Question

I have a string in a bash script that contains a line of a log entry such as this:

Oct 24 12:37:45 10.224.0.2/10.224.0.2 14671: Oct 24 2012 12:37:44.583 BST: %SEC_LOGIN-4-LOGIN_FAILED: Login failed [user: root] [Source: 10.224.0.58] [localport: 22] [Reason: Login Authentication Failed] at 12:37:44 BST Wed Oct 24 2012

To clarify; the first IP listed there “10.224.0.2” was the machine the submitted this log entry, of a failed login attempt. Someone tried to log in, and failed, from the machine at the 2nd IP address in the log entry, “10.224.0.58”.

I wish to replace the first occurrence of the IP address “10.224.0.2” with the host name of that machine, as you can see presently is is “IPADDRESS/IPADDRESS” which is useless having the same info twice. So here, I would like to grep (or similar) out the first IP and then pass it to something like the host command to get the reverse host and replace it in the log output.

I would like to repeat this for the 2nd IP “10.224.0.58”. I would like to find this IP and also replace it with the host name.

It’s not just those two specific IP address though, any IP address. So I want to search for 4 integers between 1 and 3, separated by 3 full stops ‘.’

Is regex the way forward here, or is that over complicating the issue?

Many thanks.

Answer 1

Replace a fixed IP address with a host name:

$ cat log | sed -r 's/10\.224\.0\.2/example.com/g'

Replace all IP addresses with a host name:

$ cat log | sed -r 's/[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}/example.com/g'

If you want to call an external program, it’s easy to do that using Perl (just replace host with your lookup tool):

$ cat log | perl -pe 's/(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})/`host \1`/ge'

Hopefully this is enough to get you started.