Configure Network Teaming on CentOS / RHEL 8/7

Question: How to configure Network Teaming on CentOS / RHEL 8? or How to configure Network bonding on CentOS / RHEL 8/7?. In this blog post, we will discuss how to configure Network Teaming on RHEL / CentOS 8/7. Network teaming can also be called NIC teaming, channel bonding or link aggregation.

Teaming involves combining or aggregating of network links to provide a logical link with higher throughput or to provide redundancy. This concept of teaming is widely referred to as bonding as implemented in the Linux kernel.

Teaming Concepts & terminologies

• teamd – An application daemon which uses the libteam library to implement load-balancing and round-robin logic. It listens and communicates via Unix Domain Sockets.
• teamdctl – A utility used to control a running instance of teamd using D-bus. It can be used during run time to read the configuration, the state of link-watchers, check and change the state of ports, add and remove ports, and to change ports between active and backup states.
• runners – Separate units of code which implements functions that are unique to the different load sharing and backup methods, such as round-robin. The user specifies the runner in the JSON format configuration file and the code is then compiled into an instance of teamd when the instance is created

Available runners (Team running modes)

• broadcast – data is transmitted over all ports
• active-backup – one port or link is used while others are kept as a backup
• round-robin – data is transmitted over all ports in turn
• loadbalance – With active Tx load balancing and BPF-based Tx port selectors
• lacp – implements the 802.3ad Link Aggregation Control Protocol

Network teaming link watchers

• ethtool – used to watch for link state changes. This is the default if no other link-watcher is specified in the configuration file.
• arp_ping – used to monitor the presence of a far-end hardware address using ARP packets.
• nsna_ping – Used to monitor the presence of a neighbor’s interface

Configure network teaming daemon on RHEL / CentOS 8

To install teamd, issue the following command in your termina:

sudo dnf -y install teamd

Version of teamd installed can be obtained with:

$ rpm -qi teamd
 Name        : teamd
 Version     : 1.27
 Release     : 9.el8
 Architecture: x86_64
 Install Date: Wed 26 Dec 2018 10:54:05 AM EAT
 Group       : System Environment/Daemons
 Size        : 291570
 License     : LGPLv2+
 Signature   : RSA/SHA256, Sat 25 Aug 2018 03:27:14 PM EAT, Key ID 199e2f91fd431d51
 Source RPM  : libteam-1.27-9.el8.src.rpm
 Build Date  : Mon 20 Aug 2018 06:31:15 PM EAT
 Build Host  : x86-vm-10.build.eng.bos.redhat.com
 Relocations : (not relocatable)
 Packager    : Red Hat, Inc. http://bugzilla.redhat.com/bugzilla
 Vendor      : Red Hat, Inc.
 URL         : http://www.libteam.org
 Summary     : Team network device control daemon
 Description :
 The teamd package contains team network device control daemon.

Configure network teaming with nmcli

First of all, identify the network interfaces that will be used as slaves.

# ip link  show
 1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
 2: enp1s0:  mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
     link/ether 52:54:00:8f:8c:86 brd ff:ff:ff:ff:ff:ff
 3: enp7s0:  mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
     link/ether 52:54:00:74:99:a9 brd ff:ff:ff:ff:ff:ff
 4: enp8s0:  mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
     link/ether 52:54:00:2b:db:9c brd ff:ff:ff:ff:ff:ff

Since I have NetworkManager running in my RHEL / CentOS 8 server machine, I’ll use nmcli network management tool.

# nmcli connection show 
NAME                UUID                                  TYPE      DEVICE 
enp1s0              498869bb-0d88-4a4c-a83a-c491d1040b0b  ethernet  enp1s0 
Wired connection 1  48b9eacf-eb8b-348b-a44d-1d74ba22a6d6  ethernet  enp7s0 
Wired connection 2  10d1def1-39dc-3468-b9cd-263e72424383  ethernet  enp8s0 

For my setup, i’ll use the two network devices – enp7s0 and enp8s0. Delete the connections with nmcli before you start the configuration.

The two devices should now be in disconnected state.

$ nmcli device status
DEVICE  TYPE      STATE         CONNECTION 
enp1s0  ethernet  connected     enp1s0     
enp7s0  ethernet  disconnected  --         
enp8s0  ethernet  disconnected  --         
lo      loopback  unmanaged     --  

To create a new team interface, with name team0:

$ nmcli connection add type team con-name team0 ifname team0 \
  config '{ "runner": {"name": "loadbalance"}, "link_watch": {"name": "ethtool"}}'

You can replace loadbalance with:

• activebackup
• broadcast
• random
• roundrobin

NetworkManager will write configuration file to /etc/sysconfig/network-scripts/ifcfg-team-team0. To view the other values assigned, run:

$ nmcli con show team0

Use command below to view Team interfaces just configured.

$ nmcli con show

Assign team interface an IP address, DNS and auto-connection.

nmcli con mod team0 ipv4.addresses 192.168.121.10/24
nmcli con mod team0 ipv4.gateway 192.168.121.1
nmcli con mod team0 ipv4.dns 8.8.8.8
nmcli con mod team0 ipv4.method manual
nmcli con mod team0 connection.autoconnect yes

Add Network Slave to the team.

nmcli con add type team-slave con-name team0-slave0 ifname enp7s0 master team0
nmcli con add type team-slave con-name team0-slave1 ifname enp8s0 master team0

Sample output.

