Your browser is out-of-date!

Update your browser to view this website correctly. Update my browser now

×

Joyn migrates three petabytes of content to the cloud with AWS

Content from the archive that sees a lot of interest is sorted into a frequent access tier, while content that draws less attention is stored in an infrequent access tier

German streaming service Joyn has moved three petabytes of content from an on-premises facility into Amazon Simple Storage Service (Amazon S3) in under three months.

The project used 40 AWS Snowball appliances, with the files ultimately stored in Amazon S3 Glacier. The 3.4 PB of new media complements an existing catalogue of hundreds of thousands of video titles already stored on Amazon S3.

“Historically, media asset storage and archives have been relegated to on-premises, but in the last few years, as the cloud has proven its value as a secure, cost-effective alternative, more content providers are moving to the cloud as hardware phases out. It was the right move for us, also considering that Joyn is a cloud-first service and was built on AWS services,” said Stefan Haufe, media engineer, Joyn.

“We knew that moving our archive using AWS Snowball would be fast, simple and secure, but we’ve also been impressed by how cost-effective it was. Now that all of our content is easily accessible in the cloud, we can quickly transcode and deliver content to audiences that they can’t find on any other streaming service.”

Joyn engineers collaborated with AWS to optimise the storage architecture, moving a significant portion of the streamer’s library to Amazon S3 Intelligent-Tiering. According to AWS, by utilising Amazon S3 Intelligent-Tiering, Joyn can keep all of its content online and also optimise storage automatically as access patterns change. Content from the archive that sees a lot of interest is sorted into a frequent access tier, while content that draws less attention is stored in an infrequent access tier.

“It used to be that we’d have to be selective about which content we’d retrieve from our deep archive, or in some cases, what we’d keep on the archive, but now it’s a no brainer,” added Haufe. “We were able to grow our storage volume by a factor of 3x for the same total cost of ownership by using S3 Intelligent-Tiering.”