The timeout is hardcoded: https://github.com/openvswitch/ovs/blob/d4bd0a2ad54b63c1e767ea2a7bff761d4676514a/rhel/etc_sysconfig_network-scripts_ifup-ovs#L25

There's a bug report where ovs-ctl command took longer than 10 seconds, apparently because there's a large number of ports: https://bugzilla.redhat.com/show_bug.cgi?id=2280910

While the original report should probably clarify why the particular command took so long, it may make sense to allow the timeout to be higher in particular situations. This issue is to track this ability.

F.e. the command that timed out for customer is:

 
ovs-vsctl -t 10 – --if-exists del-port br-vnf1 dpdkbond-tenant1 – add-bond br-vnf1 dpdkbond-tenant1 dpdkbond-tenant1-port0 dpdkbond-tenant1-port1 bond_mode=active-backup other_config:miimon=100 other_config:updelay=200 other_config:downdelay=200 – set interface dpdkbond-tenant1-port0 type=dpdk – set interface dpdkbond-tenant1-port1 type=dpdk – set Interface dpdkbond-tenant1-port0 options:dpdk-devargs=0000:b1:00.0 – set Interface dpdkbond-tenant1-port1 options:dpdk-devargs=0000:b1:00.1 – set Interface dpdkbond-tenant1-port0 mtu_request=9000 – set Interface dpdkbond-tenant1-port1 mtu_request=9000 – set Interface dpdkbond-tenant1-port0 options:n_rxq=2 – set Interface dpdkbond-tenant1-port1 options:n_rxq=2