The disklog.sh script accepts the log file path as a command-line argument or uses the default log file. The -e $logfile checks whether the file exists or not. If the log file does not exist, it is initialized with a column header. The list of remote machine IP addresses can be hardcoded in IP_LIST, delimited with spaces, or the nmap command can be used to scan the network for available nodes. If you use the nmap call, adjust the IP address range for your network.

A for loop iterates through each of the IP addresses. The ssh application sends the df -H command to each node to retrieve the disk usage information. The df output is stored in a temporary file. A while loop reads that file line by line and invokes awk to extract the relevant data and output it. An egrep command extracts the percent full value and strips %. If this value is less than 80, the line is marked SAFE, else it's marked ALERT. The entire output string must be redirected to the log file. Hence, the for loop is enclosed in a subshell () and the standard output is redirected to the log file.