There's been an issue for a while for users running jumbo frames on storage networks with the 1000v and not being able to boot VSMs after a complete outage because the default VEM mtu is 1500. Until the VSM is up and communicates with the VEM to change the MTU storage is not available. This creates the "chicken and egg" scenario. For further details, you can read Trey's explanation. At the end of his post, he says that this is being fixed. I recently installed SV1(3) for a customer and noticed there's now a command called "system mtu" that can be applied to ethernet port profiles. Documentation says that this sets the MTU for a "system" VLAN. Does anyone have an idea if this is our fix?
The wording on this is a bit confusing to me.
"For system uplinks ports, set the system mtu command in the system port profile applied to the port. This ensures the MTU is preserved for uplink interfaces with system VLANs.
For non-system uplinks, the VSM sets the correct MTU based on the interface configuration.The system mtu command does not apply to non-system profiles or ports."
This is a vethernet command (per veth port-profile that happens to have a system vlan directive), or an ethernet command (per uplink port-profile carrying my system vlans)? Seems to be the latter. Just want to be sure; I don't have a lab to mess with this... yet.
Is it required to set the MTU on the interface/port-channel level, or just the utilized profile? I guess this is a bit curious to me, since all my hosts use the same uplink port profile, and this change is an all-in-one-shot impact if that's the case, unless I create a new uplink port-profile and have them switch to it.
The 'system mtu' command is for uplink port profile that has system vlans. Once you have this configuration in a system uplink profile (which will be used by Ethernet interfaces of VEMs), it will be applied to all the Ethernet interfaces that uses this profile. You need not configure MTU on the Ethernet/port channel interface once you have it in the profile.