SAA-C03 Amazon Web Services AWS Certified Solutions Architect - Associate (SAA-C03) Free Practice Exam Questions (2025 Updated)

To store and view engineering drawings with caching support, Amazon S3 and Amazon CloudFront are suitable solutions. Amazon S3 can store any amount of data with high durability, availability, and performance. Amazon CloudFront can distribute the engineering drawings to edge locations closer to the users, which can reduce the latency and improve the user experience. Amazon CloudFront can also cache the engineering drawings at the edge locations, which can minimize the amount of time that users wait for the drawings to load.

To resize images dynamically and serve appropriate formats to clients, a Lambda@Edge function with an external image management library can be used. Lambda@Edge allows running custom code at the edge locations of CloudFront, which can process the images on the fly and optimize them for different devices and browsers. An external image management library can provide various image manipulation and optimization features.

https://aws.amazon.com/rds/aurora/serverless/

Joining the FSx for Windows File Server file system to the on-premises Active Directory will allow the company to use the existing Active Directory groups to restrict access to the file shares, folders, and files after the move to AWS. This option allows the company to continue using their existing access controls and management structure, making the transition to AWS more seamless.

https://aws.amazon.com/blogs/aws/introducing-amazon-s3-object-lambda-use-your-code-to-process-data-as-it-is-being-retrieved-from-s3/

S3 Object Lambda is a new feature of Amazon S3 that enables customers to add their own code to process data retrieved from S3 before returning it to the application. By using S3 ObjectLambda, the data can be processed and transformed in real-time, without the need to store multiple copies of the data in separate S3 buckets or DynamoDB tables.

In this case, the Pll can be removed from the data by the code added to S3 Object Lambda before returning the data to the two applications that do not need to process Pll. The one application that requires Pll can be pointed to the original S3 bucket where the Pll is still stored.

Using S3 Object Lambda is the simplest and most cost-effective solution, as it eliminates the need to maintain multiple copies of the same data in different buckets or tables, which can result in additional storage costs and operational overhead.

https://docs.aws.amazon.com/vpc/latest/privatelink/vpc-endpoints.html

Amazon S3 is an object storage service that offers industry-leading scalability, data availability, security, and performance. You can use Amazon S3 to host static content for your website, such as HTML files, images, videos, etc. Amazon Elastic Container Service (Amazon ECS) is a fully managed container orchestration service that allows you to run and scale containerized applications on AWS. AWS Fargate is a serverless compute engine for containers that works with both Amazon ECS and Amazon EKS. Fargate makes it easy for you to focus on building your applications by removing the need to provision and manage servers. You can use Amazon ECS with AWS Fargate for compute power for your containerized application logic tier. Amazon RDS is a managed relational database service that makes it easy to set up, operate, and scale a relational database in the cloud. You can use a managed Amazon RDS cluster for the database tier of your application. This solution will simplify deployment and reduce operational costs for your three-tier application. References: https://docs.aws.amazon.com/AmazonS3/latest/userguide/WebsiteHosting.html https://docs.aws.amazon.com/AmazonECS/latest/developerguide/Welcome.htmlhttps://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/Welcome.html

https://docs.aws.amazon.com/prescriptive-guidance/latest/patterns/three-aws-glue-etl-job-types-for-converting-data-to-apache-parquet.html

AWS Glue is a serverless, fully managed ETL (Extract, Transform, and Load) service. It is specifically designed for data transformation tasks such as converting .csv files to Apache Parquet format. Using AWS Glue requires minimal development effort because it includes prebuilt transformations and integrates seamlessly with Amazon S3.

Option A:While Amazon EMR with Apache Spark offers extensive flexibility, it requires setting up and managing a cluster, writing custom Spark code, and handling resource scaling, which increases development effort compared to AWS Glue.

Option C:AWS Batch requires creating job definitions, specifying execution environments, and potentially writing custom scripts for the transformation process, which involves more setup compared to AWS Glue.

Option D:AWS Lambda could handle the transformation but is better suited for smaller-scale processing or real-time transformations. Handling hundreds of files daily with Lambda would require more complex orchestration and is not the most efficient solution for this scale of batch processing.