Connection 'team0-slave0' (daa6fc23-cdef-40b1-9b9d-5157d6ff3910) successfully added.
Connection 'team0-slave1' (a20cf7ee-fb08-4270-a6e7-b6e20cb490dc) successfully added.

Confirm connection details.

Restart the connection.

nmcli connection down team0 && nmcli connection up team0

You should see assigned IP information.

$ ip addr show  dev  team0
 8: team0:  mtu 1500 qdisc noqueue state UP group default qlen 1000
     link/ether 52:54:00:74:99:a9 brd ff:ff:ff:ff:ff:ff
     inet 192.168.121.10/24 brd 192.168.121.255 scope global noprefixroute team0
        valid_lft forever preferred_lft forever
     inet6 fe80::4c51:96b:c24e:ede9/64 scope link noprefixroute 
        valid_lft forever preferred_lft forever
$ ping -c 1 192.168.121.1
 PING 192.168.121.1 (192.168.121.1) 56(84) bytes of data.
 64 bytes from 192.168.121.1: icmp_seq=1 ttl=64 time=0.134 ms
 --- 192.168.121.1 ping statistics ---
 1 packets transmitted, 1 received, 0% packet loss, time 0ms
 rtt min/avg/max/mdev = 0.134/0.134/0.134/0.000 ms

The team status can also be checked with teamdctl .

$ teamdctl team0 state
 setup:
   runner: loadbalance
 ports:
enp7s0
     link watches:
       link summary: up
       instance[link_watch_0]:
         name: ethtool
         link: up
         down count: 0
enp8s0
     link watches:
       link summary: up
       instance[link_watch_0]:
         name: ethtool
         link: up
         down count: 0

How to Delete Team with nmcli

Show connection details.

$ nmcli connection show 
 NAME          UUID                                  TYPE      DEVICE 
 enp1s0        498869bb-0d88-4a4c-a83a-c491d1040b0b  ethernet  enp1s0 
 team0         f763a709-3956-497f-b92c-5c06f848bee7  team      team0  
 team0-slave0  daa6fc23-cdef-40b1-9b9d-5157d6ff3910  ethernet  enp7s0 
 team0-slave1  a20cf7ee-fb08-4270-a6e7-b6e20cb490dc  ethernet  enp8s0 

Bring down team connection.

nmcli connection down team0

Delete slaves.

$ nmcli connection delete team0-slave0 team0-slave1
Connection 'team0-slave0' (daa6fc23-cdef-40b1-9b9d-5157d6ff3910) successfully deleted.
Connection 'team0-slave1' (a20cf7ee-fb08-4270-a6e7-b6e20cb490dc) successfully deleted.

Delete team created.

$ nmcli connection delete team0
Connection 'team0' (f763a709-3956-497f-b92c-5c06f848bee7) successfully deleted.

Confirm deletion.

$ nmcli connection show 
 NAME    UUID                                  TYPE      DEVICE 
 enp1s0  498869bb-0d88-4a4c-a83a-c491d1040b0b  ethernet  enp1s0 

Creating a Network Team Using ifcfg Files

To create a networking team using ifcfg files, create a file in the /etc/sysconfig/network-scripts/ directory as follows:

$ cat /etc/sysconfig/network-scripts/ifcfg-team0

DEVICE=team0
NAME=team0
DEVICETYPE=Team
ONBOOT=yes
BOOTPROTO=none
IPADDR=192.168.121.10
PREFIX=24
GATEWAY=192.168.121.1
TEAM_CONFIG='{"runner": {"name": "activebackup"}, "link_watch": {"name": "ethtool"}}'

To create a port to be a member of team0, create one or more files in the /etc/sysconfig/network-scripts/ directory as follows:

$ cat /etc/sysconfig/network-scripts/ifcfg-team0-slave0

NAME=team0-slave0
DEVICE=enp7s0
ONBOOT=yes
TEAM_MASTER=team0
DEVICETYPE=TeamPort

$ cat /etc/sysconfig/network-scripts/ifcfg-team0-slave1

NAME=team0-slave1
DEVICE=enp8s0
ONBOOT=yes
TEAM_MASTER=team0
DEVICETYPE=TeamPort 

Start team interface.

$ sudo ip link set enp7s0 down
$ sudo ip link set enp8s0 down
$ sudo ifup team0

List team ports.

$teamnl team0 ports
 4: enp8s0: up 4294967295Mbit FD 
 3: enp7s0: up 4294967295Mbit FD 

Key notes on managing Team interfaces

1. Starting the master interface does not automatically start the port interfaces.
2. Starting a port interface always starts the master interface.
3. Stopping the master interface also stops the port interfaces.
4. A master without ports can start static IP connections.
5. A master without ports waits for ports when starting DHCP connections.
6. A master with a DHCP connection waiting for ports completes when a port with a carrier is added.
7. A master with a DHCP connection waiting for ports continues waiting when a port without a carrier is added.

You can also check:

How to Create a Linux Network Bridge on RHEL / CentOS 8

Adding a Secondary IP Address to RHEL / CentOS 8 Network Interface

Install and Configure OpenVPN Server on RHEL / CentOS 8