Getting started with AWS Elemental MediaLive Anywhere

The media and entertainment industry has been requesting more flexibility in how they manage their on-premises infrastructure. This includes how they distribute content locally, and contribute content to Media Services on Amazon Web Services (AWS), while maintaining quality and reliability. With the new feature AWS Elemental MediaLive Anywhere (MediaLive Anywhere), media customers now have the option to deploy hybrid workflows using a cloud-based, centralized management control plane, with visibility into their entire on-premises encoding fleet. This blog shows how to set up and configure MediaLive Anywhere and provides best practices for operators.

Getting started with MediaLive Anywhere

MediaLive Anywhere is a feature of AWS Elemental MediaLive (MediaLive) that allows you to bring your own hardware infrastructure for live encoding. It is important to know if your current hardware will meet the required specifications. To purchase new, qualified hardware you should work with an AWS Partner. AWS can also provide information on converting existing AWS Elemental Live appliances into MediaLive Anywhere nodes.

Once the hardware is physically connected in an office or datacenter, install Red Hat Enterprise Linux Server (RHEL). (Note, there may be a cost associated with use of this software as customers must bring their own RHEL license) Next, you create a reference to that hardware in MediaLive and add the node to an encoding cluster. Finally, you run a script provided by the service to install the remaining software and activate the node, giving MediaLive permission to create channels on that node. You control the network routing, redundancy architectural design, and channel groupings to match your specific needs.

Steps to get up and running

Following are the high-level steps to get started with MediaLive Anywhere:

• Choose an AWS Partner to purchase new hardware or to qualify existing hardware
• Install the hardware in its physical location and connect it to the network
• Install the base operating system (OS) RHEL Server 9.4
• Use the AWS Management Console to configure Networks, Clusters, Nodes, and Channel placement groups in MediaLive Anywhere by referencing the service documentation
• Run the service-generated installation script on each hardware node
• Setup a sample MediaLive channel and start the job through the console
• Validate the output stream is working as expected

Key benefits

When creating new channels (or restarting existing channels), you gain access to the newest software immediately. This includes not only the newest features, but also new codecs (such as AV1). Your broadcast engineers can learn cloud skills and familiarize themselves with AWS services while maintaining their traditional role of managing and monitoring the hardware fleet—ensuring redundancy deployments are operational.

Whether the infrastructure relies on unicast, multicast, IP, SDI, or a mix of these protocols, operators can use the same AWS services to monitor their channels without worrying about the underlying formats being used. Additionally, developers now have one set of APIs to use for channels, allowing for code consolidation and improvement.

Lastly, there are no fees for push inputs or idle channels, as the IP management is handled by your network on premises. Be sure to review the pricing information to understand how the service is billed, as there is no limit to the amount of outputs per channel, which can help to reduce overall cost.

AWS Elemental Live encoders

Customers with existing AWS Elemental Live encoders from either the L800 or L900 series can use their hardware as MediaLive nodes. After setting up Integrated Dell Remote Access Controller (iDRAC) for lights-out-management, kickstarting the device with a fresh RHEL OS will be the next step. Customers should reference the on-boarding document to get started with the transition to MediaLive Anywhere from AWS Elemental Live.

Customers with Elemental Support have access to both AWS Elemental Live software and AWS Elemental MediaLive. This means that customers with AWS Elemental Live appliances can try out the MediaLive Anywhere feature. This can help with deciding which service best fits in their architecture—using the pay-as-you-go model.

Under the hood

AWS provides fully-managed services that may be built on top of other AWS services, usually those additional services are veiled from the end user for convenience and simplicity reasons. This relies on the shared responsibility model, which means AWS is responsible for securing the underlying infrastructure, while customers are responsible for securing the services and applications they deploy on AWS.

In the context of AWS services running on a customer’s infrastructure, like MediaLive Anywhere, the shared responsibility model means that AWS is responsible for the security and maintenance of the service itself. The customer is responsible for securing the underlying infrastructure and properly configuring and managing the service within their own environment.