AWS Documentation References:

AWS Glue Overview

Transforming Data Using AWS Glue

A service control policy (SCP) is a type of policy that you can use to manage permissions in your organization. SCPs offer central control over the maximum available permissions for all accounts in your organization, allowing you to ensure your accounts stay within your organization’s access control guidelines. You can use an SCP to deny the ec2:CreateVolume action when the ec2:Encrypted condition equals false, which means that any user or role in the accounts under the root OU will not be able to create unencrypted EBS volumes. This solution will have minimal effect on employees who create EBS volumes, as they can still create encrypted volumes as needed.References:https://docs.aws.amazon.com/organizations/latest/userguide/orgs_manage_policies_scps.html

AWS Storage Gateway is a service that connects an on-premises software appliance with cloud-based storage to provide seamless and secure integration between an organization's on-premises IT environment and AWS's storage infrastructure. By deploying a file gateway as a virtual machine on each clinic's premises, the medical research lab can provide low-latency access to the data stored in the S3 bucket while maintaining read-only permissions for each clinic. Thissolution allows the clinics to access the data files directly from their on-premises file-based applications without the need for data transfer or migration.

{

"Version": "2012-10-17",

"Statement": [

{

"Action": [

"s3:ListBucket",

"s3:DeleteObject"

],

"Resource": [

"arn:aws:s3:::"

],

"Effect": "Allow",

},

{

"Action": "s3:*DeleteObject",

"Resource": [

"arn:aws:s3:::/*" # <- The policy clause kludge "added" to match the solution (Q248.1) example

],

"Effect": "Allow"

}

]

}

AWS DataSync is a service that makes it easy to move large amounts of data between AWS storage services and on-premises storage systems. AWS DataSync can copy files from an S3 bucket to an EFS file system and another S3 bucket continuously, as well as overwrite only the files that have changed in the source. This solution will meet the requirements with the least operational overhead, as it does not require any code development or manual intervention.

https://aws.amazon.com/blogs/big-data/manage-fine-grained-access-control-using-aws-lake-formation/

The AWS Health API provides programmatic access to the AWS Health information that is presented in the AWS Personal Health Dashboard. You can use the API operations to get information about AWS Health events that affect your AWS services and resources. You can also use the API to enable or disable health-based insights for your organization. You can use the AWS Health API at the start of each deployment to check on AWS infrastructure health andpause all new deployments if the API returns any issues.References:https://docs.aws.amazon.com/health/latest/APIReference/Welcome.html

An IAM policy is a document that defines the permissions for an IAM identity (such as a user, group, or role). You can use IAM policies to grant permissions to existing users and groups based on department. You can create an IAM policy that grants least privilege permission, which means that you only grant the minimum permissions required for the users to perform their tasks. You can then attach the policy to the IAM groups, which will apply the policy to all the users in those groups. This solution will reduce operational costs and simplify configuration and management of permissions.References:https://docs.aws.amazon.com/IAM/latest/UserGuide/access_policies.html

AWS DataSync is a data transfer service that makes it easy for you to move large amounts of data online between on-premises storage and AWS storage services over the internet or AWS Direct Connect. DataSync automatically encrypts your data in transit using TLS encryption, and verifies data integrity during transfer using checksums. DataSync can transfer data up to 10 times faster than open-source tools, and reduces operational overhead by simplifying and automating tasks such as scheduling, monitoring, and resuming transfers. References:https://aws.amazon.com/datasync/

API Gateway + Lambda is the perfect solution for modern applications with serverless architecture.

https://aws.amazon.com/storagegateway/file/

File Gateway provides a seamless way to connect to the cloud in order to store application data files and backup images as durable objects in Amazon S3 cloud storage. File Gateway offers SMB or NFS-based access to data in Amazon S3 with local caching. It can be used for on-premises applications, and for Amazon EC2-based applications that need file protocol access to S3 object storage.

https://aws.amazon.com/storagegateway/volume/

Volume Gateway presents cloud-backed iSCSI block storage volumes to your on-premises applications. Volume Gateway stores and manages on-premises data in Amazon S3 on your behalf and operates in either cache mode or stored mode. In the cached Volume Gateway mode, your primary data is stored in Amazon S3, while retaining your frequently accessed data locally in the cache for low latency access.

