Table of Contents

Toggle

Introduction

Network Teaming, Bonding or Link Aggregation, We are about to explore how to create a Linux Team interface. To create a network team interface we need a minimum of two numbers of physical or virtual interfaces.

By combing number of interfaces, a team will be created, It’s created to provide more link redundancy to increase the speed of connection, to keep the uptime of server up and running while anyone of interface failure or cable disconnectivity. This makes the critical business production servers running smooth by avoiding any downtimes.

Let’s see the type of available network teaming in Linux.

Types of Network Teaming

Modes	Mode No	A Short Description
RoundRobin	0	Send traffic over the available slave ports in a round-robin fashion.
Active-Backup	1	Monitors the link for changes and selects the active port to send the traffics.
Balance-XOR	2	Transmit based on the selected transmit hash policy. Transmit based on the selected transmit hash policy.
Broadcast	3	Sends packets on all slave ports.
LACP	4	Provides load balancing by implementing the Link Aggregation Control Protocol 802.3ad on the slave ports.
LoadBalance	5	While passive mode, it uses the Berkeley Packet Filter (BPF) hash function to select the port that is used to send packets. While active mode, it uses a balancing algorithm to distribute outgoing packets over the available slaves.
Random	6	Randomly selects a slave to send each outgoing packet.

If you are planning to set up network teaming manually by creating interface files, the sample interface files will be underlying under /usr/share/doc/teamd/.

[root@prodsrv1 ~]# ls -lthr /usr/share/doc/teamd/example_ifcfgs/*/*
 -rw-r--r--. 1 root root 293 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/3/ifcfg-team_test0
 -rw-r--r--. 1 root root 107 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/3/ifcfg-eth2
 -rw-r--r--. 1 root root 123 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/3/ifcfg-eth1
 -rw-r--r--. 1 root root 257 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/2/ifcfg-team_test0
 -rw-r--r--. 1 root root  74 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/2/ifcfg-eth2
 -rw-r--r--. 1 root root  74 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/2/ifcfg-eth1
 -rw-r--r--. 1 root root 157 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/1/ifcfg-team_test0
 -rw-r--r--. 1 root root  73 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/1/ifcfg-eth2
 -rw-r--r--. 1 root root  73 Mar 17  2017 /usr/share/doc/teamd/example_ifcfgs/1/ifcfg-eth1
[root@prodsrv1 ~]#

Else, if you plan to create with config options the samples will be under below location.

[root@prodsrv1 ~]# ls -lthr /usr/share/doc/teamd/example_configs/
 total 64K
 -rw-r--r--. 1 root root 244 Mar 17  2017 roundrobin_2.conf
 -rw-r--r--. 1 root root  97 Mar 17  2017 roundrobin.conf
 -rw-r--r--. 1 root root  93 Mar 17  2017 random.conf
 -rw-r--r--. 1 root root 183 Mar 17  2017 loadbalance_3.conf
 -rw-r--r--. 1 root root 140 Mar 17  2017 loadbalance_2.conf
 -rw-r--r--. 1 root root  98 Mar 17  2017 loadbalance_1.conf
 -rw-r--r--. 1 root root 209 Mar 17  2017 lacp_1.conf
 -rw-r--r--. 1 root root  96 Mar 17  2017 broadcast.conf
 -rw-r--r--. 1 root root 318 Mar 17  2017 activebackup_tipc.conf
 -rw-r--r--. 1 root root 285 Mar 17  2017 activebackup_nsna_ping_1.conf
 -rw-r--r--. 1 root root 447 Mar 17  2017 activebackup_multi_lw_1.conf
 -rw-r--r--. 1 root root 241 Mar 17  2017 activebackup_ethtool_3.conf
 -rw-r--r--. 1 root root 212 Mar 17  2017 activebackup_ethtool_2.conf
 -rw-r--r--. 1 root root 194 Mar 17  2017 activebackup_ethtool_1.conf
 -rw-r--r--. 1 root root 465 Mar 17  2017 activebackup_arp_ping_2.conf
 -rw-r--r--. 1 root root 305 Mar 17  2017 activebackup_arp_ping_1.conf
[root@prodsrv1 ~]#

Creating a Team Interface Using nmcli

Listing Physical Interfaces

Let’s start to list the available physical interfaces in our virtual or physical server to create a network teaming interface.

[root@prodsrv1 ~]# nmcli device 
 DEVICE  TYPE      STATE         CONNECTION 
 ens160  ethernet  connected     ens160     
 ens224  ethernet  disconnected  --         
 ens256  ethernet  disconnected  --         
 lo      loopback  unmanaged     --         
[root@prodsrv1 ~]#

# nmcli device     ## Full form
       or
# nmcli dev        ## Short form

Listing the connection profile

To list the connection profiles we can use the full form or short form as well.

# nmcli connection   ## Full form
       or
# nmcli con          ## short form

The output of connections shows it not yet in the upstate.

[root@prodsrv1 ~]# nmcli connection 
 NAME    UUID                                  TYPE      DEVICE 
 ens160  d849fd31-6cdc-4363-9ffe-c8ea9d4af0dc  ethernet  ens160 
 ens224  5824aa6d-68ee-4145-80b9-21403812d856  ethernet  --     
 ens256  cc479e9f-64bf-4e0e-b7a2-80e4eabeff4b  ethernet  --     
[root@prodsrv1 ~]#

Let’s delete the profile of ens224 and ens256 before creating our teaming.

Creating a Team network

First of all, create a team interface in your choice of name. In this guide, I’m about to use team0.

# nmcli connection add con-name team0 ifname team0 type team config '{"runner":{"name": "activebackup"}}'
# nmcli connection

Once after creating team0 interface assign the IP address, gateway, DNS server information etc.