Broadcast operators, who are tasked with ensuring a channel is up and stable, will see that MediaLive Anywhere works seamlessly with existing AWS services. It provides visibility into your channel’s uptime and statistics, providing direct access to the auxiliary services used.

It’s important to become familiar with the underlying services so you can make adjustments more quickly during a live event—especially with this deployment running as a hybrid model on your hardware.

MediaLive Anywhere is integrated with the following services:

• Amazon Elastic Container Service (Amazon ECS) Anywhere: Docker containers are deployed to correspond to each node you integrate in MediaLive. Monitoring the health of the container directly corresponds to the channels assigned to that node.
• Amazon CloudWatch: Reviewing the list of MediaLive metrics for both inputs and outputs can be useful in validating incoming and outgoing bandwidth, framerates, video quality, and frame drops. Multiple services communicate with CloudWatch, allowing dashboards to be created showing multiple near real-time monitors of a single channel.
• Amazon EventBridge: Events will be thrown by the AWS services and can be immediately acted on when you set rules in EventBridge. The MediaLive documentation shows a straightforward way to catch all events from that service.
• AWS Systems Manager (Systems Manager): Purpose built for fleet management, this service is useful for software and security patching. It provides direct visibility into CPU usage, disk I/O, and network throughput, which is critical when an operator is trying to locate a bottleneck.

Workflows that contribute from MediaLive Anywhere to AWS Elemental MediaConnect (for further distribution) or to an origin like AWS Elemental MediaPackage and a content delivery network (CDN) like Amazon CloudFront are fully supported.

Figure 1: This diagram shows a MediaLive Anywhere node running Red Hat Enterprise Linux communicating with other AWS services for monitoring and video delivery.

Building out redundancy

The feature MediaLive Anywhere addresses redundancy in multiple ways. The first step is to add node-level redundancy. MediaLive Anywhere can be configured in an N+M redundancy model, where specific nodes are flagged as Active nodes and other are flagged as Backup nodes. This allows MediaLive Anywhere to detect and remediate node failure. When detected, it can relocate entire Channel placement groups from a failed node to a stable node, while keeping the cluster grouping isolated from other clusters in your on-premises facility.

Figure 2: This diagram shows an example of node resiliency using an active/backup (N+M) redundancy model.

Ensure proper zonal or regional redundancy is completed through embedded timecode. Having the same timecode source allows MediaLive and MediaPackage to synchronize multiple inputs and provide seamless failover. This timecode can be provided by a camera, timecode generator, or by using epoch time.

MediaLive Anywhere can be used for both contribution or distribution workflows. When contributing content to the cloud for downstream processing, MediaConnect or MediaLive can accept contribution feeds from the on-premises encoders. Secure reliable transport (SRT) is one of the preferred contribution protocols.

When using two availability zones in a single region, MediaLive synchronizes the feeds when a Standard channel configuration is used. If two regions are required, two Single Pipeline channels may be used with two MediaPackage endpoints. In this scenario, the failover is handled using multiple CloudFront origins. In both cases, CMAF Ingest should be used as the output group from MediaLive to MediaPackage.

Figure 3: This architectural diagram shows an example of availability zone failover using two MediaLive Anywhere nodes and embedded timecode.

In the event that the adaptive-bitrate (ABR) encoding happens on the MediaLive Anywhere nodes, CMAF Ingest can be used to push to a local packager or up to MediaPackage for packaging and origination. This enables MediaPackage to synchronize the inputs and provide seamless failover within a region.

Figure 4: This architectural diagram shows an example of CMAF Ingest being sent from the on-premises encoders to MediaPackage in AWS.

Conclusion

Media & entertainment customers are constantly exploring and innovating with their hybrid workflows. The new feature MediaLive Anywhere helps bridge the gap between on-premises and cloud-based architectures, making it easier to deploy and manage your channels. Now that you are familiar with MediaLive Anywhere and the services that it uses, the next step is to get a node up and running. Reach out to one of our AWS Partners to get started today. As always, reach out to your account team for any assistance or read the documentation to learn more. Keep building.

Check out more AWS Partners or contact an AWS Representative to know how we can help accelerate your business.