Amazon EFS is a fully managed file system that can be mounted by multiple Linux instances in AWS and on premises through native protocols such as NFS and SMB. Amazon EFS has no minimum size requirements and can scale up and down automatically as files are added andremoved. Amazon EFS also supports high availability and durability by allowing multiple mount targets in different Availability Zones within a region. Amazon EFS meets all the requirements of the question, while the other options do not. References:

https://aws.amazon.com/efs/

https://docs.aws.amazon.com/wellarchitected/latest/performance-efficiency-pillar/storage-architecture-selection.html

https://aws.amazon.com/blogs/storage/from-on-premises-to-aws-hybrid-cloud-architecture-for-network-file-shares/

A cross-account IAM role is a way to grant users from one AWS account access to resources in another AWS account. The cross-account IAM role can have a read-only access policy attached to it, which allows the users to download objects from the S3 buckets without modifying or deleting them. The cross-account IAM role also reduces the operational overhead of managing multiple IAM users and policies in each account. The cross-account IAM role meets all the requirements of the question, while the other options do not. References:

https://docs.aws.amazon.com/AmazonS3/latest/userguide/example-walkthroughs-managing-access-example2.html

https://aws.amazon.com/blogs/storage/setting-up-cross-account-amazon-s3-access-with-s3-access-points/

https://docs.aws.amazon.com/IAM/latest/UserGuide/id_roles_create_for-user_externalid.html

https://awscli.amazonaws.com/v2/documentation/api/latest/reference/ds/index.html

Amazon API Gateway is a fully managed service that makes it easy for developers to create, publish, maintain, monitor, and secure APIs at any scale. AWS Lambda is a serverless compute service that lets you run code without provisioning or managing servers. Lambda scales automatically based on the incoming requests, but it may take some time to initialize new instances of your function if there is a sudden increase in demand. This may result in high latency or cold starts for your API. To avoid this, you can use provisioned concurrency, which ensures that your function is initialized and ready to respond at any time. Provisioned concurrency also helps you achieve consistent low latency for your API by reducing the impact of scaling on performance. References:https://docs.aws.amazon.com/apigateway/latest/developerguide/http-api-develop-integrations-lambda.htmlhttps://docs.aws.amazon.com/lambda/latest/dg/configuration-concurrency.html

https://aws.amazon.com/blogs/security/how-to-monitor-and-visualize-failed-ssh-access-attempts-to-amazon-ec2-linux-instances/

https://aws.amazon.com/rds/aurora/global-database/

https://aws.amazon.com/blogs/architecture/using-amazon-aurora-global-database-for-low-latency-without-application-changes/

Before you begin: Decide which two Availability Zones you will use for your EC2 instances. Configure your virtual private cloud (VPC) with at least one public subnet in each of these Availability Zones. These public subnets are used to configure the load balancer. You can launch your EC2 instances in other subnets of these Availability Zones instead.

1. database shows no signs of being overloaded. CPU, memory, and disk access metrics are all low==>A and C out. We cannot only add nodes instance or add read replica, because database workload is totally fine, very low. 2. "least operational overhead"==>B out, because b need to configure lambda. 3. ROS proxy: Shares infrequently used connections; High availability with failover; Drives increased efficiency==>proxy can leverage failover to redirect traffic from timeout rds instance to healthy rds instance. So D is right.

Amazon Athena is the best choice for running one-time queries on streaming data. Although Amazon Kinesis Data Analytics provides an easy and familiar standard SQL language to analyze streaming data in real-time, it is designed for continuous queries rather than one-time queries[1]. On the other hand, Amazon Athena is a serverless interactive query service that allows querying data in Amazon S3 using SQL. It is optimized for ad-hoc querying and is ideal for running one-time queries on streaming data[2].AWS Lake Formation uses as a central place to have all your data for analytics purposes (E). Athena integrate perfect with S3 and can makes queries (A).

https://aws.amazon.com/cn/blogs/compute/cost-optimization-and-resilience-eks-with-spot-instances/

https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/aws-resource-ec2-placementgroup.html

"A cluster placement group is a logical grouping of instances within a single Availability Zone that benefit from low network latency, high network throughput"

https://computingforgeeks.com/stream-logs-in-aws-from-cloudwatch-to-elasticsearch/

