Even if private interface is configured with MTU 9000 IP packets larger than 1500 do not reach destination.
Steps to reproduce:
1. Provision 2 instances in the same tenant network (L2 domain) and on different compute nodes
2. Run iperf between nodes:
ubuntu@shaker-apzokt-master-0:~$ iperf -c 10.0.0.5 -t 10 -M 1460 -- succeed, shows > 9G
ubuntu@shaker-apzokt-master-0:~$ iperf -c 10.0.0.5 -t 10 -M 1470 -- hangs
Deployment:
3 controllers, 3 computes, VLAN, Ubuntu
Network configuration:
eth0 - public
eth1 - <none>
eth2 - admin, storage, mgmt
eth3 - private, MTU 9000
On all interfaces 'disable offloading' switched off.
VERSION:
feature_groups:
- mirantis
production: "docker"
release: "6.1"
openstack_version: "2014.2.2-6.1"
api: "1.0"
build_number: "521"
build_id: "2015-06-08_06-13-27"
nailgun_sha: "4340d55c19029394cd5610b0e0f56d6cb8cb661b"
python-fuelclient_sha: "4fc55db0265bbf39c369df398b9dc7d6469ba13b"
astute_sha: "7766818f079881e2dbeedb34e1f67e517ed7d479"
fuel-library_sha: "f43c2ae1af3b493ee0e7810eab7bb7b50c986c7d"
fuel-ostf_sha: "7c938648a246e0311d05e2372ff43ef1eb2e2761"
fuelmain_sha: "bcc909ffc5dd5156ba54cae348b6a07c1b607b24"
our preliminary analysis shows that the possible reason - jumbo frames are not allowed on routers. moving to scalelab