On Sat, May 25, 2019 at 5:06 AM <[email protected]> wrote: > Hello, > > I've a problem, all my ovirt hosts and vms are linked with a bonding mode > 1（Active-Backup）2x10Gbps > ovirt version：4.3 > topology： > ----------eno2 > vm----------ovirtmgmt------bond0-----------eno1 > > ifcfg-bond0： > # Generated by VDSM version 4.30.9.1 > DEVICE=bond0 > BONDING_OPTIOS='mode=1 miion=100' > BRIDGE=ovirtmgmt > MACADDR=a4:be:26:16:e9:b2 > ONBOOT=yes > MTU=1500 > DEFROUTE=no > NM_CONTROLLER=no > IPV6INIT=no > > ifcfg-eno1: > # Generated by VDSM version 4.30.9.1 > DEVICE=eno1 > MASTER=bond0 > SLAVE=yes > ONBOOT=yes > MTU=1500 > DEFROUTE=no > NM_CONTROLLER=no > IPV6INIT=no > > ifcfg-eno2: > # Generated by VDSM version 4.30.9.1 > DEVICE=eno2 > MASTER=bond0 > SLAVE=yes > ONBOOT=yes > MTU=1500 > DEFROUTE=no > NM_CONTROLLER=no > IPV6INIT=no > > ifcfg-ovirtmgmt: > # Generated by VDSM version 4.30.9.1 > DEVICE=ovirtmgmt > TYPE=Brodge > DELAY=0 > STP=off > ONBOOT=yes > IPADDR=x.x.x.x > NEYMASK=255.255.255.0 > GATEWAY=x.x.x.x > BOOTPROTO=none > MTU=1500 > DEFROUTE=yes > NM_CONTROLLER=no > IPV6INIT=yes > IPV6_AUTOCONF=yes > > > cat /proc/net/bonding/bond0 > Ethernet Chanel Bonding Driver:v3.7.1(April 27, 2011) > > Bonding Mode:fault-tolerance(active-ackup) > Primary Slave:none > Currently Active Slave:eno1 > MII Status:up > MII Polling Intercal (ms):100 > Up Delay (ms) : 0 > Down Delay (ms) : 0 > > Slave Interface :eno1 > MII Status:up > Speed : 10000 Mbps > Link Failure Count : 0 > Permanent HW addr :a4:be:26:16:e9:b2 > Slave queue ID: 0 > > Slave Interface :eno2 > MII Status:up > Speed : 10000 Mbps > Link Failure Count : 0 > Permanent HW addr :a4:be:26:16:e9:b2 > Slave queue ID: 0 > > ping vm from different subnet. > > Eveything is okay if I don't change bond link interface。When I unplug > Currently Active Slave eno1,bond link change to eno2 as expected but vm > become unreachable until external physical switch MAC Table ageing time > expired.It seems that vm doesn't sent gratuitous ARP when bond link change. > How can I fix if? >
There is no reason for the VM OS to send anything as it is unaware of the change you have done in the network. It should work fine if you perform this operation from oVirt Management, as it will cause the interfaces to be set down and up again (I would expect the links to go down as a result), causing the switch ports to flush its mac address table. > vm os is Centos 7.5 > ovirt version 4.2 also tested. > _______________________________________________ > Users mailing list -- [email protected] > To unsubscribe send an email to [email protected] > Privacy Statement: https://www.ovirt.org/site/privacy-policy/ > oVirt Code of Conduct: > https://www.ovirt.org/community/about/community-guidelines/ > List Archives: > https://lists.ovirt.org/archives/list/[email protected]/message/CUC67VZ7WNW5M4L7IBBDIUZKK7SRLMLQ/ >
_______________________________________________ Users mailing list -- [email protected] To unsubscribe send an email to [email protected] Privacy Statement: https://www.ovirt.org/site/privacy-policy/ oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/[email protected]/message/FAAET3PUF5OWWH7GUA72G3ICWK3MLSRU/
