Why is this in my firewall-start script!? logger -t $(basename $0) $1

[Deleted member 27741]

I added a few iptables rules yesterday, they did not work. Turns out, I had the command-

logger -t $(basename $0) $1

at the top of my firewall-start script. Of course, that kept the other rules from running because apparently that command hangs. I can't remember for the life of my why I would put that command in firewall-start. Does anyone know why it would be there? Does it look like a snippit of any script you know of, because I did not write the code!

 

[john9527]

I added a few iptables rules yesterday, they did not work. Turns out, I had the command-

logger -t $(basename $0) $1

at the top of my firewall-start script. Of course, that kept the other rules from running because apparently that command hangs. I can't remember for the life of my why I would put that command in firewall-start. Does anyone know why it would be there? Does it look like a snippit of any script you know of, because I did not write the code!

A lot of people use that, or similar, to write a syslog entry for executing the script
$(basename $0) = the name of the script
$1 = passed parameter

It shouldn't hang, just write to the syslog...

Re your script not running....how did you edit it? Are you sure you didn't introduce DOS CR-LF sequences when you added the lines? Since you are running my fork, you can run 'dos2unix' to fix it up....
dos2unix /jffs/scripts/firewall-start

 

[ColinTaylor]

logger -t $(basename $0) $1

There is a "gotcha" with that command (that keeps tripping me up).

If you're running it from the command line (i.e. for testing) then you must supply a parameter for $1. If you don't then the logger command will hang waiting for input.

Maybe I should change it to:

logger -t $(basename $0) ${1:-missing}

or is that just getting overly complicated?

 

[Deleted member 27741]

Looks like it works like this-
logger -t $(basename $0) $

Take out the 1 out and it is all good, writes to syslog and runs without hanging. Thanks, guys!


********************Edit*******************
Changed it to this-
logger -t $(basename $0) THIS SCRIPT HAS RUN

In the log it looks like this-
firewall-start: THIS SCRIPT HAS RUN

I guess it was the 1 that was the problem, leaving just the $ looked odd in the log so I changed it to something more descriptive.

Now I am kicking myself in the head for not having my firewall-start running correctly for months now. Guess I should have actually LOOKED at iptables to make sure the script was run. Live and learn.

 

[ColinTaylor]

Well the purpose of the "$1" was so that you could see what the value of the parameter was in the syslog. If you change it to "$" then it's meaningless. You might as well put something a bit more user friendly there:

logger -t $(basename $0) Starting

 

[Deleted member 27741]

If someone knows how to get the command working with the paramater value, that would be cool!

 

[ColinTaylor]

Thinking about it a bit more the answer is obvious. Use something like this:

logger -t $(basename $0) $1 "Starting"

That way the logger command will always have something to print even if the user forgets to supply a value for $1.

 

[john9527]

Or something like this....

logger -t $(basename $0) "Starting, Parameter: $1"