Posts Tagged ‘shell scripting’

How to install BASH and use shell scripting on Windows ?

Thursday, June 26th, 2025

install-bash-on-windows-run-and-use-shellscripting-on-windows-howto

Bash (Bourne Again SHell) is definitely a technology that will stay for years to come its simplicity and multi-platoformness is a factor that will definitely continue for many years thus even though it is mostly used on Linux / BSD / Unix, its application on Windows OS-es nowadays is perhaps increasing. Hence since so many people use Winodws nowdays (for work) it is really useful to have Bash set-up on Windows host machine.
In this article, I'll shortly explain how this is done, the article will not have anything too much interesting for the advanced admin or dev ops guy, but I hope people who are entering the business of system administration and high level computing and still orienting might benefit from it.

To install and use Bash shell terminal in Windows there are at least 3 ways:

  • Use Git Bash (Download and install it directly precompiled on WIndows)
  • Use Windows WSL emulation (install some Linux distro)
  • Use Virtualbox / Vagrant / VMware / Hyper-V emulation and install VM from public ISO image.

As a Free Software Lover, I would recommend and  always prefer to use the Free Software alternative if that is possible and thanksfully usually I use and install Git Bash or completely install Cygwin (Full set of Linux tools to run like native on Windows together with  Mobaxterm) together.

 

1. Installing Git Bash on Windows (uses MinGW Minimalist GNU for Windows)

Some might prefer to not use Microsoft for managing their bash especially the more freedom in mind people who like GNU and Free software and people.

MinGW is well known among free and open source enthusiasts.
It includes a port of the GNU Compiler Collection (GCC), GNU Binutils for Windows (assembler, linker, archive manager), a set of freely distributable Windows specific header files and static import libraries which enable the use of the Windows API, a Windows native build of the GNU Project's GNU Debugger, and miscellaneous utilities.

MinGW does not rely on third-party C runtime dynamic-link library (DLL) files, and because the runtime libraries are not distributed using the GNU General Public License (GPL), it is not necessary to distribute the source code with the programs produced, unless a GPL library is used elsewhere in the program.

 

MinGW can be run either on the native Microsoft Windows platform, cross-hosted on Linux (or other Unix), or "cross-native" on Cygwin.


To install Bash via Git, you can use Git for Windows, which includes Git Bash — a lightweight Bash emulator.


Steps to Install Git Bash on Windows
 

a. Download Git for Windows

Go to the official Git website:

https://git-scm.com/download/win

The download should start automatically.

b. Run the Installer

  • Open the downloaded .exe file
  • Follow the installation prompts

Recommended Settings:

  • Select components: Keep default
  • Editor: Choose your preferred text editor (e.g., Notepad++ or Vim)
  • Adjust PATH environment: Choose “Git from the command line and also from 3rd-party software”
  • Choose SSH executable: Use Built-in OpenSSH
  • Choose HTTPS transport backend: Use the default (OpenSSL)
  • Configure line endings: Select “Checkout Windows-style, commit Unix-style line endings”
  • Terminal emulator: Choose “Use MinTTY (the default terminal)”

Click Next through the remaining steps and then Install.

c. Launch Git Bash

After installation:

  • Press Windows key, type "Git Bash"
  • Click to launch the terminal

Now you're using a Bash shell on Windows.

Perhaps most common way is to use Windows Subsystem for Linux (WSL), people follow. WSL is a technology which is native Windows but gives MS Windows the opportunity to act in a way similar to docker containers. WSL lets you run a full Linux environment (including Bash) directly on Windows without using a virtual machine and is really fast and easy on Machine system resources. 


 2. Installing WSL bash easy from Windows 10 / 11 using  Win GUI menus


Steps to install WSL on Windows 10 / 11

Microsoft has since only continued to improve its Windows Subsystem for Linux, and an update in a Windows 10 preview build back in mid-2020 made it easier to install Bash.

That method also works the same as on Win 10 as well as on Win 11.
To install Bash shell emulation, hence open Windows Terminal as an admin user. You can do this by right-clicking the Windows icon and selecting “Windows Terminal (Admin)” from the power user menu.

(If you’re on Windows 10, you should see it listed as “Windows Powershell (Admin)” in the menu.)

