Issue

Logging is useful for determining:

• Admin state (whether the port/bridge/bond up or down)
• Physical state (whether the connection for the port/bridge/bond up or down)

... and the current level is not enough in /var/log/.

Environment

All versions of Cumulus Linux.

Resolution

• Utilizing ip monitor
• Running in the background
• Automatically logging on boot
• Looking at the log output
• Admin/physical state troubleshooting table

Utilizing ip monitor

Use the Linux ip monitor tool to monitor link state.

cumulus@switch:~$ ip -timestamp monitor link

• The -timestamp command allows us to timestamp each message
• The 'link' keyword specifies to only pay attention to link changes

Running in the Background

What if I want to run it in the background and redirect it to a log?

Cumulus Linux is Linux so use the & and > to redirect the output:

cumulus@switch:~$ sudo sh -c "ip -timestamp monitor link > /var/log/link.log" &

The redirect (>) in the previous example requires you to use the quotation marks around the two commands subjected to the redirection clause (>) so that both commands are run as the root user. If using the root account (disabled by default on Cumulus Linux 2.0 and later) there is no need:

root@switch:~# ip -timestamp monitor link > /var/log/link.log &

Automatically Logging on Boot

What if I automatically want to log link activity, even if the switch is rebooted?

Utilize the rc.local startup script located at /etc/rc.local. Here rc.local is already configured on a Cumulus Linux switch:

    
    cumulus@switch:~$ cat /etc/rc.local
    #!/bin/sh -e
    #
    # rc.local
    #
    # This script is executed at the end of each multiuser runlevel.
    # Make sure that the script will "exit 0" on success or any other
    # value on error.
    #
    # In order to enable or disable this script just change the execution
    # bits.
    #
    # By default this script does nothing.
    ip -timestamp monitor link > /var/log/link.log
    
    exit 0
    
    cumulus@switch:~$
    

Looking at the Log Output

What does admin state vs. physical state look like in the log?

Let's go ahead and turn a port off and on and look at the log.

cumulus@switch:~$ sudo ip link set swp1 down

cumulus@switch:~$ sudo ip link set swp1 up

cumulus@switch:~$ cat /var/log/link.log
Timestamp: Wed Jul 16 19:00:35 2014 350805 usec
3: swp1: <BROADCAST,MULTICAST,SLAVE> mtu 1500 qdisc pfifo_fast master bond24 state DOWN
    link/ether 44:38:39:00:25:d9 brd ff:ff:ff:ff:ff:ff
Timestamp: Wed Jul 16 19:00:37 2014 840831 usec
3: swp1: <NO-CARRIER,BROADCAST,MULTICAST,SLAVE,UP> mtu 1500 qdisc pfifo_fast master bond24 state DOWN
    link/ether 44:38:39:00:25:d9 brd ff:ff:ff:ff:ff:ff

Admin/Physical State Troubleshooting Table

Now let's look at every possible scenario to make troubleshooting easier.

swp1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT qlen 500

swp1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast state DOWN mode DEFAULT qlen 500

swp1: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_fast state DOWN mode DEFAULT qlen 500

swp1: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_fast state DOWN mode DEFAULT qlen 500

Warning Notice how Admin down and down/down look identical.

For additional troubleshooting of physical ports, read the documentation on duplex, speed and auto-negotiation settings.