cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1990
Views
0
Helpful
1
Replies

LACP (802.11ad) bonding WS-C3750G-24TS-E and LInux Centos 6.7

I have two lacp bonds configured on my linux box, identical to each other.  One is on an iscsi vlan, the other on my access vlan.

The iscsi bond looks fine:

# cat /proc/net/bonding/bond1
Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)

Bonding Mode: IEEE 802.3ad Dynamic link aggregation
Transmit Hash Policy: layer3+4 (1)
MII Status: up
MII Polling Interval (ms): 100
Up Delay (ms): 0
Down Delay (ms): 0

802.3ad info
LACP rate: slow
Min links: 0
Aggregator selection policy (ad_select): stable
Active Aggregator Info:
    Aggregator ID: 1
    Number of ports: 2
    Actor Key: 9
    Partner Key: 12
    Partner Mac Address: 00:13:80:73:2f:80

Slave Interface: eth2
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 40:f2:e9:0d:e1:de
Aggregator ID: 1
Slave queue ID: 0

Slave Interface: eth3
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 40:f2:e9:0d:e1:df
Aggregator ID: 1
Slave queue ID: 0

but the access bond is funny in that it shows Number of ports: 1

# cat /proc/net/bonding/bond0
Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)

Bonding Mode: IEEE 802.3ad Dynamic link aggregation
Transmit Hash Policy: layer2 (0)
MII Status: up
MII Polling Interval (ms): 0
Up Delay (ms): 0
Down Delay (ms): 0

802.3ad info
LACP rate: slow
Min links: 0
Aggregator selection policy (ad_select): stable
Active Aggregator Info:
    Aggregator ID: 1
    Number of ports: 1
    Actor Key: 9
    Partner Key: 11
    Partner Mac Address: 00:13:80:73:2f:80

Slave Interface: eth0
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 40:f2:e9:0d:e1:dc
Aggregator ID: 1
Slave queue ID: 0

Slave Interface: eth1
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 40:f2:e9:0d:e1:dd
Aggregator ID: 2
Slave queue ID: 0

From both the host and the switch side, both bonds are configured identically:

# cat /etc/sysconfig/network-scripts/ifcfg-bond0
DEVICE=bond0
TYPE=Ethernet
ONBOOT=yes
NM_CONTROLLED=no
BOOTPROTO=static
DEFROUTE=yes
IPV6INIT=no
NAME="System bond0"
PEERDNS=yes
PEERROUTES=yes
IPADDR=10.1.11.60
NETMASK=255.255.255.0
GATEWAY=10.1.11.1
BONDING_OPTS=mode=4 miimon=100 xmit_hash_policy=layer3+4

# cat /etc/sysconfig/network-scripts/ifcfg-bond1
DEVICE=bond1
TYPE=Ethernet
ONBOOT=yes
NM_CONTROLLED=no
BOOTPROTO=static
IPV6INIT=no
IPADDR=192.43.216.60
NETMASK=255.255.255.0
MTU=9000
BONDING_OPTS="mode=4 miimon=100 xmit_hash_policy=layer3+4"

and from the switch side, it looks fine

interface Port-channel11
 description 'access'
 switchport mode access
 switchport nonegotiate
 load-interval 30
 spanning-tree portfast
!
interface Port-channel12
 description 'iscsi'
 switchport access vlan 216
 switchport mode access
 switchport nonegotiate
 load-interval 30
 spanning-tree portfast

!

interface GigabitEthernet1/0/17
 switchport access vlan 216
 switchport mode access
 switchport nonegotiate
 load-interval 30
 channel-group 12 mode active
 spanning-tree portfast

!
interface GigabitEthernet1/0/18
 switchport access vlan 216
 switchport mode access
 switchport nonegotiate
 load-interval 30
 channel-group 12 mode active
 spanning-tree portfast
!
interface GigabitEthernet1/0/19
 switchport mode access
 switchport nonegotiate
 load-interval 30
 channel-group 11 mode active
 spanning-tree portfast
!
interface GigabitEthernet1/0/20
 switchport mode access
 switchport nonegotiate
 load-interval 30
 channel-group 11 mode active
 spanning-tree portfast

switch#sho lacp counters
             LACPDUs         Marker      Marker Response    LACPDUs
Port       Sent   Recv     Sent   Recv     Sent   Recv      Pkts Err
---------------------------------------------------------------------
Channel group: 11
Gi1/0/19    90     20       0      0        0      0         0     
Gi1/0/20    75     45       0      0        0      0         0     

             LACPDUs         Marker      Marker Response    LACPDUs
Port       Sent   Recv     Sent   Recv     Sent   Recv      Pkts Err
---------------------------------------------------------------------
Channel group: 12
Gi1/0/17    93     57       0      0        0      0         0     
Gi1/0/18    89     57       0      0        0      0         0

So - why is the host showing Number of ports: 1 for the access bond?

1 Reply 1

The problem appears to be on the linux side.  eth1 is not being used to send any data.  Both ports are receiving.  I swapped cables at the host side and the problem stayed with eth1, then I swapped them on the switch side and the problem stayed with eith1, so the switch and the cables are no longer suspect.  Rebooting failed to change the behavior.


# dmesg | grep eth1
igb 0000:06:00.1: added PHC on eth1
igb 0000:06:00.1: eth1: (PCIe:5.0Gb/s:Width x4) 40:f2:e9:0d:e1:dd
igb 0000:06:00.1: eth1: PBA No: 106100-000
bond0: Adding slave eth1
ADDRCONF(NETDEV_UP): eth1: link is not ready
bond0: Enslaving eth1 as a backup interface with an up link
igb 0000:06:00.1: eth1: igb: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX