
subaud is authored by Court Schuett (he/him). While Court is an 8090 Inc employee, opinions and statements on this blog are his own.


Court Schuett


In this demo, we will see how to build an application that will accept streaming audio sent via gRPC and have the audio transcribed using [OpenAI Whisper](https://openai.com/research/whisper). This demo builds off of the previous [gRPC Streaming Audio with Node](https://subaud.io/blog/node-grpc-server) blog post. However, in this demo, the ECS uses an [EC2 based deployment that supports GPUs](https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ecs-gpu.html). This requires several changes to the deployment that will be highlighted.

## Overview

![Overview](/images/blog/streaming-whisper-ecs-hosted/StreamingWhisper.png)

The basic concepts of the previous [gRPC Streaming Audio with Node](https://subaud.io/blog/node-grpc-server) remain largely the same. An audio stream will be generated from a local gRPC client and sent to a gRPC server hosted in [Amazon Elastic Container Service (ECS)](https://aws.amazon.com/ecs/). The server will transcribe the audio and return the results to the gRPC client. While the previous model used a Fargate hosted container, this demo uses an EC2 deployed container in order to use a GPU. Additionally, while the previous model used [Amazon Transcribe](https://aws.amazon.com/transcribe/) to provide the Speech To Text processing, this uses Whisper as the ASR and is included in the container that is built and deployed. Because Whisper does not normally support streaming audio, we will have to introduce a chunking mechanism that Whisper can use.

## GitHub Repo

https://github.com/schuettc/hosted-whisper-streaming

## ECS Deployment

```typescript
const cluster = new Cluster(this, 'cluster', {
  vpc: props.vpc,
});

const launchTemplate = new LaunchTemplate(this, 'LaunchTemplate', {
  machineImage: EcsOptimizedImage.amazonLinux2(AmiHardwareType.GPU),
  instanceType: new InstanceType('g4dn.2xlarge'),
  requireImdsv2: true,
  userData: UserData.forLinux(),
  securityGroup: ecsServiceSecurityGroup,
  role: ec2Role,
  blockDevices: [
    {
      deviceName: '/dev/xvda',
      volume: BlockDeviceVolume.ebs(300),
    },
  ],
});

const autoScalingGroup = new AutoScalingGroup(this, 'AutoScalingGroup', {
  vpc: props.vpc,
  launchTemplate: launchTemplate,
  desiredCapacity: 1,
  minCapacity: 1,
  maxCapacity: 2,
  vpcSubnets: { subnetType: SubnetType.PUBLIC },
});

const capacityProvider = new AsgCapacityProvider(this, 'AsgCapacityProvider', {
  autoScalingGroup,
});

cluster.addAsgCapacityProvider(capacityProvider);
```

Some changes are required in the CDK to use an EC2 based deployment of ECS. Here we see the [Launch Template](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ec2-launch-templates.html) that defines the EC2 instance parameters. An autoscaling group is also configured to allow this demo to scale up if needed.

```typescript
const taskDefinition = new Ec2TaskDefinition(this, 'taskDefinition', {
  taskRole: hostedWhisperStreamingRole,
  networkMode: NetworkMode.BRIDGE,
});

taskDefinition.addContainer('HostedWhisperStreaming', {
  image: ContainerImage.fromAsset('src/resources/whisperServer'),
  environment: {
    ECS_LOGLEVEL: props.logLevel,
  },
  gpuCount: 1,
  memoryLimitMiB: 31690,
  cpu: 8192,
  portMappings: [{ containerPort: 50051, hostPort: 50051 }],
  logging: new AwsLogDriver({ streamPrefix: 'HostedWhisperStreaming' }),
});

this.ecsService = new Ec2Service(this, 'ECSService', {
  cluster: cluster,
  taskDefinition: taskDefinition,
  capacityProviderStrategies: [
    { capacityProvider: capacityProvider.capacityProviderName, weight: 1 },
  ],
});
```

Next we configure the [Task Definition](https://docs.aws.amazon.com/AmazonECS/latest/developerguide/task_definitions.html) that will use the included Docker container image.

## Docker Image

```Dockerfile
FROM --platform=linux/amd64 nvidia/cuda:12.2.2-cudnn8-runtime-ubuntu22.04 AS base

RUN apt-get update && apt-get install -y --no-install-recommends \
    python3 python3.10-dev python3-pip python3.10-venv libsndfile1 build-essential curl git  && \
    rm -rf /var/lib/apt/lists/*

FROM base AS builder

COPY src/requirements.txt ./

RUN pip3 install --upgrade pip setuptools wheel && \
    pip3 install --user -r requirements.txt && \
    pip3 install --user git+https://github.com/openai/whisper.git

FROM base
ENV MODEL=${MODEL}
ENV LOG_LEVEL=${LOG_LEVEL}

COPY --from=builder /root/.local /root/.local
COPY src/* ./
RUN chmod +x /entrypoint.sh

EXPOSE 50051

ENTRYPOINT ["/entrypoint.sh"]


```

The Dockerfile included creates a Docker image that uses the [nvidia/cuda Docker image](https://hub.docker.com/r/nvidia/cuda/) as the base. This image will allow us to use the GPUs on the EC2 instance.

The `pip3 install` commands are particularly important when using this image. These will update and install the necessary tools to install and configure Whisper. In order to ensure the model is downloaded and installed, `python3 -c "import whisper; whisper.load_model('base')"` is run as part of the build.

As part of the `entrypoint.sh` script, we will be sure to set the correct path for the cuDNN libraries.

```bash
export LD_LIBRARY_PATH=`python3 -c 'import os; import nvidia.cublas.lib; import nvidia.cudnn.lib; print(os.path.dirname(nvidia.cublas.lib.__file__) + ":" + os.path.dirname(nvidia.cudnn.lib.__file__))'`
```

This solves the `Could not load library libcudnn_ops_infer.so.8. Error: libcudnn_ops_infer.so.8: cannot open shared object file: No such file or directory` issue that is caused by the `LD_LIBRARY_PATH` not containing the correct locations for the libraries.

Finally, we will start the server.

```bash
python3 server.py
```

## Whisper Server

The server starts with two Servicers - AudioTranscriberServicer and HealthCheckServicer. The HealthCheckServicer is used to validate to the Application Load Balancer

```python
def serve():
    server = grpc.server(futures.ThreadPoolExecutor(max_workers=10))
    audio_transcriber_pb2_grpc.add_AudioTranscriberServiceServicer_to_server(
        AudioTranscriberServicer(), server
    )
    audio_transcriber_pb2_grpc.add_HealthCheckServiceServicer_to_server(
        HealthCheckServicer(), server
    )
    server.add_insecure_port("[::]:50051")
    server.start()
    logger.info("Server started. Listening on port 50051.")
    server.wait_for_termination()
```

The AudioTranscriberServicer will be used to process the audio stream and return the transcriptions to the gRPC client. Whisper does not natively support streaming audio, so we must break the streamed audio into chunks that Whisper can use. To do this, audio processor will process the audio frames and use VAD to detect speech segments. These frames are buffered until a non-speech frame is encountered. When this happens, the frames are joined into a chunk. If the chunk is long enough, it is processed by [faster-whisper](https://github.com/SYSTRAN/faster-whisper) using the `large-v2` model. This allows for rapid transcription of streaming audio while allowing Whisper to provide the best results.

## Notes and Warnings

Because this deployment uses GPU based instance(s), be sure to check the [prices of the instances](https://aws.amazon.com/ec2/pricing/on-demand/) that will be used.

![Pricing](/images/blog/streaming-whisper-ecs-hosted/Pricing.png)

## Testing

This demo requires a [domain hosted in Route 53](https://docs.aws.amazon.com/Route53/latest/DeveloperGuide/registrar.html) so that a certificate can be associated with the [Application Load Balancer](https://docs.aws.amazon.com/elasticloadbalancing/latest/application/load-balancer-target-groups.html#target-group-protocol-version) listener. To configure the domain within the deployment, create a `.env` file with a `DOMAIN_NAME=` associated with a [Hosted Zone Name](https://docs.aws.amazon.com/Route53/latest/DeveloperGuide/hosted-zones-working-with.html) in the account.

Once the `.env` has been configured, you can deploy the CDK from the cloned repo:

```bash
yarn deploy
```

After deploying the CDK, you can test the functionality by running the included client. This client will capture audio from your microphone and stream it to the server using gRPC. Output from the CDK will be used to configure the client in the `/client/.env` file. To run the client:

```bash
cd client
yarn build
yarn start
```

![ClientStreaming](/images/blog/streaming-whisper-ecs-hosted/CLIExample.gif)

## Cleanup

In order to delete this CDK, you should remove the Auto Scaling Group first. Once that is done, you can delete the Stack from Cloudformation.


A CDK stack that will deploy a will create an ECS hosted application for streaming audio with gRPC to Whisper and getting the transcription output

Using Whisper with Streaming Audio


This simple bash script can be used to get the public and private IP addresses of the Tasks within a Fargate Cluster. With the `CLUSTER_NAME` defined as an environment variable, this script will iterate through all of the Tasks in a Cluster and create a `.env` file with both the public and private addresses of the Tasks.

```bash
#!/bin/bash

> taskIps.env

TASK_LIST=$(aws ecs list-tasks --cluster $CLUSTER_NAME)
TASK_ARNS=$(echo "$TASK_LIST" | jq -r '.taskArns[]')

while IFS= read -r TASK_ARN; do
    NETWORK_INTERFACE_ID=$(aws ecs describe-tasks --cluster $CLUSTER_NAME --tasks "$TASK_ARN" --query 'tasks[0].attachments[0].details[?name==`networkInterfaceId`].value' --output text)
    PRIVATE_IP=$(aws ecs describe-tasks --cluster $CLUSTER_NAME --tasks "$TASK_ARN" --query 'tasks[0].attachments[0].details[?name==`privateIPv4Address`].value' --output text)
    PUBLIC_IP=$(aws ec2 describe-network-interfaces --network-interface-ids "$NETWORK_INTERFACE_ID" --query 'NetworkInterfaces[*].Association.PublicIp' --output text)
    TASK_NAME=$(aws ecs describe-tasks --cluster $CLUSTER_NAME --tasks "$TASK_ARN" --query 'tasks[0].containers[0].name' --output text | tr '[:lower:]' '[:upper:]')
    echo "${TASK_NAME}_PUBLIC_IP=$PUBLIC_IP" >> taskIps.env
    echo "${TASK_NAME}_PRIVATE_IP=$PRIVATE_IP" >> taskIps.env
done <<< "$TASK_ARNS"


```

The basic steps involved:

- Get the Tasks in the Cluster
- Get the ARNs of the Tasks
- Loop through the Task ARNs
  - Get the Network Interface ID of the Task using the Task ARN and Cluster Name
  - Get the Private IP using the Task ARN and Cluster Name
  - Get the Public IP using the Network Interface ID
  - Get the Task Name from the Task ARN and Cluster Name
  - Write everything to the the `.env` file

### GitHub Gist

https://gist.github.com/schuettc/5e805ec2ca6c21f9b3af3f9cfa75fade


This simple bash script can be used to get the public and private IP addresses of a Fargate Cluster

Bash script to get ECS Fargate Cluster IP Addresses


# CDK Pipeline for ECS hosted NextJS App

In this demo we will explore and see how to use two different, but complementary techniques:

- Using CDK Pipelines to deploy a Stack
- Deploying a NextJS application to an ECS Fargate container

In each of these examples, we will look at several different mechanisms to provide security and scaling. This blog has been converted to using this method of hosting.

