Description of problem: If a user tries to configure jumbo frames on baremetal by following the guide at https://docs.openshift.com/container-platform/4.4/installing/installing_bare_metal/installing-bare-metal-network-customizations.html and creating a manifest cluster-network-03-config.yml with the SDN MTU set to 8950, while the veth interfaces and tunnel interfaces end up at 8950, the baremetal interface on the host itself is still at 1500 MTU instead of 9000. It would be god to have the baremetal interface MTU also to be set based off the manifests, instead of requiring a workaround like using an ifcfg file to set MTU on the baremetal interface to 9000 using ignition manually. Version: 4.4.6 How reproducible: 100% Steps to Reproduce: 1. follow the procedure in https://docs.openshift.com/container-platform/4.4/installing/installing_bare_metal/installing-bare-metal-network-customizations.html to configure the OpenShiftSDN MTU to 8950 2. Deploy the cluster 3. Check that MTU of veth and tunnel interfaces and see that while they are set to 8950, the baremetal interface itself is set to 1500 Actual results: veth interfaces and tunnel interfaces (vxlan) are set to 8950, but without setting MTU through an ifcfg file and ignition, the baremetal interface MTU is at 1500 Expected results: Based off the MTU in the manifests, baremetal interface MTU should be automatically configured to 9000 Additional info:
Note that you do not need to set the MTU manually, *as long as* all the masters have their BM NIC MTU set correctly. The Cluster Network Operator will read the MTU of the machine it runs on (a master) and use that as the default MTU for all nodes in the cluster. TLDR; make sure your BM NIC MTU is correctly set, and the cluster will use that MTU automatically.
*** This bug has been marked as a duplicate of bug 1846485 ***