Getting Started With Jekyll on Windows with Amazon S3
I have created this guide in the interest of helping other people figure out how to set up Jekyll on Windows (I use Linux all day at work, but my home machine is a Windows laptop). Some of the guides out there didn’t work for me since the packages they referenced were Mac-only, but hopefully this will work out for you.
A brief overview of Jekyll
Jekyll is a static site generator written in Ruby (requires no Ruby knowledge to use). Many people use it for generating their personal sites / blogs because the cost of hosting it on Amazon S3 is orders of magnitude cheaper than ordinary shared hosting, and orders of magnitude more reliable in serving the occasional high traffic spike. People also like it because there’s less code between you and your writing. Since the system is just a way to convert Markdown or HTML formatted posts into a templated HTML output, it requires very little configuration and setup to get a site up and running.
First, get started by downloading Ruby for Windows along with Ruby Development Kit. You will also need to put the rubygems/bin folder into your PATH (Right click “Computer” -> Properties -> Advanced System Settings -> Environment Variables -> Edit Path variable and append the path to the bin folder).
gem install jekyll
Configure and run jekyll
Download some boilerplate code for your new jekyll site by cloing this convenient jekyll boilerplate repository by necolas. Then, use it to create a git repository of your own. Help in doing this can be found at github.
Log into the s3 console and create two buckets, one called yourdomain.com and another called www.yourdomain.com. In the properties of the www.yourdomain.com* bucket, go to “Static Website Hosting” and select “Enable website hosting”. Then for the index document use “index.html” and for th error document use “error.html”. You can create error.html later as it’s not part of the boilerplate code. Then, configure the yourdomain.com bucket by editing its properties in the same way, except select “Redirect all requests to another host name”, and redirect all the requests to www.yourdomain.com.
Setup your DNS
Go to your DNS settings and for your zone apex record (Usually donated with an @, it represents the record that is used when a user types in “yourdomain.com”), have it set to URL Redirect to “www.yourdomain.com” and for the www record, have it point to the endpoint of your “www.yourdomain.com” record, which you can find in its properties under “Static Website Hosting”
Setup the s3 interface
Next, you’ll set up an easy way to push your new site to s3. Then, download an s3 interface for the Windows command line. (I use this instead of the jekyll-s3 package recommended on the main Jekyll repository because for some reason jekyll-s3 would not upload my images correctly to s3)
Lastly, cd to the directory your jekyll blog is at and use the following command in your command line:
s3 put www.bucketname.com _site/ /sub:withdelete /sync /acl:public-read
This command uploads your the _site folder to your bucketname.com bucket, deleting files on s3 that don’t exist anymore locally, only uploading files that have changed since last upload, and making all the files on s3 public.
For your convenience, you can put these commands as a batch file you can run when you want to update your site. Simply create a file in your blog folder called update_blog.bat and add the following text to it:
cd C:\Put\the\path\to\your\blog\here jekyll --no-auto --no-server git commit -am "updating blog" git push origin master s3 put www.yourdomain.com _site/ /sub:withdelete /sync /acl:public-read
All that’s needed to do now is to edit the templates and posts in a similar fashion to the boilerplate, and run the batch script.