Skip to main content

Limiting Bitrate and Network Throttling

We may limit incoming or outcoming data rates to/from our infrastructure to maintain the stability of our service for customers. Bitrate limitation is an action to limit the number of bits that can be passed through a transmission channel in a period of time. Network throttling is an intentional action to slow down transmission speed in a network channel. It is not only about limiting bitrate but also limiting the allowed number of requests in a period of time. There are several tools and techniques that can be used to apply bitrate limitation and network throttling.


It is an easy-to-use tool for Linux and is already in the package repository. It can limit the bit rate that can be achieved by network interfaces in the system. We can install it by running the following command.

apt install wondershaper

We can choose an interface to have a limitation either or both on download and upload.

wondershaper <interface-name> <download-rate-in-bps> <upload-rate-in-bps>

For example, the following code will limit eth0 to 40k bits/s of download rate and 20k bits/s of upload rate.

wondershaper eth0 40000 20000

To clear the setting, we can run the following command.

wondershaper clear eth0


Based on its manual page, Trickle is a tool to limit the download or upload rate of any applications that utilize the socket interface. We can read more information about this tool here. We need to remember that only programs with dynamic linking can work with Trickle. We can test whether a binary can work with Trickle by the following command.

ldd $(which [binary]) | grep
For example,
ldd $(which wget) | grep

If it shows the location of, it means the binary can work with Trickle. For instance, if we want to limit the download rate of the wget tool, we can use the following example.

trickle -d [rate in KB/s] wget [protocol://file-URL] 


Nginx is actually a proxy server that provides plenty of features to help us serve our web-based services such as the request rate and bandwidth limiting capability. This feature is used to prevent DDoS attacks by preventing our server to handle too many requests. It needs a defined key to differentiate a client from another, a memory zone to keep states of all the keys, and a rate-setting that states the number of requests per second or minute. The examples are as follows.

http {
    limit_req_zone $binary_remote_addr zone=one:10m rate=1r/s;

    server {
        location /search/ {
            limit_req zone=one;
        location /search2/ {
        	# if request rate > specified rate, 5 requests are queued
            limit_req zone=one burst=5; 
        location /search3/ {
        	# 5 excessive requests are served immediately
            limit_req zone=one burst=5 nodelay; 
        location /search4/ {
        	# 3 excessive requests are served immediately, 2 requests are delayed
            limit_req zone=one burst=5 delay=3; 

Bandwidth limiting is applied per connection while Nginx originally allows multiple connections. It means if we want the exact bitrate value to be provided for specific addresses, we can set Nginx to allow only one connection.

http {
    limit_conn_zone $binary_remote_address zone=addr:10m;
    # mapping for dynamic configuration
    map $ssl_protocol $response_rate {
      "TLSv1.1" 10k;
      "TLSv1.2" 100k;
      "TLSv1.3" 1000k;

    server {
        root /www/data;
        limit_conn addr 5;

        location /download/ {
            limit_conn       addr 1;
            # rate limiting is applied after 1MB data is passed
            limit_rate_after 1m; 
            limit_rate       50k;
        # dynamic configuration
        location /secure/ {
            limit_rate       $response_rate;

We can get more information from Nginx documentation.


Popular posts from this blog

Setting Up Next.js Project With ESLint, Typescript, and AirBnB Configuration

If we initiate a Next.js project using the  create-next-app tool, our project will be included with ESLint configuration that we can apply using yarn run lint . By default, the tool installs eslint-config-next and extends next/core-web-vitals in the ESLint configuration. The Next.js configuration has been integrated with linting rules for React and several other libraries and tools. yarn create next-app --typescript For additional configuration such as AirBnB, it is also possible. First, we need to install the peer dependencies of eslint-config-airbnb . We also add support for Typescript using eslint-config-airbnb-typescript . yarn add --dev eslint-config-airbnb eslint-plugin-import eslint-plugin-jsx-a11y eslint-plugin-react eslint-plugin-react-hooks yarn add --dev eslint-config-airbnb-typescript @typescript-eslint/eslint-plugin @typescript-eslint/parser After that, we can update the .eslintrc.json file for the new configuration. { "extends": [ "airb