 

windows-run-powershell-from-start-menu-screenshot

 


To complete WSL install with Virtualized Ubuntu OS

In Windows Terminal, run this command:

PS C:\Users\MyUser> wsl –install

Once everything is downloaded needed to run WSL emulation and Ubuntu Linux distribution,  Restart the PC.

Once your PC rebooted, installation will continue automatically.

After Ubuntu installed successfully, you’ll next be prompted to create a username and password and Ubuntu will fire up, and you will have your bash in Windows

 

Install-WSL-linux-subsystem-for-windows-from-powershell-prompt-screenshot


a. Enabling and Intalling BASH via command line (if WSL Linux subsystem for Windows is not enabled on Windows


It might be your Windows has no configured Windows Subsystem for Linux, hence if that is the case you will need to enable it following below few steps.

b. Enable WSL via dism.exe cmd

Open PowerShell as Administrator and run:

Powershell

PS C:\Users\MyUser> wsl –install

This installs WSL 2 and a default Linux distribution (like Ubuntu).

If you're on Windows 10 or on a PC where whoever installed the OS has not installed the Win Subsystem for Linux, you may need to manually enable WSL:

Launch Powershell

PS C:\Users\MyUser> dism.exe /online /enable-feature /featurename:Microsoft-Windows-Subsystem-Linux /all /norestart

PS C:\Users\MyUser> dism.exe /online /enable-feature /featurename:VirtualMachinePlatform /all /norestart

Then restart your computer and run:

from the Windows Magnifier run Powershell and type in PS1 prompt:

PS C:\Users\MyUser> wsl –set-default-version 2

c. Installing other Linux Distribution (different from Ubuntu)

If not already installed during wsl –install, open the Microsoft Store and search for:

  • Ubuntu
  • Debian
  • Kali Linux
  • etc.

Click Install on the one you want.

d. Launching WSL / Bash terminal

Once installed:

  • Open Start Menu
  • Search for your Linux distro you just installed (e.g., “Ubuntu”)
  • Launch it

This opens a Bash shell where you can run Linux commands, like in regular Linux but on your Microsoft Windows OS.
 

Sum it up

What we learned is how to install bash via Bash Git and start using it to have more hybrid environment Windows / Linux. The article explained the two main methods using GIt Bash and using embedded Windows emulator WSL with an emulated Linux distro.

Enjoy ! 🙂

 

 


 

Linux: Generating Web statistics from Old Apache logs with Webalizer

Thursday, July 25th, 2013

Webalizer generate and visualize in web page statistics of old websites howto webalizer static html google analytics like statistics on linux logo

Often it happens, that some old hosted websites were created in a way so no Web Statistics are available. Almost all modern created websites nowadays are already set to use Google AnalyticsAnyhow every now and then I stumble on hosting clients whose websites creator didn't thought on how to track how many hits or unique visitors site gets in a month / year etc.
 Thanksfully this is solvable by good "uncle" admin with help with of Webalizer (with custom configuration) and a little bit of shell scripting.

The idea is simple, we take the old website logs located in lets say 
/var/log/apache2/www.website-access.log*,
move files to some custom created new directory lets say /root/www.website-access-logs/ and then configure webalizer to read and generate statistics based on log in there.

For the purpose, we have to have webalizer installed on Linux system. In my case this is Debian GNU / Linux.

For those who hear of Webalizer for first time here is short package description:

debian:~# apt-cache show webalizer|grep -i description -A 2

Description-en: web server log analysis program
The Webalizer was designed to scan web server log files in various formats
and produce usage statistics in HTML format for viewing through a browser.

 If webalizer is not installed still install it with:

debian:~# apt-get install --yes webalizer
...
.....