# nmcli connection modify team0 ipv4.addresses 192.168.107.105/24 ipv4.method manual
# nmcli connection modify team0 ipv4.gateway 192.168.107.2 ipv4.dns 192.168.107.2 ipv4.dns-search linuxsysadmins.local
# nmcli connection modify team0 autoconnect yes ipv4.method manual ipv6.method ignore

By running the individual command to assign IP, DNS we can run all in one command as well. To configure the interface in a single command we need to know about what are the parameters in an interface and what is the use of each line. To know more about an interface file entry read Configure a Network interface on CentOS8.

# nmcli connection modify team0 ipv4.addresses 192.168.107.105/24 ipv4.method manual ipv4.gateway 192.168.107.2 ipv4.dns 192.168.107.2 ipv4.dns-search linuxsysadmins.local ipv4.method manual ipv6.method ignore connection.autoconnect yes

Once configured team interface, bring it online.

# nmcli connection up team0

Verify the status. Now we should get the team interface runner mode without any slaves.

[root@prodsrv1 ~]# teamdctl team0 state 
 setup:
   runner: activebackup
 runner:
   active port: 
[root@prodsrv1 ~]#

Let’s start to create slave interfaces and add it to master team interface team0.

Creating and adding Slave Interfaces

To add a slave interface into existing team interface we can use below commands.

# nmcli connection add type ethernet con-name team0-slave1 ifname ens224 master team0
# nmcli connection add type ethernet con-name team0-slave2 ifname ens256 master team0

Output for your reference

[root@prodsrv1 ~]# nmcli connection add type ethernet con-name team0-slave1 ifname ens224 master team0 
 Connection 'team0-slave1' (58c5f0af-2f95-43e5-b0b4-07965376c748) successfully added.
[root@prodsrv1 ~]# nmcli connection add type ethernet con-name team0-slave2 ifname ens256 master team0 
 Connection 'team0-slave2' (095f7444-fbf7-4727-91e5-acc652b74881) successfully added.
[root@prodsrv1 ~]#

Once created with master team interface team0 and slave interfaces print using “nmcli connection“.

[root@prodsrv1 ~]# nmcli connection 
 NAME          UUID                                  TYPE      DEVICE 
 ens160        d849fd31-6cdc-4363-9ffe-c8ea9d4af0dc  ethernet  ens160 
 team0         3394c156-e6fd-41ae-be40-3b52ca1b4bdf  team      team0  
 team0-slave1  58c5f0af-2f95-43e5-b0b4-07965376c748  ethernet  ens224 
 team0-slave2  095f7444-fbf7-4727-91e5-acc652b74881  ethernet  ens256 
[root@prodsrv1 ~]#

Bring up the Slave interfaces.

By following bring up all the slave interfaces once after configuring IP for master team interface (team0).

[root@prodsrv1 ~]# nmcli connection up team0-slave1 
 Connection successfully activated (D-Bus active path: /org/freedesktop/NetworkManager/ActiveConnection/6)
[root@prodsrv1 ~]#
 
[root@prodsrv1 ~]# nmcli connection up team0-slave2 
 Connection successfully activated (D-Bus active path: /org/freedesktop/NetworkManager/ActiveConnection/7)
[root@prodsrv1 ~]#

Verify the Status

Once after creating the teaming interface check for the status using

# teamdctl team0 state

The output clearly shows Link of both interfaces ens224 and ens256 are in the upstate.

[root@prodsrv1 ~]# teamdctl team0 state
 setup:
   runner: activebackup
 ports:
   ens224
     link watches:
       link summary: up
       instance[link_watch_0]:
         name: ethtool
         link: up
         down count: 0
   ens256
     link watches:
       link summary: up
       instance[link_watch_0]:
         name: ethtool
         link: up
         down count: 0
 runner:
   active port: ens224
[root@prodsrv1 ~]#

Verify the Availability of Interfaces

Disconnect any one of the physical interface ens224 or ens256 from the team0 interface and check the availability.

# nmcli device disconnect ens224
# teamdctl team0 state

Once again verify the status of team0 interface.

[root@prodsrv1 ~]# nmcli device disconnect ens224 
 Device 'ens224' successfully disconnected.
[root@prodsrv1 ~]#

[root@prodsrv1 ~]# teamdctl team0 state
 setup:
   runner: activebackup
 ports:
   ens256
     link watches:
       link summary: up
       instance[link_watch_0]:
         name: ethtool
         link: up
         down count: 0
 runner:
   active port: ens256
[root@prodsrv1 ~]#

Below is the proof of Interface status and ping status from source to the destination. While running a continuous ping and bringing down any one of the interfaces will not break the teaming.

That’s it we have successfully created a Team interface on a systemd based Linux server.

Conclusion

To create a network teaming we need to follow only 4 steps, By following the above guide we are good with creating our first network teaming interface in any flavour of Linux servers. Subscribe to newsletter for more networking-related articles.

jos on How to Configure Sudo email alertMarch 4, 2022
Hi! Some way to send commands write as root by email? Thanks!
paul on Monitoring Proxmox with InfluxDB and Grafana in 4 Easy stepsMarch 1, 2022
Tried 12966 per your example, got error "failed to import dashboard."
Anonymous on Migrate existing Iptables to Nftables in RHEL8/CentOSFebruary 27, 2022
I followed the procedures and successfully loaded some weird 8081 port configurations needed for the app, however I discovered it…
sy on Install Ansible AWX on Kubernetes in 5 minutesFebruary 17, 2022
tried many times - followed all steps but was unable to setup custom context I always get -> The connection…
Faraz Ahmed on Prometheus with Grafana for Linux server monitoring in 5 easy stepsFebruary 7, 2022
Hello Babin Lonston, Your tutorial is very helpful to me to Install the prometheus and nodeexporter on the server. Can…