Migrating to Amazon MQ reduces the overhead on the queue management. C and D are dismissed. Deciding between A and B means deciding to go for an AutoScaling group for EC2 or an RDS for Postgress (both multi- AZ). The RDS option has less operational impact, as provide as a service the tools and software required. Consider for instance, the effort to add an additional node like a read replica, to the DB. https://docs.aws.amazon.com/amazon-mq/latest/developer-guide/active-standby-broker-deployment.htmlhttps://aws.amazon.com/rds/postgresql/

Many applications, including those built on modern serverless architectures, can have a large number of open connections to the database server and may open and close database connections at a high rate, exhausting database memory and compute resources. Amazon RDS Proxy allows applications to pool and share connections established with the database, improving database efficiency and application scalability.https://aws.amazon.com/id/rds/proxy/

https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/private-content-restricting-access-to-s3.html

https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/distribution-web-awswaf.html

AWS Fargate is a serverless experience for user applications, allowing the user to concentrate on building applications instead of configuring and managing servers. Fargate also automates resource management, allowing users to easily scale their applications in response to demand.

This solution meets the requirements of moving the files automatically, running Lambda functions on the copied data, and sending the data files to SageMaker Pipelines with the least operational overhead. S3 replication can copy the files from the initial S3 bucket to the analysis S3 bucket as they arrive. The analysis S3 bucket can send event notifications to Amazon EventBridge (Amazon CloudWatch Events) when an object is created.EventBridge can trigger Lambda and SageMaker Pipelines as targets for the ObjectCreated rule. Lambda can run pattern-matching code on the copied data, and SageMaker Pipelines can execute a pipeline with the data files.

Option A is incorrect because creating a Lambda function to copy the files to the analysis S3 bucket is not necessary when S3 replication can do that automatically. It also adds operational overhead to manage the Lambda function. Option B is incorrect because creating a Lambda function to copy the files to the analysis S3 bucket is not necessary when S3 replication can do that automatically. It also adds operational overhead to manage the Lambda function. Option C is incorrect because using S3 event notification with multiple destinations can result in throttling or delivery failures if there are too many events.

In compliance mode, a protected object version can't be overwritten or deleted by any user, including the root user in your AWS account. When an object is locked in compliance mode, its retention mode can't be changed, and its retention period can't be shortened. Compliance mode helps ensure that an object version can't be overwritten or deleted for the duration of the retention period. In governance mode, users can't overwrite or delete an object version or alter its lock settings unless they have special permissions. With governance mode, you protect objects against being deleted by most users, but you can still grant some users permission to alter the retention settings or delete the object if necessary. In Governance mode, Objects can be deleted by some users with special permissions, this is against the requirement.

Compliance:

- Object versions can't be overwritten or deleted by any user, including the root user

- Objects retention modes can't be changed, and retention periods can't be shortened

Governance:

- Most users can't overwrite or delete an object version or alter its lock settings

- Some users have special permissions to change the retention or delete the object

Use a group email address for the management account's root userhttps://docs.aws.amazon.com/organizations/latest/userguide/orgs_best-practices_mgmt-acct.html#best-practices_mgmt-acct_email-address

In a typical development environment, dev and test databases are mostly utilized for 8 hours a day and sit idle when not in use. However, the databases are billed for the compute and storage costs during this idle time. To reduce the overall cost, Amazon RDS allows instances to be stopped temporarily. While the instance is stopped, you’re charged for storage and backups, but not for the DB instance hours. Please note that a stopped instance will automatically be started after 7 days. This post presents a solution using AWS Lambda and Amazon EventBridge that allows you to schedule a Lambda function to stop and start the idle databases with specific tags to save on compute costs. The second post presents a solution that accomplishes stop and start of the idle Amazon RDS databases using AWS Systems Manager.

https://aws.amazon.com/premiumsupport/knowledge-center/cloudfront-access-to-amazon-s3/

https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/private-content-restricting-access-to-s3.html#private-content-restricting-access-to-s3-overview

Service control policies (SCPs) are one type of policy that you can use to manage your organization. SCPs offer central control over the maximum available permissions for all accounts in your organization, allowing you to ensure your accounts stay within your organization's access control guidelines. Seehttps://docs.aws.amazon.com/organizations/latest/userguide/orgs_manage_policies_scp.html.

