You are currently viewing Migration of 350 GB Video Data to Amazon CloudFront with S3
vteams- blog

Migration of 350 GB Video Data to Amazon CloudFront with S3

  • Post published:February 19, 2016

A web portal of music tutorial software has thousands of customers, gazillions of orders and video data over 350 GB that offers monthly and annual subscriptions. A fast, secure and reliable solution was required by client of vteam #252 for downloading videos. Additionally a problem occurred with the video content being on portal’s own server. It caused the portal to hang completely while browsing. Security of the video content was the real concern as it was all paid by the users.

Challenge:

The main challenge was to download videos fast and make transition of data from portal’s server to cloud computing server smooth and efficiently once portal received download requests. Due to bandwidth consumption, requests should not be delayed as it was a limitation from client side.

Solution:

vteams engineers Raza Mehdi & Aaqib Javed did some research and suggested the following two available options:

  • Akamai cloud computing service
  • Amazon CloudFront with S3
    .

The first priority was to implement Akamai CDN (File & Media Downloads) for which a script was written which automatically synchronized the data between the client server & Akamai. The problem encountered with Akamai was that if a user cancelled the download in between or the download link itself stopped working, the client was still charged for the full download.

After Akamai services analysis, it was decided that Amazon CloudFront should be implemented with Amazon S3 providing the back-end storage for client’s video content. The main caveat was migrating the content in real-time, so the original script for migration was modified to synchronize the data with Amazon. After this, the client’s video content of over 350GB was successfully migrated to Amazon without any major downtime. The content was then delivered to the user through Amazon CloudFront with URL tokenization. This resulted in a fast, secure, reliable and cost-effective solution for the client.