## GitHub Repo

https://github.com/schuettc/cdk-pipeline-ecs-hosted-nextjs

## CDK Pipeline

![CDKPipelineOverview](/images/blog/cdk-pipeline-ecs-hosted-nextjs/CDKPipelineOverview.png)

This demo uses [AWS CodeStar Connections](https://docs.aws.amazon.com/dtconsole/latest/userguide/connections-create-github.html) to create a link to a GitHub repo. This will allow us to use GitHub commits to push changes to our NextJS App through the CDK Pipeline.

The [cdk-pipeline-ecs-hosted-nextjs.ts](https://github.com/schuettc/cdk-pipeline-ecs-hosted-nextjs/blob/main/src/cdk-pipeline-ecs-hosted-nextjs.ts) file will create a Stack with a self-mutating CodePipeline. This is the Pipeline contains a Stage that will deploy the the NextJSApp Stack. This Pipeline is triggered by the GitHub repo and branch associated.

```typescript
const pipeline = new CodePipeline(this, 'Pipeline', {
  synth: new CodeBuildStep('Synth', {
    input: CodePipelineSource.connection(props.githubRepo, props.githubBranch, {
      connectionArn: props.connectionArn,
    }),
    commands: ['yarn install --frozen-lockfile', 'yarn build', 'npx cdk synth'],
    rolePolicyStatements: [
      new PolicyStatement({
        actions: ['sts:AssumeRole'],
        resources: ['*'],
        conditions: {
          StringEquals: {
            'iam:ResourceTag/aws-cdk:bootstrap-role': 'lookup',
          },
        },
      }),
    ],
  }),
  selfMutation: true,
  dockerEnabledForSynth: true,
});

pipeline.addStage(
  new NextJSAppStage(this, 'NextJSApp', props as NextJSAppProps),
);
```

## NextJS App

![NextJSAppOverview](/images/blog/cdk-pipeline-ecs-hosted-nextjs/ECSFargateNextJSApp.png)

The demo optionally uses a pre-configured Route 53 Hosted Zone domain. Otherwise the traffic will route through the following services:

- Route 53 (Optional)
- Cloudfront
- Application Load Balancer
- Fargate with Auto Scaling Group

### Securing Traffic

In order to ensure that traffic uses this path, we will establish the follow security mechanisms.

#### Fargate

```typescript
const fargateTargetGroup = new ApplicationTargetGroup(
  this,
  'nextJsAppTargetGroup',
  {
    vpc: props.vpc,
    port: 3000,
    protocol: ApplicationProtocol.HTTP,
    targets: [this.fargateService],
    healthCheck: {
      path: '/',
      protocol: Protocol.HTTP,
      port: '3000',
    },
  },
);
```

When configuring the `ApplicationTargetGroup` that points to the Fargate application, a Security Group will be created that allows traffic from the Application Load Balancer to access the Fargate task.

![FargateSecurityGroup](/images/blog/cdk-pipeline-ecs-hosted-nextjs/FargateSecurityGroup.png)

This Security Group will use the Security Group of the Application Load Balancer.

### Application Load Balancer

```typescript
this.applicationLoadBalancerSecurityGroup.addIngressRule(
  Peer.anyIpv4(),
  Port.tcp(80),
);
```

This Security Group will allow traffic in from any IPv4 address on TCP Port 80. In order to ensure that traffic from sources besides Cloudfront are rejected, we will use a configuration of the Listener.

```typescript
const fargateListener = this.applicationLoadBalancer.addListener(
  'fargateListener',
  {
    port: 80,
    protocol: ApplicationProtocol.HTTP,
    open: true,
    defaultAction: ListenerAction.fixedResponse(403),
  },
);

fargateListener.addAction('ForwardFromCloudFront', {
  action: ListenerAction.forward([fargateTargetGroup]),
  conditions: [
    ListenerCondition.httpHeader(props.customHeader, [props.randomString]),
  ],
  priority: 1,
});
```

This Listener will by default reject all traffic that is sent to it. An additional action will be used to allow traffic with a specific header and value to forward to the Fargate application. This information is created separately and configured during deployment. For more information on this mechanism, see [building-a-websocket-server-with-fargate-and-cdk](https://subaud.io/blog/building-a-websocket-server-with-fargate-and-cdk).

