We knew how many items (units) are there to process, every once in a while we could check how many have been alread processed (by looking up in a database).

The same can be used even if the processing is done in the same program where we have the monitoring.We just need to calculate how much time - on average - took to process one unit (dividing the elapsed time by the number of units done) and then multiplying it by the number of units remaining.

It won't be exact, but can give you a better feeling to see how much more you have to wait.

examples/python/time_left.py

import time
import datetime

class Timer(object):
    def __init__(self, total):
        self.start = datetime.datetime.now()
        self.total = total

    def remains(self, done):
        now  = datetime.datetime.now()
        #print(now-start)  # elapsed time
        left = (self.total - done) * (now - self.start) / done
        sec = int(left.total_seconds())
        if sec < 60:
           return "{} seconds".format(sec)
        else:
           return "{} minutes".format(int(sec / 60))


# There are 12987 units in this task
t = Timer(12987)

# do some work

time.sleep(1)
# after some time passed, or after some units were processed
# eg.  37 units, calculate, and print the remaining time:

print(t.remains(37))

# let the process continue and once in a while print the remining time.