We recommend that you use On-Demand Instances for applications with short-term, irregular workloads that cannot be interrupted.https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ec2-on-demand-instances.html

Composite alarms determine their states by monitoring the states of other alarms. You can **use composite alarms to reduce alarm noise**. For example, you can create a composite alarm where the underlying metric alarms go into ALARM when they meet specific conditions. You then canset up your composite alarm to go into ALARM and send you notifications when the underlying metric alarms go into ALARM by configuring the underlying metric alarms never to take actions. Currently, composite alarms can take the following actions:https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/Create_Composite_Alarm.html

This solution meets the requirements of creating digital copies for a large collection of historical written records, analyzing the documents, extracting the medical information, and storing the documents so that an application can run SQL queries on the data. Writing the document information to an Amazon S3 bucket can provide scalable and durable storage for the scanned files. Using Amazon Athena to query the data can provide serverless and interactive SQL analysis on data stored in S3. Creating an AWS Lambda function that runs when new documents are uploaded can provide event-driven and serverless processing of the scanned files. Using Amazon Textract to convert the documents to raw text can provide accurate optical character recognition (OCR) and extraction of structured data such as tables and forms from documents using artificial intelligence (AI). Using Amazon Comprehend Medical to detect and extract relevant medical information from the text can provide natural language processing (NLP) service that uses machine learning that has been pre-trained to understand and extract health data from medical text.

Option A is incorrect because writing the document information to an Amazon EC2 instance that runs a MySQL database can increase the infrastructure overhead and complexity, and it may not be able to handle large volumes of data. Option C is incorrect because creating an Auto Scaling group of Amazon EC2 instances to run a custom application that processes the scanned files and extracts the medical information can increase the infrastructure overhead and complexity, and it may not be able to leverageexisting AI and NLP services such as Textract and Comprehend Medical. Option D is incorrect because using Amazon Rekognition to convert the documents to raw text can provide image and video analysis, but it does not support OCR or extraction of structured data from documents. Using Amazon Transcribe Medical to detect and extract relevant medical information from the text can provide speech-to-text transcription service for medical conversations, but it does not support text analysis or extraction of health data from medical text.

https://docs.amazonaws.cn/en_us/IAM/latest/UserGuide/reference_policies_examples_s3_rw-bucket.html

The policy is separated into two parts because the ListBucket action requires permissions on the bucket while the other actions require permissions on the objects in the bucket. You must use two different Amazon Resource Names (ARNs) to specify bucket-level and object-level permissions. The first Resource element specifies arn:aws:s3:::AdminTools for the ListBucket action so that applications can list all objects in the AdminTools bucket.

https://docs.aws.amazon.com/lambda/latest/dg/configuration-vpc.html#vpc-managing-eni

For standard accelerators, Global Accelerator uses the AWS global network to route traffic to the optimal regional endpoint based on health, client location, and policies that you configure, which increases the availability of your applications. Endpoints for standard accelerators can be Network Load Balancers, Application Load Balancers, Amazon EC2 instances, or Elastic IP addresses that are located in one AWS Region or multiple Regions.

https://docs.aws.amazon.com/global-accelerator/latest/dg/what-is-global-accelerator.html

https://docs.aws.amazon.com/whitepapers/latest/aws-vpc-connectivity-options/aws-direct-connect-aws-transit-gateway.html

This solution meets the requirements of directing users to a backup static error page if the primary website is unavailable, minimizing changes and infrastructure overhead. Route 53active-passive failover configuration can route traffic to a primary resource when it is healthy or to a secondary resource when the primary resource is unhealthy. Route 53 health checks can monitor the health of the ALB endpoint and trigger the failover when needed. The static error page can be hosted in an S3 bucket that is configured as a website, which is a simple and cost-effective way to serve static content.

Option A is incorrect because using a latency routing policy can route traffic based on the lowest network latency for users, but it does not provide failover functionality. Option C is incorrect because using an active-active configuration with the ALB and an EC2 instance can increase the infrastructure overhead and complexity, and it does not guarantee that the EC2 instance will always be healthy. Option D is incorrect because using a multivalue answer routing policy can return multiple values for a query, but it does not provide failover functionality.