The result:

![Listener](/images/blog/cdk-pipeline-ecs-hosted-nextjs/Listener.png)

### Cloudfront

```typescript
this.distribution = new Distribution(this, 'CloudfrontDistribution', {
  minimumProtocolVersion: SecurityPolicyProtocol.TLS_V1_2_2021,
  defaultBehavior: {
    origin: new LoadBalancerV2Origin(props.applicationLoadBalancer, {
      httpPort: 80,
      protocolPolicy: OriginProtocolPolicy.HTTP_ONLY,
      originId: 'default-origin',
    }),
    viewerProtocolPolicy: ViewerProtocolPolicy.REDIRECT_TO_HTTPS,
    cachePolicy: CachePolicy.CACHING_DISABLED,
    allowedMethods: AllowedMethods.ALLOW_GET_HEAD_OPTIONS,
    originRequestPolicy: OriginRequestPolicy.ALL_VIEWER,
  },
  domainNames: props.domainName ? [props.domainName] : undefined,
  certificate: certificate,
  priceClass: PriceClass.PRICE_CLASS_100,
});
```

The Distribution created includes a domain and SSL certificate to provide HTTPS access to the Fargate task. Requests to HTTP will redirect to HTTPS.

#### Domain (optional)

```typescript
if (props.domainName) {
  hostedZone = HostedZone.fromLookup(this, 'HostedZone', {
    domainName: props.domainName,
  });

  certificate = new Certificate(this, 'Certificate', {
    domainName: props.domainName,
    validation: CertificateValidation.fromDns(hostedZone),
  });
}

if (hostedZone && props.domainName) {
  new AaaaRecord(this, 'mainAaaaRecord', {
    zone: hostedZone,
    target: RecordTarget.fromAlias(new CloudFrontTarget(this.distribution)),
  });

  new ARecord(this, 'mainARecord', {
    zone: hostedZone,
    target: RecordTarget.fromAlias(new CloudFrontTarget(this.distribution)),
  });
}
```

If a domain is included in the deployment, a certificate will be created and used in addition to the default Cloudfront certificate. DNS records will be created associated with this domain.

#### Custom Header

To create the header used to ensure traffic is sourced from Cloudfront, the Cloudfront Distribution Origin will be configured to send it using a Custom Resource.

```typescript
new CustomResource(this, 'customHeaderCustomResource', {
  serviceToken: customHeaderCustomResourceProvider.serviceToken,
  properties: {
    DistributionId: this.distribution.distributionId,
    Origins: [
      {
        OriginId: 'default-origin',
        CustomHeaders: [
          {
            HeaderName: props.customHeader,
            HeaderValue: props.randomString,
          },
        ],
      },
    ],
  },
});
```

The result will look something like this:

![Origin](/images/blog/cdk-pipeline-ecs-hosted-nextjs/Origin.png)

### Fargate Task

```typescript
const ecsTask = new FargateTaskDefinition(this, 'ecsTask', {
  memoryLimitMiB: 2048,
  cpu: 1024,
  runtimePlatform: {
    operatingSystemFamily: OperatingSystemFamily.LINUX,
    cpuArchitecture: CpuArchitecture.X86_64,
  },
});

ecsTask.addContainer('NextJsContainer', {
  image: ContainerImage.fromAsset('nextjs-docker'),
  containerName: 'nextJsApp',
  portMappings: [{ containerPort: 3000, hostPort: 3000 }],
  logging: LogDrivers.awsLogs({
    streamPrefix: 'nextJsApp',
  }),
  environment: {},
});

const taskSecurityGroup = new SecurityGroup(this, 'taskSecurityGroups', {
  vpc: props.vpc,
});

this.fargateService = new FargateService(this, 'nextJsAppFargateService', {
  cluster: cluster,
  taskDefinition: ecsTask,
  assignPublicIp: true,
  desiredCount: 1,
  vpcSubnets: { subnetType: SubnetType.PUBLIC },
  securityGroups: [taskSecurityGroup],
});
```

The Fargate Service will be configured with a port mapping for port 3000. This corresponds to the Application Load Balancer that forwards traffic to port 3000. The Fargate task will be deployed in the VPC created during the deployment and assigned a public IP address.

