Ingest media - Kinesis Video Streams

Ingest media

The following limits are in place:

  • Session duration: one hour, maximum

  • Signaling channels: maximum of 100 per account with storage configuration enabled

Ingest media from a browser

Important

Chrome is currently the only supported browser.

  1. Open the Amazon Kinesis Video Streams with WebRTC SDK in the JavaScript sample page.

  2. Complete the following information:

    • KVS Endpoint - In the Region field, select your region.

      For example, us-west-2.

    • AWS Credentials

      Complete the following fields:

      • Access Key ID

      • Secret Access Key

      • Session Token - The sample application supports both temporary and long-term credentials. Leave this field blank if you're using long-term IAM credentials. See Temporary security credentials in IAM for more information.

    • Signaling Channel - In the Channel Name field, type the name of the signaling channel you configured earlier. For more information, see Configure destination.

    • Tracks - Select Send Video and Send Audio.

    • WebRTC Ingestion and Storage - Expand the node and select Automatically determine ingestion mode. This option makes the sample application call the DescribeMediaStorageConfiguration API to determine in which mode to run.

  3. Select Start Master.

    If the signaling channel is configured for ingestion using the DescribeMediaStorageConfiguration API, the sample application will automatically invoke the JoinStorageSession API immediately after connecting to the Signaling channel to start the WebRTC ingestion workflow.

Ingest media from WebRTC C SDK

Follow the Amazon Kinesis Video Streams with WebRTC SDK in C for embedded devices procedures to build the sample application.

  1. Setup your environment with your AWS account credentials:

    export AWS_ACCESS_KEY_ID=YourAccessKey export AWS_SECRET_ACCESS_KEY=YourSecretKey export AWS_DEFAULT_REGION=YourAWSRegion

    If you're using temporary AWS credentials, also export your session token:

    export AWS_SESSION_TOKEN=YourSessionToken
  2. Run the samples:

    Master sample

    Navigate to the build folder and use "1" as the second argument. Type:

    ./samples/kvsWebrtcClientMaster channel-name 1

    GStreamer master sample

    Navigate to the build folder and use "audio-video-storage" as the second argument. Type:

    ./samples/kvsWebrtcClientMasterGstSample channel-name audio-video-storage testsrc

This starts WebRTC ingestion.

Note

Your supplied signaling channel must be configured for storage. Use the DescribeMediaStorageConfiguration API to confirm.

Add viewers to the ingestion session

Note

WebRTC ingest with multi-viewer support (Preview) is being provided in Preview as defined in the AWS Service Terms and is subject to change. It is currently only available in us-east-1 (IAD).

To participate in the preview, email .

Once the signaling channel is in WebRTC Ingestion mode, viewer participants no longer connect directly to the master participant. Viewer participants connect directly to the storage session. Viewer participants receive the media sent by the master participants, and viewer participants can send back optional audio to the master participant. Any audio sent back by viewers will be sent to all other peers connected to the storage session and ingested to the Kinesis Video Stream, as long as the master participant is connected to the storage session.

The following limits are in place:

  • Maximum number of viewers: 3

  • Maximum time viewer participants can be connected to the storage session without a master participant present: 3 minutes

Important

If a viewer disconnects from the storage session (closes the peer connection), their quota (viewer limit) remains consumed for 1 minute. During this 1-minute period, the viewer can invoke this API with the same Client ID to rejoin the session without consuming an additional viewer quota. After 1 minute, the viewer quota is released and available for other viewers to join.

Browser

Important

Chrome is the only supported browser.

  1. Open another tab in the Amazon Kinesis Video Streams with WebRTC SDK in the JavaScript sample page. All the information from the previous page will be automatically populated. If not, complete the following information:

    • KVS Endpoint - In the Region field, select your region.

      For example, us-west-2.

    • AWS Credentials

      Complete the following fields:

      • Access Key ID

      • Secret Access Key

      • Session Token - The sample application supports both temporary and long-term credentials. Leave this field blank if you're using long-term IAM credentials. See Temporary security credentials in IAM for more information.

    • Signaling Channel - In the Channel Name field, type the name of the signaling channel you configured earlier. For more information, see Configure destination.

    • Tracks - Select Send Audio. Note that if Send Video is checked, it will automatically be unchecked when choosing Start Viewer.

    • WebRTC Ingestion and Storage - Expand the node and select Automatically determine ingestion mode. This option makes the sample application call the DescribeMediaStorageConfiguration API to determine in which mode to run.

  2. Select Start Viewer.

    The application automatically calls the JoinStorageSessionAsViewer API immediately after connecting to the signaling channel to trigger an SDP offer send to the viewer from the session.

Note

With peer-to-peer WebRTC, the viewer participant is the controlling peer and the master participant is the controlled peer. In WebRTC ingestion mode, the storage session is now the controlling peer. After connecting to signaling and invoking JoinStorageSessionAsViewer, the viewer will need to respond to the SDP offer and establish a connection to the storage session through WebRTC.

Note

The storage session will only send TURN candidates. When nominating an ICE candidate pair from the perspective of participants, the remote candidate will always be of type relay.