How to limit bandwidth and allow multiple downloads when downloading a file?

Question

I have this code, that is working when only 1 download is running at a time

using (System.IO.FileStream fs = System.IO.File.OpenRead(@"C:\HugeFile.GBD"))
{
    using (System.IO.BinaryReader br = new System.IO.BinaryReader(fs))
    {
        Response.AddHeader("Cache-control", "private");
        Response.AddHeader("Content-Type", "application/octet-stream");
        Response.AddHeader("Content-Length", fs.Length.ToString());
        Response.AddHeader("Content-Disposition", "filename=HugeFile.GBD");
        Response.Flush();
        float kbs = 20f;

        while (fs.Position < fs.Length)
        {
            if (!Response.IsClientConnected)
                break;
            byte[] bytes = br.ReadBytes((int)Math.Truncate(1024 * kbs));
            char[] c = UTF8Encoding.Default.GetChars(bytes);

            Response.Write(c, 0, c.Length);


            Response.Flush();
            System.Threading.Thread.Sleep(1000);
        }
        Response.Flush();
    }
}

But if I make two simultaneous connections (start a second download on the same browser) the second one doesn't execute until the first finishes.

Using a thread or a task results in an error when adding the Headers to the Response...

How can I make it so I can execute 2+ downloads on the same browser at the same time?

Answer 1

It seems that Google Chrome handles a download with a unique URL, so when I try to access the same URL it doesn't even run the action.

By adding anything to the URL, ex:

home/download?x=1
home/download?x=2
home/download?x=3

Chrome will think it is a different download and get it. So, to download a file, even if a different generated file every time, we need to change the URL.

And as for the throttle I created a FileThrottleResult which inherits from FilePathResult:

using System;
using System.IO;
using System.Threading;
using System.Web.Mvc;

namespace Mvc3Test
{
    public class FileThrottleResult : FilePathResult
    {
        float rate = 0;

        /// <summary>
        /// 
        /// </summary>
        /// <param name="fileName"></param>
        /// <param name="contentType"></param>
        /// <param name="rate">Kbps</param>
        public FileThrottleResult(string fileName, string contentType, float rate)
            : base(fileName, contentType)
        {
            if (!File.Exists(fileName))
            {
                throw new ArgumentNullException("fileName");
            }
            this.rate = rate;
        }

        protected override void WriteFile(System.Web.HttpResponseBase response)
        {
            int _bufferSize = (int)Math.Round(1024 * this.rate);
            byte[] buffer = new byte[_bufferSize];

            Stream outputStream = response.OutputStream;

            using (var stream = File.OpenRead(FileName))
            {
                response.AddHeader("Cache-control", "private");
                response.AddHeader("Content-Type", "application/octet-stream");
                response.AddHeader("Content-Length", stream.Length.ToString());
                response.AddHeader("Content-Disposition", String.Format("filename={0}", new FileInfo(FileName).Name));
                response.Flush();

                while (true)
                {
                    if (!response.IsClientConnected)
                        break;

                    int bytesRead = stream.Read(buffer, 0, _bufferSize);
                    if (bytesRead == 0)
                        break;

                    outputStream.Write(buffer, 0, bytesRead);
                    response.Flush();
                    Thread.Sleep(1000);
                }
            }
        }
    }
}

Usage:

public FileThrottleResult Download()
{
    string testFile = @"C:\hugefile.zip";
    return new FileThrottleResult(testFile, "application/octet-stream", 200);
}

This will successfully limit the bandwidth and will not have problems with simultaneous downloads of the same file, as long as you trick the browser with a "different" url.