The Docker image used is the example shown here: https://github.com/vercel/next.js/tree/canary/examples/with-docker

### Result

The result is an Application Load Balancer that will only accept traffic with a specific header and value. This header and value are known only by the deployed Cloudfront Distribution. Traffic with this header will be forwarded to a Fargate task that only allows traffic from the Security Group associated with the Application Load Balancer. This deployment allows us to restrict access to the the Fargate task that uses the Cloudfront Distribution.

## Configuration and Deployment

To deploy this repo, you will need to complete the following steps:

1. Fork this repo
2. Configure Codestar connections with your GitHub repo
3. Configure a Hosted Zone in Route 53 with your domain (optional)
4. Configure your `.env` file
5. Deploy

### Configure Codestar

To configure your AWS account to support a connection to GitHub, follow the instructions here: https://docs.aws.amazon.com/dtconsole/latest/userguide/connections-create-github.html

The result will look something like this:

![Connections](/images/blog/cdk-pipeline-ecs-hosted-nextjs/Connections.png)

### Configure Hosted Zone (optional)

To create a Hosted Zone, follow the instructions here: https://docs.aws.amazon.com/Route53/latest/DeveloperGuide/CreatingHostedZone.html

The result will look something like this:

![HostedZone](/images/blog/cdk-pipeline-ecs-hosted-nextjs/HostedZone.png)

### Configure `.env` file

Before deploying, you will need to configure a `.env` file with the following fields:

```bash
CONNECTION_ARN=arn:aws:codestar-connections:us-east-1:112233445566:connection/XXXXXXXXXXXXXXXXXXXXXXXX
GITHUB_REPO=XXXXXXXXXXX/cdk-pipeline-ecs-hosted-nextjs
GITHUB_BRANCH=main
DOMAIN_NAME=example.com
```

`DOMAIN_NAME` is optional

### Deploy

To deploy the Pipeline:

```bash
yarn launch
```

Deploying the Pipeline will begin the process of deploying the NextJSApp stack.


A CDK Pipeline example that will deploy a NextJS App on ECS

CDK Pipeline for ECS hosted NextJS App


# Minecraft CDK Deployment using AWS Fargate

This [repo](https://github.com/schuettc/cdk-minecraft) will deploy an AWS Fargate based Minecraft server that uses Route 53 to update the service and activate the Minecraft server.

Much of the work here is based on the work by [doctorrat117](https://github.com/doctorray117) from [minecraft-ondemand](https://github.com/doctorray117/minecraft-ondemand).

## Overview

![Overview](/images/blog/Minecraft.png)

## .env File

The .env file used by this deployment includes several variables that will be referenced later. A sample .env file is included: [.env.sample](/.env.sample) and should be updated and renamed to `.env`.

```bash
MINECRAFT_EDITION = 'java'
SERVER_SUBDOMAIN = 'minecraft'
DOMAIN = 'example.com'
HOSTED_ZONE_ID = 'Z00000000000000000000'
MEMORY_SIZE = '8192'
CPU_SIZE = '4096'
SNS_EMAIL = 'admin@example.com'
STARTUP_MIN = '10'
SHUTDOWN_MIN = '20'
DEBUG = 'true'
```

A Route 53 domain and Hosted Zone ID is required for this deployment. These will be used to start the service.

### AWS Fargate Service

This deployment uses an AWS Fargate service to deploy two containers - the Minecraft server and a watchdog server. The Minecraft server is the reason for deploying this and will be used to play the game. The watchdog container is used to ensure that the Minecraft server is still being used.

