[ovs-discuss] Cannot connect Openstack VM to OVS in DPDK mode (vhost-user client)

Sat May 26 22:43:18 UTC 2018

Hi
I am having issues trying to connect a VM in Openstack with OVS in DPDK
mode (dpdkvhostuserclient)

Apparently, the vhostuser client port, is not working in OVS, for some
memory issues
2018-05-26T22:07:08.669Z|00189|netdev_dpdk|ERR|Failed to create memory pool
for netdev vhud402d58e-c9, with MTU 1500 on socket 0: Invalid argument
2018-05-26T22:07:08.669Z|00190|dpif_netdev|ERR|Failed to set interface
vhud402d58e-c9 new configuration

ovsdb-server and ovs-vsswitchd are running in docker containters, but this
should not be the problem, because those containters work in kernel mode.
the problem is when using netdev and vhost-user (client)

Openstack was configured for using netdev
[OVS]
datapath_type = netdev
vhostuser_socket_dir = /var/run/openvswitch

Bridge is created
    Bridge br-int
        Controller "tcp:127.0.0.1:6633"
            is_connected: true
        fail_mode: secure
        Port "vhud402d58e-c9"
            tag: 1
            Interface "vhud402d58e-c9"
                type: dpdkvhostuserclient
                options: {vhost-server-path=
"/var/run/openvswitch/vhud402d58e-c9"}
        Port int-dpdk_bridge
            Interface int-dpdk_bridge
                type: patch
                options: {peer=phy-dpdk_bridge}
        Port patch-tun
            Interface patch-tun
                type: patch
                options: {peer=patch-int}
        Port br-int
            Interface br-int
                type: internal

with netdev
_uuid               : d031a04b-e392-4fd5-9b14-68ff04289976
auto_attach         : []
controller          : [64f24e69-c84b-4ab9-b1b0-d792968b7f76]
datapath_id         : "00004aa031d0d54f"
datapath_type       : netdev
datapath_version    : "<built-in>"
external_ids        : {}
fail_mode           : secure
flood_vlans         : []
flow_tables         : {}
ipfix               : []
mcast_snooping_enable: false
mirrors             : []
name                : br-int
netflow             : []
other_config        : {}
ports               : [730ecb57-d985-4e6e-9e4c-1d33c355672f,
745d44c2-70af-49d3-b3e8-72d4acc3410d, 97cc5449-8a47-416e-8f35-0dbf839109a8,
fff53944-6891-438c-80e8-aceb940601f0]
protocols           : ["OpenFlow10", "OpenFlow13"]
rstp_enable         : false
rstp_status         : {}
sflow               : []
status              : {}
stp_enable          : false

and port is dpdkvhostuserclient
(ovsdpdk-db)[root at s131002 /]# ovs-vsctl list Interface
_uuid               : 6bde0358-3cb1-4648-958d-cac81dc683b4
admin_state         : up
bfd                 : {}
bfd_status          : {}
cfm_fault           : []
cfm_fault_status    : []
cfm_flap_count      : []
cfm_health          : []
cfm_mpid            : []
cfm_remote_mpids    : []
cfm_remote_opstate  : []
duplex              : []
error               : []
external_ids        : {attached-mac="fa:16:3e:ba:d0:e1", iface-id=
"d402d58e-c9c7-4b65-94e0-2b38c3cfd926", iface-status=active, vm-uuid=
"c7946cf9-6529-446a-9eb3-d2cfefbfddd7"}
ifindex             : 7907161
ingress_policing_burst: 0
ingress_policing_rate: 0
lacp_current        : []
link_resets         : 0
link_speed          : []
link_state          : down
lldp                : {}
mac                 : []
mac_in_use          : "00:00:00:00:00:00"
mtu                 : 0
mtu_request         : 1500
name                : "vhud402d58e-c9"
ofport              : 3
ofport_request      : []
options             : {vhost-server-path=
"/var/run/openvswitch/vhud402d58e-c9"}
other_config        : {}
statistics          : {"rx_1024_to_1522_packets"=0, "rx_128_to_255_packets"=0,
"rx_1523_to_max_packets"=0, "rx_1_to_64_packets"=0, "rx_256_to_511_packets"=0,
"rx_512_to_1023_packets"=0, "rx_65_to_127_packets"=0, rx_bytes=0,
rx_dropped=0, rx_errors=0, rx_packets=0, tx_bytes=0, tx_dropped=0,
tx_packets=0}
status              : {}
type                : dpdkvhostuserclient

The virtual machine is running in qemu with this arguments
-netdev vhost-user,chardev=charnet0,id=hostnet0
-chardev socket,id=charnet0,path=/var/run/openvswitch/vhud402d58e-c9,server
-device virtio-net-pci,netdev=hostnet0,id=net0,mac=fa:16:3e:ba:d0:e1,bus=pci
.0,addr=0x3

The socket is created
srwxrwxr-x 1 42436 42436 0 May 26 17:07 /var/run/openvswitch/vhud402d58e-c9

Nova logs does not evidence errors
2018-05-26 17:07:09.779 7 INFO nova.virt.libvirt.driver [req-4f6cbb38-652d-4
a6b-a205-9a8bf1096bfc f2ef1bf11ce8480980301a68fbaad0ac
ca4f4f14e4c6448d94e34064dfd12eaf - default default] [instance: c7946cf9-6529
-446a-9eb3-d2cfefbfddd7] Instance rebooted successfully.
2018-05-26 17:07:09.888 7 INFO nova.compute.manager [req-4f6cbb38-652d-4
a6b-a205-9a8bf1096bfc f2ef1bf11ce8480980301a68fbaad0ac
ca4f4f14e4c6448d94e34064dfd12eaf - default default] [instance: c7946cf9-6529
-446a-9eb3-d2cfefbfddd7] VM Started (Lifecycle Event)
2018-05-26 17:07:28.155 7 INFO nova.compute.resource_tracker
[req-4f6cbb38-652d-4a6b-a205-9a8bf1096bfc f2ef1bf11ce8480980301a68fbaad0ac
ca4f4f14e4c6448d94e34064dfd12eaf - default default] Final resource view:
name=s131002.nocix.net phys_ram=24164MB used_ram=7168MB phys_disk=250GB
used_disk=61GB total_vcpus=16 used_vcpus=4 pci_stats=[]

Do you have any clue about what could be happening??

I would really appreciate any guidance on this topic..

Thanks
JM
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openvswitch.org/pipermail/ovs-discuss/attachments/20180526/09c40e92/attachment-0001.html>