Then make backup copy of original / default webalizer.conf (very important step especially if server is already processing Apache log files with some custom webalizer configuration:

debian:~# cp -rpf /etc/webalizer/webalizer.conf /etc/webalizer/webalizer.conf.orig

Next step is to copy webalizer.conf with a name reminding of website of which logs will be processed, e.g.:

debian:~# cp -rpf /etc/webalizer/webalizer.conf /etc/webalizer/www.website-webalizer.conf

In www.website-webalizer.conf config file its necessary to edit at least 4 variables:

LogFile /var/log/apache2/access.log
OutputDir /var/www
#Incremental no
ReportTitle Usage statistics for

 Make sure after modifying 3 vars read something like:  
LogFile /root/www.website/access_log_merged_1.log
OutputDir /var/www/www.website
Incremental yes
ReportTitle Usage statistics for Your-Website-Host-Name.com

Next create /root/www.website and /var/www/www.website, then copy all files you need to process from /var/log/apache2/www.website* to /root/www.website:

debian:~# mkdir -p /root/www.website
debian:~# cp -rpf /var/log/apache2/www.website* /root/www.website

On Debian Apache uses logrotate to archive old log files, so all logs except www.website-access.log and wwww.website-access.log.1 are gzipped:

debian:~#  cd /root/www.website
debian:~# ls 
www.website-access.log.10.gz
www.website-access.log.11.gz
www.website-access.log.12.gz
www.website-access.log.13.gz
www.website-access.log.14.gz
www.website-access.log.15.gz
www.website-access.log.16.gz
www.website-access.log.17.gz
www.website-access.log.18.gz
www.website-access.log.19.gz
www.website-access.log.20.gz
...
 

Then we have to un-gzip zipped logs and create one merged file from all of them ready to be red later by Webalizer. To do so I use a tiny shell script like so:

for n in {52..1}; do gzip -d www.dobrudzhatour.net-access.log.$n.gz; done
for n in {52..1}; do cat www.dobrudzhatour.net-access.log.$n >> access_log_merged_1.log;
done

First look de-gzips and second one does create a merged file from all with name access_merged_1.log The range of log files in my case is from www.website-access.log.1 to www.website-access.log.52, thus I have in loop back number counting from 52 to 1.

Once access_log_merged_1.log is ready we can run webalizer to process file (Incremental) and generate all time statistics for www.website:

debian:~# webalizer -c /etc/webalizer/webalizer-www.website-webalizer.conf

Webalizer V2.01-10 (Linux 2.6.32-27-server) locale: en_US.UTF-8
Using logfile /root/www.website/access_log_merged_1.log (clf)
Using default GeoIP database Creating output in /var/www/webalizer-www.website
Hostname for reports is 'debian'
Reading history file… webalizer.hist
Reading previous run data.. webalizer.current
333474 records (333474 ignored) in 37.50 seconds, 8892/sec

To check out just generated statistics open in browser:

http://yourserverhost/webalizer-www.website/

or

http://IP_Address/webalizer-www.website

 You should see statistics pop-up, below is screenshot with my currently generated stats:

Webalizer website access statistics screenshot Debian GNU Linux

Make daily Linux MySQL database backups with shell script

Thursday, May 23rd, 2013

Creating database backup with MySQL with mysqlbackupper and mysqlback shell scripts easy create mysql backups

Some time ago, I've written a tiny shell script which does dumps of Complete (SQL Script) MySQL databases. There are plenty of ways to backup MySQL database and plenty of scripts on the net but I like doing it my own way. I have few backup scripts. I prefer script database over keeping binary logs, or using some un-traditional backup methods like backing all binary data in /var/lib/mysql.

One was intended to backup with mysqldump whole database and later upload to a central server running tsh (shell). Using tsh maybe not the best method to upload, but the script can easily be modified to use ssh passwordless authentication as a method to upload.

I'm not a pro shell scripter, but MySQLBackupper script can be used as useful for learning some simple bash  shell scripting.

To use the script as intended you will have to build tsh from source. Tsh is in very early development stage (ver 0.2) but as far as I tested it before some years it does great what it is intended for. You can  MySQLBackupper.sh script from here.
Earlier, I used MysqlBackupper.sh to upload all SQL dumps to /backups directory on central backup storage server, thus I had written secondary script to classify uploaded backups based on backup archive name. Script used is mysqldumps-classify.sh and can be viewed here. Though this way of making backups, needs a bit of custom work for managing backups up to 10 / 20 servers it worked well.

I have written also another mysqlbackup script which is much more simplistic and only dumps with mysqldump and stores copies on hard disk in tar.gz archive. You can download my other simple mysqkbackup.sh here.

Only inconvenient thing about above scripts is they dump all SQL databases. Hence whether necessary to get content for single database from (complete) All database SQL (script backup), I use SED (stream editor) one liner script.

It is interesting to hear how others prepare their MySQL db backups.