#### Starting the Service

To start the service and allow someone to use the Minecraft server, a query to `SERVER_SUBDOMAIN`.`DOMAIN` will result in Route 53 updating Cloudwatch logs. This Cloudwatch log will trigger an AWS Lambda Function to set the `desired-count` of the Fargate service to 1.

```python
    if desired == 0:
        ecs.update_service(
            cluster=CLUSTER,
            service=SERVICE,
            desiredCount=1,
        )
        print("Updated desiredCount to 1")
    else:
        print("desiredCount already at 1")
```

By setting the `desired-count` to 1, the containers in the ECS service will be activated and available. The AWS ECS Service includes the Minecraft server container and a watchdog container. Both of these will be started when the `desired-count` is set to 1.

```typescript
this.service = new FargateService(this, "FargateService", {
  serviceName: "MineCraftService",
  cluster: this.cluster,
  capacityProviderStrategies: [
    {
      capacityProvider: "FARGATE",
      weight: 1,
      base: 1,
    },
  ],
  taskDefinition: this.task,
  assignPublicIp: true,
  desiredCount: 0,
  vpcSubnets: { subnetType: SubnetType.PUBLIC },
  securityGroups: [props.securityGroup],
  enableExecuteCommand: true,
});

const minecraftServerContainer = this.task.addContainer(
  "MinecraftServerContainer",
  {
    image: ContainerImage.fromRegistry(props.serverConfig.image),
    environment: {
      EULA: "TRUE",
      MEMORY: "8G",
    },
    portMappings: [
      {
        containerPort: props.serverConfig.port,
        hostPort: props.serverConfig.port,
        protocol: props.serverConfig.protocol,
      },
    ],
    essential: false,
    logging: props.serverConfig.debug
      ? new AwsLogDriver({
          logRetention: RetentionDays.THREE_DAYS,
          streamPrefix: "minecraft",
        })
      : undefined,
  }
);

this.task.addContainer("WatchdogContainer", {
  image: ContainerImage.fromAsset(
    path.resolve(__dirname, "./resources/watchdogContainer")
  ),
  essential: true,
  environment: {
    CLUSTER: this.cluster.clusterName,
    SERVICE: "MineCraftService",
    DNSZONE: props.subDomainHostedZoneId,
    SERVERNAME: `${props.serverSubDomain}.${props.domain}`,
    SNSTOPIC: props.snsTopic?.topicArn || "",
    STARTUPMIN: props.startupMin,
    SHUTDOWNMIN: props.shutdownMin,
  },
  logging: props.serverConfig.debug
    ? new AwsLogDriver({
        logRetention: RetentionDays.THREE_DAYS,
        streamPrefix: "minecraft",
      })
    : undefined,
});
```

#### Stopping the Service

During normal operations, both the Minecraft container and Watchdog container will be active. The Minecraft server is used to play the game while the Watchdog server will monitor the activity of the Minecraft server. When the watchdog server determines that the Minecraft server has not been used for `SHUTDOWN_MIN`, the watchdog server will set the `desired-count` of the Fargate Service to 0. By setting the `desired-count` to 0, the containers will be stopped to reduce costs.

### Notification

When the AWS Fargate service starts or stops, a notification is sent to SNS.

```bash
function send_notification ()
{
  [ "$1" = "startup" ] && MESSAGETEXT="${SERVICE} is online at ${SERVERNAME}"
  [ "$1" = "shutdown" ] && MESSAGETEXT="Shutting down ${SERVICE} at ${SERVERNAME}"

  ## SNS Option
  [ -n "$SNSTOPIC" ] && \
  echo "SNS topic set, sending $1 message" && \
  aws sns publish --topic-arn "$SNSTOPIC" --message "$MESSAGETEXT"
}
```


This demo will use CDK to deploy a Minecraft server on AWS Fargate using Route 53 as the trigger

Ecs

Using Whisper with Streaming Audio

Bash script to get ECS Fargate Cluster IP Addresses

CDK Pipeline for ECS hosted NextJS App

Deploy a Minecraft server with AWS Fargate and AWS CDK