Deploy the application servers by using Amazon EC2 instances in an Auto Scaling group across multiple Availability Zones. Use an Amazon RDS DB instance in a Multi-AZ configuration. To make an existing application highly available and resilient while avoiding any single points of failure and giving the application the ability to scale to meet user demand, the best solution would be to deploy the application servers using Amazon EC2 instances in an Auto Scaling group across multiple Availability Zones and use an Amazon RDS DB instance in a Multi-AZ configuration. By using an Amazon RDS DB instance in a Multi-AZ configuration, the database is automatically replicated across multiple Availability Zones, ensuring that the database is highly available and can withstand the failure of a single Availability Zone. This provides fault tolerance and avoids any single points of failure.

You can use CloudFront to deliver video on demand (VOD) or live streaming video using any HTTP origin. One way you can set up video workflows in the cloud is by using CloudFront together with AWS Media Services.https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/on-demand-streaming-video.html

CloudFront uses a local cache to provide the response, AWS Global accelerator proxies requests and connects to the application all the time for the response.

https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/private-content-restricting-access-to-s3.html#private-content-granting-permissions-to-oai

Cloudfront for rapid response and s3 to minimize infrastructure.

Migrate the database to Amazon Aurora MySQL - this will let the DB scale on it's own; it'll scale automatically without needing adjustment. Create AMI of the web app and using a launch template - this will make the creating of any future instances of the app seamless. They can then be added to the auto scaling group which will save them money as it will scale up and down based on demand. Using a spot fleet to launch instances- This solves the "MOST cost-effective" portion of the question as spot instances come at a huge discount at the cost of being terminated at any time Amazon deems fit. I think this is why there's a bit of disagreement on this. While it's the most cost effective, it would be a terrible choice if amazon were to terminate that spot instance during a busy period.

EC2 instance Savings Plan saves 72% while Compute Savings Plans saves 66%. But according to link, it says "Compute Savings Plans provide the most flexibility and help to reduce your costs by up to 66%. These plans automatically apply to EC2 instance usage regardless of instance family, size, AZ, region, OS or tenancy, and also apply to Fargate and Lambda usage." EC2 instance Savings Plans are not applied to Fargate or Lambda

This solution meets the requirements of a disaster recovery solution to back up the data that is generated by an analytics application, stored in JSON format, and must be accessible in milliseconds if it is needed. Amazon S3 Standard is a durable and scalable storage class for frequently accessed data. It can store any amount of data and provide high availability and performance. It can also support millisecond access time for data retrieval.

Option A is incorrect because Amazon OpenSearch Service (Amazon Elasticsearch Service) is a search and analytics service that can index and query data, but it is not a backup solution for data stored in JSON format. Option B is incorrect because Amazon S3 Glacier is a low-cost storage class for data archiving and long-term backup, but it does not support millisecond access time fordata retrieval. Option D is incorrect because Amazon RDS for PostgreSQL is a relational database service that can store and query structured data, but it is not a backup solution for data stored in JSON format.

Amazon CloudFront is a content delivery network (CDN) that speeds up the delivery of static and dynamic web content, such as HTML pages, images, and videos. By using CloudFront, the HTML pages will be served to users from the edge location that is closest to them, resulting in faster delivery and a better user experience. CloudFront can also handle the high traffic and large number of requests expected for the global event, ensuring that the HTML pages are available and accessible to users around the world.

https://aws.amazon.com/premiumsupport/knowledge-center/cloudwatch-memory-metrics-ec2/

The details are revealed in below url:https://docs.aws.amazon.com/AWSSimpleQueueService/latest/SQSDeveloperGuide/FIFO-queues.html

FIFO (First-In-First-Out) queues are designed to enhance messaging between applications when the order of operations and events is critical, or where duplicates can't be tolerated. Examples of situations where you might use FIFO queues include the following: To make sure that user-entered commands are run in the right order. To display the correct product price by sending price modifications in the right order. To prevent a student from enrolling in a course before registering for an account.

https://aws.amazon.com/blogs/compute/microservice-delivery-with-amazon-ecs-and-application-load-balancers/