of networks with no need for multiple host NICs or VLAN use - ability to set up new network without any changes to the physical infrastructure - easier to maintain due to centralized management - good traffic isolation due to Openflow rules - improved performance over neutron OVS plugin (https://blog.russellbryant.net/2016/12/19/comparing-openstack-neutron-ml2ovs-and-ovn-control-plane/) Session outline: - OVN overview - oVirt external network provider API overview - oVirt provider for OVN overview Session link: https://www.youtube.com/watch?v=vGeouWfKJwA Feature Page: https://www.ovirt.org//develop/release-management/features/ovirt-ovn-provider/ Blog Post: https://www.
Re: [ovirt-users] Ovirt + OpenVSwitch
I've been giving this a new try. Unfortunately the setup script for hosted engine seems to be hardcoded to create a bridge on a physical network interface. I was somewhat successful by creating a dummy ovirtmgmt bridge interface and patching bridge.py to give back a static ip. Then the engine started up but failed later in the setup due to a faulty firewall rule. I then thought that ok I'll restore the network config to the original settings, run the setup and then add the OpenVSwitch config. That failed even more miserably before I got that far.. As soon as vdsm activates it will write ifcfg files for eth0 and ovirtmgmt but with a config that doesn't work in my case as there are some additional parameters that are needed. For example it is missing the HWADDR setting in ifcfg-eth0. This is using the beta1 release of 4.0 Even if I boot rescue image and correct the network configuration it will again be overwritten by vdsm on boot. I really need to make Ovirt understand that it must NOT change any network configuration on the host, just use what is provided to it and if there is something that it doesn't like then log the settings that should be changed. How can I accomplish that? /Sverker On 2016-04-13 14:50, Dan Kenigsberg wrote: On Mon, Apr 11, 2016 at 01:24:51PM +0200, Sverker Abrahamsson wrote: Hi My usecase is that I am setting up a lab to emulate a customers production environment which has multiple vlans. As this is done in a hosting environment where I don't have any control over the switches, I've created virtual switches with OpenVSwitch which connects together a couple of physical servers which are running virtual machines to emulate their servers. Currently I'm running a few virtual machines created manually with libvirt but looking for a more user friendly way of handling them. I'm not looking at managing the network setup from Ovirt, I just want it to use the network that is present and create VM's there. In this case, you may want to define your ovs-based network in libvirt (virsh net-define bla.xml) and use vdsm-hook-extnet to let ovirt consume it. The tweaking that I did with 3.6 was to run the hosted-engine setup but where it failed I put the right answer in the file so that it would skip over that part. I got that far that it attempts to start the vm but as it doesn't give right parameters to libvirt it fails to create a port on the bridge. I'm not having much luck with the 4.0 alpha though, this is a separate machine that I set up for this testing but when I run hosted-engine --deploy it just returns without any message. I did run the cleanup script found at http://www.ovirt.org/documentation/how-to/hosted-engine/ and rebooted in between but still no luck /Sverker ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [ovirt-users] Ovirt + OpenVSwitch
Hi My usecase is that I am setting up a lab to emulate a customers production environment which has multiple vlans. As this is done in a hosting environment where I don't have any control over the switches, I've created virtual switches with OpenVSwitch which connects together a couple of physical servers which are running virtual machines to emulate their servers. Currently I'm running a few virtual machines created manually with libvirt but looking for a more user friendly way of handling them. I'm not looking at managing the network setup from Ovirt, I just want it to use the network that is present and create VM's there. The tweaking that I did with 3.6 was to run the hosted-engine setup but where it failed I put the right answer in the file so that it would skip over that part. I got that far that it attempts to start the vm but as it doesn't give right parameters to libvirt it fails to create a port on the bridge. I'm not having much luck with the 4.0 alpha though, this is a separate machine that I set up for this testing but when I run hosted-engine --deploy it just returns without any message. I did run the cleanup script found at http://www.ovirt.org/documentation/how-to/hosted-engine/ and rebooted in between but still no luck /Sverker On 2016-04-07 13:49, Dan Kenigsberg wrote: On Wed, Apr 06, 2016 at 11:57:08AM -0400, Martin Mucha wrote: Hi, I think OpenVSwitch should be supported in 4.0. M. - Original Message - Has anybody succeeded in installing Ovirt 3.6 with hosted engine on a server which uses OpenVSwitch for the network config? I believe my issue is that Ovirt wants to control the network to create a bridge for its management and I wants it to just use whatever network is available on the host without trying to be clever about it. I was able to tweak it to get to the final stage where it fails on waiting for the engine to start. Martin is right, but we should understand your usage of OpenVSwitch first. Do you intend to use it for networking of ovirt VMs? For something else? How did you tweak "it" (ovirt? hosted engine? ovs?) ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[ovirt-users] Ovirt + OpenVSwitch
Has anybody succeeded in installing Ovirt 3.6 with hosted engine on a server which uses OpenVSwitch for the network config? I believe my issue is that Ovirt wants to control the network to create a bridge for its management and I wants it to just use whatever network is available on the host without trying to be clever about it. I was able to tweak it to get to the final stage where it fails on waiting for the engine to start. Best regards Sverker ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users