CODEWITHSUNDEEP
videoamazon-s3streamingamazon-web-servicesflowplayer
pr0tocol
pr0tocol

Reputation: 323

Video upload and playback in Flowplayer using Amazon S3 and other webservices

I'm looking to do the following. I was hoping you expert could review the steps and let me know if it is possible, or at which step does my plan fall apart. I apologize if this format is not effective, but I think seeing my thought process in steps will be best. Thank you so much.

  1. Purchase an Amazon S3 storage account (start with trial)
  2. Create uploader on my website that will allow users to upload to the Amazon storage (web uploader)
  3. Create custom-coded database that will link video metadata on website (Title, User, Views, Date, etc) to the actual file address on Amazon storage. This step is important, as it will allow a user who uploaded a video to remove a video, change the video title on the website, and more.
  4. Custom-code webpage to locate file address based on video id (step 3-database)
  5. Custom-code webpage to play that file from Amazon storage in Flowplayer flash player

Hopefully these steps convey my goal.

My questions:

  1. Flowplayer accepts the following file formats: FLV, H.264, MP4. Out of convenience, I would want to take care of file conversion for my users, as I assume that is what YouTube and the big video companies do. How can I do this in step 2 (video upload)?
  2. Step 3, which is setting up the database to link Amazon S3 storage to the actual users/video metadata on my site seems somewhat complicated. Does Amazon have a service that takes care of all of this for me (easily?)

Thanks so much!

Upvotes: 1

Views: 1323

Answers (1)

Geoff Appleford
Geoff Appleford

Reputation: 18832

  1. You would upload to an Intermediate server where you would encode the videos in the desired formats before storing them on S3. EC2 would probably be a good fit here as transfers between EC2 and S3 are free and very fast.

    FFmpeg is a common tool used to convert videos or you might consider a 3rd party solution like encoding.com Most of the big players in this space support pulling and pushing to S3 natively too.

  2. Amazon doesn't really have any service to index and track your S3 videos. Whilst you can query S3 using the APIs or add metadata to S3 objects, you're definitely better off tracking everything in your own database. Use whatever database you are comfortable with - you only need to store the S3 bucket and key (filename) of each file along with any other user data you wish to link.

Upvotes: 2

Related Questions