Rangkaian Sensor Infrared dengan Photo Dioda

Keunggulan photodioda dibandingkan LDR adalah photodioda lebih tidak rentan terhadap noise karena hanya menerima sinar infrared, sedangkan LDR menerima seluruh cahaya yang ada termasuk infrared. Rangkaian yang akan kita gunakan adalah seperti gambar di bawah ini. Pada saat intensitas Infrared yang diterima Photodiode besar maka tahanan Photodiode menjadi kecil, sedangkan jika intensitas Infrared yang diterima Photodiode kecil maka tahanan yang dimiliki photodiode besar. Jika  tahanan photodiode kecil  maka tegangan  V- akan kecil . Misal tahanan photodiode mengecil menjadi 10kOhm. Maka dengan teorema pembagi tegangan: V- = Rrx/(Rrx + R2) x Vcc V- = 10 / (10+10) x Vcc V- = (1/2) x 5 Volt V- = 2.5 Volt Sedangkan jika  tahanan photodiode besar  maka tegangan  V- akan besar  (mendekati nilai Vcc). Misal tahanan photodiode menjadi 150kOhm. Maka dengan teorema pembagi tegangan: V- = Rrx/(Rrx + R2) x Vcc V- = 150 / (150+10) x Vcc V- = (150/160) x 5

Kerusakan pada Motherboard

1. Sering terjadi hang memory tidak cocok --- ganti memory ada virus di harddisk --- scan harddisk over clock --- seting kembali clock prosesor ada bad sector di harddisk --- partisi harddisk dengan benar 2. Pembacaan data menjadi lambat memori tidak cukup --- tambah memori harddisk penuh atau ada virus --- kurangi isi harddisk, scan harddisk, atau ganti hardisk 3. CMOS failure baterai habis --- ganti baterai CMOS seting BIOS berubah --- seting kembali BIOS 4. Tidak bisa booting cache memory rusak --- disable eksternal cache memory di BIOS memori tidak cocok --- ganti memori boot sector pada harddisk rusak --- masukkan operating system baru ada bad sector pada trek awal harddisk --- partisi harddisk 5. Suara bip panjang berkali-kali memori rusak --- periksa kedudukan memori memori tidak cocok --- ganti memori memori tidak masuk slot dengan sempurna --- periksa kembali kedudukan memori 6. Suara bip bagus tetapi tidak ada tampilan / bip dua kali VGA card

Raspberry Pi Bluetooth Connection

Raspberry Pi 3 provides a built-in Bluetooth module. The latest Raspbian has been bundled with tools for enabling Bluetooth connection. The Bluetooth icon will be shown up on the top right corner of the desktop. It's a tool to discover available Bluetooth devices and connect Pi with Bluetooth devices. It is easy to connect any Bluetooth-enabled electronic device with Pi. But, sometimes Pi will fail to connect, especially for Bluetooth device that has no standardized services. From a terminal, we can use the  bluetoothctl tool to scan and connect with a Bluetooth device. You should make sure that the BlueZ protocol stack has been installed by running $ apt-get install bluez Run bluetoothctl to enter the tool command window Turn the power on by running power on (Optional) You can set AutoEnable=true in /etc/bluetooth/main.conf if you want to make the Bluetooth auto power-on after reboot. Run devices to see which devices have been paired Run scan on if your desired d

Pemrograman Paralel dengan CUDA

CUDA ( Compute Unified Device Architecture ) adalah suatu skema yang dibuat oleh NVIDIA agar NVIDIA selaku GPU ( Graphic Processing Unit ) mampu melakukan komputasi tidak hanya untuk pengolahan grafis namun juga untuk tujuan umum. Jadi, dengan CUDA, kita dapat memanfaatkan cukup banyak  processor  yang dimiliki oleh NVIDIA untuk berbagai perhitungan. GPU yang ada  saat ini seperti ATI pun sudah memiliki banyak processor di dalamnya. Pada ATI, skema yang mereka bangun disebut ATI Stream. Saat ini pemrograman paralel menjadi sangat penting karena kebutuhan kemampuan komputasi komputer yang terus meningkat seperti kemampuan multitasking  dan pengolahan grafis yang andal. Metode saat ini dalam peningkatan peforma komputer juga berbeda dengan masa lampau dimana peningkatan clock  dari processor  yang diutamakan. Peningkatan clock  juga dibatasi oleh kemampuan fisik dari perangkat digital yaitu persoalan daya dan panas. Pada 2005 berbagai industri komputer mulai menawakan komputer den

Itachi Uchiha

The Real Hero of Konoha