Trying to isolate the commands used in the failing test (works in a not too small Oracular VM):

0. Prep install and config
 apt install openvswitch-switch-dpdk dpdk-dev
 echo "NR_2M_PAGES=784" >> /etc/dpdk/dpdk.conf
 systemctl restart dpdk
 update-alternatives --set ovs-vswitchd /usr/lib/openvswitch-switch-dpdk/ovs-vswitchd-dpdk
 ovs-vsctl --no-wait init
 ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-init=true
 ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-extra="--log-level=pmd.*:error --no-pci"
1. start OVS with DPDK support (this is the point where you'd restart new test iterations)
 systemctl stop openvswitch-switch ovs-vswitchd
 rm /var/log/openvswitch/ovs-vswitchd.log
 systemctl start openvswitch-switch ovs-vswitchd
 ovs-vsctl show
2. create an internal DPDK bridge with a vhostuserclient port with defined MTU
 ovs-vsctl add-br br10 -- set bridge br10 datapath_type=netdev
 OVS_RUNDIR="/tmpdir/_debian/tests/system-dpdk-testsuite.dir/020"
 rm -rf $OVS_RUNDIR
 mkdir -p $OVS_RUNDIR
 ovs-vsctl add-port br10 dpdkvhostuserclient0 -- set Interface dpdkvhostuserclient0 type=dpdkvhostuserclient options:vhost-server-path=$OVS_RUNDIR/dpdkvhostclient0
 ovs-vsctl show
3. connect a vhostuserclient server
 # Do this in a shell of its own, it will stay up
 dpdk-testpmd --in-memory --socket-mem=512 --single-file-segments --no-pci --vdev="net_virtio_user,path=$OVS_RUNDIR/dpdkvhostclient0,server=1 --vdev=net_tap0,iface=tap0 -- -a --stats-period 2 --nb-cores 2 --rxq 2 --txq 2
4. check if it is truly ready
 ovs-vsctl show
 grep "virtio is now ready for processing" /var/log/openvswitch/ovs-vswitchd.log
 # should have a happy dpdkvhostuserclient0
 ovs-vsctl get Interface dpdkvhostuserclient0 link_state
 # should say "up"
5. set happy MTU
 ovs-vsctl set Interface dpdkvhostuserclient0 mtu_request=9702
 ovs-vsctl get Interface dpdkvhostuserclient0 mtu
 # should deliver 9702
6. set the bigger MTU
 ovs-vsctl set Interface dpdkvhostuserclient0 mtu_request=9711
 # the ovs-vswitchd.log will say (expected)
   dpdkvhostuserclient0: unsupported MTU 9711
   failed to set MTU for network device dpdkvhostuserclient0: Invalid argument
 ovs-vsctl get Interface dpdkvhostuserclient0 mtu
   # will still deliver 9702
7. cleanup the port
 ovs-vsctl del-port br10 dpdkvhostuserclient0
 # Should work without bad RC, log in ovs-vswitchd.log will say:
    bridge|INFO|bridge br10: deleted interface dpdkvhostuserclient0 on port 1
    dpif_netdev|INFO|PMD thread on numa_id: 0, core id:  1 destroyed.
    dpdk(pmd-c01/id:8)|INFO|PMD thread released DPDK lcore 1.
    dpdk|INFO|VHOST_CONFIG: (/tmpdir/_debian/tests/system-dpdk-testsuite.dir/020/dpdkvhostclient0) free connfd 50
    netdev_dpdk|INFO|vHost Device '/tmpdir/_debian/tests/system-dpdk-testsuite.dir/020/dpdkvhostclient0' not found
 ovs-vsctl del-br br10
 ovs-vsctl show
 # should show an empty switch


^^ Here I expected the issue with the new code (OVS + DPDK being new).
But sadly I have proven that the "not found" on the path was a red herring as well.
It comes up in the good path as well.
The good case just does not push the log output to the test-output.
Hence it seemed to be an issue "only in the bad case" but that was a fluke.