Python shell script, displays memory usage for a user's processes. Also displays warning message if the usage is over a configurable amount.

watch_memory_usage.py

#!/usr/bin/python

import subprocess
from datetime import datetime
import time
import sys
import itertools

# Configurable Values
MEM_LIMIT = 600000
delay = 10 # polling interval for running the following command:
cmd = "ps -u CHANGEME_USERNAME -o rss,etime,pid,command | awk '{ SUM += $1 } END { print SUM }';"


# generic highlighting function, poached from somewhere / not my own
def hilite(string, status, bold):
  if not sys.stdout.isatty():
    return string

  attr = []
  if status is True:
    # green
    attr.append('32')
  elif status is False:
    # red
    attr.append('31')
  elif int(status) > 0:
    attr.append(status)

  if bold:
    attr.append('1')
  return '\x1b[%sm%s\x1b[0m' % (';'.join(attr), string)


# Creates the output, which is a combination of timestamp, memory usage, and warning msg if appropriate.
def make_output(text):
  msgs = list()

  msg = text[:-1] # trim off trailing \r\n
  time_str = datetime.now().strftime("%Y-%d-%m, %H:%M:%S")

  if int(msg) > MEM_LIMIT:
    msgs.append(hilite("%s: *** Warning: mem usage is above %s ***" % (time_str, MEM_LIMIT), False, True))

  hl_msg = hilite(msg, '33', False)
  msgs.append("%s: Current mem usage: %s" % (hilite(time_str, True, True), hl_msg))
  return "\r\n".join(msgs)


# Main uses iter() with a lambda, executing the `cmd` value via subprocess.Popen. Then it sleeps for the specified interval.
def main():
  for (out, err) in iter(lambda: subprocess.Popen([cmd], stdout=subprocess.PIPE, shell=True).communicate(), False):
    print(make_output(out))
    time.sleep(delay)


if __name__ == '__main__':
  main()