Summary and Quiz

This lesson will summarize what we have learned about Simple Storage Service (S3). Also, we'll test our knowledge through a quiz.

Summary

Here is a summary of key takeaways from the Simple Storage Service (S3) section:

• S3 access management: We learned the key security mechanisms to secure an S3 bucket:

  • Resource-based policies: These are IAM policies that attach to a bucket and define which IAM user or role is allowed to access the bucket in the principal element.

  • ACLs: These are attached to the S3 buckets and their objects as a sub-resource to define the allowed users and allowed actions.

  • Access grants: These explicitly grant permissions to entities, such as users, groups, or services, to perform specific actions on S3 buckets and objects. Each Access Grant instance can give up to 100,000 grants.

  • Block Public Access Settings: These restrict public access to a bucket. Block public access provides four settings to manage access, which apply to buckets, users, and access points.

• S3 static website hosting: S3 allows to host static websites designed in HTML, CSS, and JavaScript. These websites are accessible over a unique bucket website URL after enabling static website hosting.

• Object versioning: S3 allows us to enable versioning on an S3 bucket to prevent accidental overwriting and deletion of objects. Through S3 Object versioning, we can restore the previous version of an object.

• S3 replication: To increase fault tolerance, S3 allows us to replicate objects and original metadata across buckets. S3 Replication allows real-time and batch replication across regions and accounts.

• Performance optimization: S3 encourages some practices to optimize performance, which include avoiding bottleneck prefixes, monitoring slow 500 slow-down errors, timeouts, and retries, and maximizing throughput through parallelism, S3 Transfer Acceleration, and Common Runtimes.

• PreSigned URLs: S3 offers PreSigned URLs to grant temporary access to users to download and upload objects. PreSigned URLs reflect the IAM permissions of the generator of the URL.

• S3 events: These allow us to invoke AWS services such as Lambda function, EventBridge, SQS, and SNS in response to events such as uploading or downloading objects to a bucket.

• S3 Batch Operations: These allow us to perform a similar operation on a batch of objects in a bucket. The S3 Batch Operations job creates a task for each object, allowing us to monitor the progress and generate the results after the completion of the job.

• S3 object storage classes: S3 offers various storage classes with variable costs and retrieval times. These storage classes can be categorized among three major access tiers.

  • Frequent access: The S3 standard class is the one with the lowest access latency and highest cost. Thus, it is used for the most frequently accessed objects.

  • Infrequent access: S3 offers S3 Standard Infrequent Access and S3 Standard One Zone Express IA for infrequently accessed objects.

  • Archive: S3 offers three storage classes for archival: S3 Glacier Flexible Retrieval. S3 Glacier Instant Retrieval and S3 Glacier Deep Archive. These storage classes vary in cost and the time required to restore the object.

  • S3 Intelligent-Tiering: This storage class automates the transition of objects between multiple storage classes based on the access pattern.

• S3 lifecycle policies: The life cycle policies allow us to automate transitioning and expiring objects after a specific time period.

• Access points: These enable us to get and put objects into a bucket through S3 access endpoints. Each access point has its own policy attached to it; thus, we can define multiple access configurations with access points.

• S3 Object Lambda: These invoke a Lambda function when an access point is accessed and allow us to add business logic to process the data before it is downloaded from a bucket.

• S3 Select and S3 Glacier Select: This mechanism allows us to fetch only a subset of data from a large object stored in the S3 Standard of S3 Glacier Instant Retrieval. This decreases time latency and costs required to fetch an object.