Automatically grab the contents of a web page on a timer and discard the results. Useful for debugging servers or pretending your website's more popular than it is (sure we can justify the extra dyno now that we're seeing a bump in traffic!).

gravepinger

#!/bin/sh
#
#    gravepinger - Pings a website on a timer.
#
test -n "$2" || {
    echo "Usage: gravepinger <target> <frequency>" >&2
    echo "" >&2
    echo "Invokes wget on <target> every <frequency> timespan." >&2
    echo "Target can be any web page, e.g. \`http://chriscummins.cc'." >&2
    echo "Timespan may be any number suffixed with \`s' for seconds," >&2
    echo "\`m' for minutes, \`h' for hours or \`d' for days." >&2
    exit 1
}

target="$1"
frequency="$2"

# Written by Hugo Duksis. See: https://gist.github.com/duksis/3891185
urlencode() {
    echo -n "$1" | perl -MURI::Escape -ne 'print uri_escape($_)'
}

logfile=~/.gravepinger.$(urlencode "$target").log
tmpfile=/tmp/gravepinger

touch "$logfile"                                     # Create logfile

while true; do
    echo "$(date)  Ping" >> "$logfile"
    wget "$target" -O "$tmpfile" >/dev/null 2>&1 \
        || echo "$(date)  Ping failed" >> "$logfile" # Download target
    rm -f "$tmpfile"                                 # Remove downloaded file

    tail -n-99 "$logfile" > "$logfile.tmp"           # Rotate logfile
    mv "$logfile.tmp" "$logfile"

    sleep "$frequency"                               # Rinse and repeat...
done