How Cloud Computing Supports DirecTV Become America’s #1 Satellite TV Service That Makes Moving Easy (*)

May 31, 2019

DirecTV

DirecTV, one of top 10 satellite TV providers ruling the American houses (source), is a direct broadcast satellite service provider based in El Segundo, California. Its satellite service, launched on June 17, 1994, transmits digital satellite television and audio to households in the United States, Latin America, Americas and the Caribbean.

DirecTV is one of top 10 satellite TV providers ruling the American houses

On July 24, 2015, AT&T acquired DirecTV in a transaction valued at $67.1 billion after receiving approval from the United States Federal Communications Commission and United States Department of Justice.

The main technologies DirecTV uses are High Definition Television (HDTV) provides customers videos in 10 times the resolution, depending on how much their TV can handle.

DirecTV is also considered as the undisputed leader in sports.

Challenge and Risk

With four regional data centers to provide over 20M customers throughout US and the unstoppable rising number of movie and TV show, data is growing so fast so the transcoding capability needs to be more scalability. No huge latency is accepted.

More pressure comes when customers expect the content must be served smoothly based on their destination location.

Currently, if users travel from California to New York, the system detects users are from California and the content will be served from California origin. And vice versa the scenario will happen with travelled users from New York.  

Current Architecture

DirecTV finds these challenges could not be solved by the current architecture which is not efficient as large media content is being sent across US. It could create latency, higher network bandwidth cost, and sometimes, interruption in service or long lag time.  

The architecture of DirecTV before the upgrade

The solution suggested by KaizenTP

KaizenTP has come with a direction: A system has to be optimized based on user location. It is included:

  1. Contents will be replicated to S3 bucket in multi-region.
  2. Using CloudFront as content delivery network, content will be served based on user’s geolocation.
  3. Route 53 also be used to check the health of resource and failover to other CDN in case of failure.
  4. Web application is scalable using scaling group.
  5. Using cross region replica for database to maximize the performance.

AWS Services in use

  1. AWS CloudFront as a content delivery network
  2. AWS EC2 for hosting web application
  3. AWS RDS as database for storing customer information
  4. AWS S3 for storing all media and static content, and access logs
  5. Route 53 as DNS solution and check the health of resources
  6. Elastic Transcoder for extending the transcoding capability
The architecture of DirecTV after the upgrade

Achievements

The upgraded architecture lets DirecTV’s customers view all their favorite programs in perfect digital format, along with accompanying sound effects to make the shows more realistic wherever they are. For detailed:

  • Content delivery: Faster
  • Network latency: Lower
  • Availability: Higher
  • User experience: Better
  • Transcode capability: Scalable
  • Database and Web Application: Scalable

Lesson Learned

If your company is facing the same challenge or you want your product/service to be more competent in the market, you may consider these lessons learnt from DirecTV cloud project:

  • Deep dive on Synchronous replication vs Asynchronous replication to determine which was the best solution based on the requirements
  • Enterprise architecture data flow and hosted infrastructureIt would have been much easier to understand the business and technologies requirements if we had involved EA team to support us on the enterprise data flow and hosted infrastructure
  • Build redundant systems from the start – our scope was for travelling customers, Ưbut it was apart of a complexed enterprise ecosystem.  We could have done a better job at asking these questions to EA and to make sure our solution adheres to enterprise standards and integrated to the overall enterprise architecture
  • Understand our network workloads; from high-volume transaction processing to video transcoding at scale.  Each of those workloads has unique requirements when it comes to the network. This would have helped us to set up a better initial VPC and scale from there.  

Want something simpler and quicker? Contact KaizenTP team for a free assessment today.

Categories: Uncategorized
Kaizen

© 2019 Kaizen Technology Partners. All Rights Reserved