6
\$\begingroup\$

I'm using Parallel.ForEach to download 500K URLs and I want to monitor the number of URLs that have been successfully downloaded each minute.

        int elapsedMinutes = 0, cnt = 0;
        Parallel.ForEach(list, tuple =>
        {
            var currMinutes = (int) ((DateTime.Now - startTime).TotalMinutes);
            lock (Object)
            {
                if (currMinutes > elapsedMinutes)
                {
                    elapsedMinutes = currMinutes;
                    Console.WriteLine("{0}: Extracted {1} urls", DateTime.Now, cnt);
                }
            }

            //download urls ...

            Interlocked.Increment(ref cnt);
        });

It looks fine now but the lock statement part seems a bit ugly to me. Is there a better way to achieve this?

\$\endgroup\$
9
\$\begingroup\$

Instead of checking the time from each thread on each iteration, I would use a Timer:

int count = 0;

TimeSpan reportPeriod = TimeSpan.FromMinutes(0.1);

using (new Timer(
    _ => Console.WriteLine("{0}: Extracted {1} urls", DateTime.Now, count),
    null, reportPeriod, reportPeriod))
{
    Parallel.ForEach(
        list, tuple =>
        {
            //download urls ...

            Interlocked.Increment(ref count);
        });
}

Console.WriteLine("{0}: Done", DateTime.Now);

I have also renamed cnt to count, those two characters are not worth it to make your code less readable.

\$\endgroup\$
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.