nova doesn't use Neutron MTUs when wiring up VIFs
Bug #1549408 reported by
Kevin Benton
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Compute (nova) |
Fix Released
|
Medium
|
Kevin Benton | ||
Liberty |
Fix Released
|
Medium
|
Matt Riedemann |
Bug Description
Neutron Network objects can have an 'MTU' field that specifies the MTU that should be defined on all interfaces attached to that network. However, Nova currently doesn't do anything with that information when wiring up the Linux bridge and OVS ports. It only references the network_device_mtu config option which is deprecated and doesn't take into account the overhead of various encapsulation protocols.
So if network_device_mtu is not specified or is smaller than the MTU of a Neutron network with jumbo frames, large packets from a VM will silently be dropped on the Linux bridge due to the mismatch of MTUs between the devices Nova has wired up and Neutron.
Changed in nova: | |
assignee: | nobody → Kevin Benton (kevinbenton) |
status: | New → In Progress |
tags: | added: network neutron |
Changed in nova: | |
importance: | Undecided → Medium |
tags: | added: liberty-backport-potential |
tags: | removed: liberty-backport-potential |
To post a comment you must log in.
https:/ /review. openstack. org/#/c/ 283847/