Extract last 5 mins of information from a log file

Since you will need extensive text processing and time manipulations, this should be done using awk, but will still need the external date command to work.

The following awk program (let's call it parse_log_range.awk) will work:

#!/bin/awk -f
BEGIN{
    "date +%s" | getline now
    close("date +%s")
}

{
    n=match($0,/\[[^]]+\]/)
    tst=substr($0,n+1,RLENGTH-2)

    gsub(/\//," ",tst)  # replace all '/'
    sub(/:/," ",tst)    # replace first ':'
    
    cmd="date +%s -d \"" tst "\""
    cmd | getline tst_s; close(cmd)

    if (now-tst_s<max_age) print
}

In the beginning, it will launch the command date +%s in the shell to query the current time and store it in a variable now.

For each line of the log file, this will

• extract the timestamp as being the element in between [ ... ]
• remove the / and the first : to make the format understandable to date
• create a shell command of the form date +%s -d edited timestamp which will convert the timestamp into "seconds since the epoch"
• run this command in the shell, and read the result into a variable tst_s

It will then check if the timestamp is less than max_age away from now, where max_age is a timespan you have to specify (in your example, 5 minutes = 300 seconds).

Run the command as

awk -v max_age=300 -f parse_log_range.awk logfile.txt