Fedora 25 Bonding 4 interfaces, but 2 gets always to stand by
Hi!
I was googling the whole day to find an answer, tried a lot of configurations, but no luck to solve my issue.
So I have a Cisco SG-300-52 type switch, which supports Link Aggregation with LACP. We already created 2 LAGs successfully to a Synology NAS and to a D-Link switch. Both uses 4 ports, and all 4 ports are active.
Now I wanted to connect a Fedora 25 server (4.11.5-200.fc25.x8664 #1 SMP Wed Jun 14 17:17:29 UTC 2017 x8664 x8664 x8664 GNU/Linux) using 4 ethernet interfaces to a 3rd LAG on this Cisco switch.
I configured the Bond using the tutorials found on various pages (fedora kb). Here are my configs:
[bond0]
DEVICE=bond0
IPADDR=192.168.0.2
GATEWAY=192.168.0.1
DNS1=192.168.0.3
BONDING_OPTS="miimon=100 updelay=0 downdelay=0 mode=4 lacp_rate=0 xmit_hash_policy=0 ad_select=1 all_slaves_active=1"
TYPE=Bond
BONDING_MASTER=yes
BOOTPROTO=none
IPV6INIT=no
IPV6_AUTOCONF=no
NAME="1. bond kapcsolat"
UUID=22a724f3-2505-4201-bc46-ee713931aa8e
ONBOOT=yes
USERCTL=no
[eno1]
DEVICE=eno1
BOOTPROTO=none
HWADDR=40:8D:5C:B0:94:1B
TYPE=Ethernet
NAME="bond0 1. alárendeltje"
UUID=bffd4d29-e98c-48c1-9c2d-c91630e393ff
ONBOOT=yes
MASTER=bond0
SLAVE=yes
USERCTL=no
[enp10s0]
DEVICE=enp10s0
HWADDR=00:13:3B:0F:C9:29
TYPE=Ethernet
NAME="bond0 2. alárendeltje"
UUID=942939ca-f682-458f-a04e-4378200cd6b0
ONBOOT=yes
MASTER=bond0
SLAVE=yes
BOOTPROTO=none
USERCTL=no
[enp8s0]
DEVICE=enp8s0
HWADDR=00:13:3B:0F:C9:28
TYPE=Ethernet
NAME="bond0 3. alárendeltje"
UUID=26c01969-f633-4c36-8c31-1b32ce9ee0b0
ONBOOT=yes
MASTER=bond0
SLAVE=yes
BOOTPROTO=none
USERCTL=no
[enp4s0]
DEVICE=enp4s0
HWADDR=00:13:3B:0F:E9:AB
TYPE=Ethernet
NAME="bond0 4. alárendeltje"
UUID=5f7f91a5-6965-4c26-9264-7c2239830b70
ONBOOT=yes
MASTER=bond0
SLAVE=yes
BOOTPROTO=none
USERCTL=no
The bond comes up correctly, however 2 interfaces (enp8s0 & enp10s0) become always "stand by" on the switch, and only the eno1 and the enp4s0 are used as active links. I have no idea, why. Both the Synology NAS and the D-Link switch can utilize the 4 connections. The LACP is enabled on the Cisco for the linux LAG group also!
[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): 100
Up Delay (ms): 0
Down Delay (ms): 0
802.3ad info
LACP rate: slow
Min links: 0
Aggregator selection policy (ad_select): bandwidth
System priority: 65535
System MAC address: 40:8d:5c:b0:94:1b
Active Aggregator Info:
Aggregator ID: 1
Number of ports: 4
Actor Key: 9
Partner Key: 1002
Partner Mac Address: 6c:50:4d:cb:b2:93
Slave Interface: eno1
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 40:8d:5c:b0:94:1b
Slave queue ID: 0
Aggregator ID: 1
Actor Churn State: none
Partner Churn State: none
Actor Churned Count: 0
Partner Churned Count: 0
details actor lacp pdu:
system priority: 65535 ...
Whenever I start (restart) networking I get following warning jún 23 16:26:17 kernel: bond0: Warning: No 802.3ad response from the link partner for any adapters in the bond
Found a link about bonding. It they use lacp_rate=1 in their configuration. It might work. LACP – How to Configure Network Bonding in Linux