S3 FAQ and Troubleshooting Guide

Products

• CloudHealth

Issue/Introduction

This guide is designed to address some common questions and provide steps to troubleshoot some of the more common S3 related issues. 

 

General S3 Information

Below are some general details associated to S3:

• There are 4 types of S3 services:  
  • Amazon S3 Standard is the default class.
  • Amazon S3 Standard Infrequent Access (IA) is designed for less frequently accessed data. Typical use cases are backup and disaster recovery solutions.
  • Amazon S3 One Zone-Infrequent Access is designed for data that is not often needed but when required, needs to be accessed rapidly. Data is stored in one zone and if that zone is destroyed, all data is lost.
  • Amazon Glacier is designed for long-term storage of data that is infrequently accessed and where retrieval latency of minutes or hours is acceptable.

• S3 bucket configuration is essential to account setup within CloudHealth.  

 

Notable CloudHealth S3 Features

CloudHealth provides visibility into cost and usage associated to the various types of S3 types available

• Reporting
  • Reporting on S3 provides cost information regarding your S3 storage.  S3 is billed by the amount of storage used, as well as the amount of data transferred out.The totals within this report reflect the cost of data transfer as well as storage.  You can select other options as well, such as API costs, costs by bucket, and/or Glacier storage costs.
  • Usage reporting shows what your S3 usage over time and which S3 storage types are used most frequently

• CloudTrail reporting
  • It's a common misconception that configuring S3 buckets will enable CloudTrail but a separate process is required with new bucket setup.  
• S3 buckets can be configured to receive subscriptions.  You may need to switch to an administrator role to ensure the appropriate permissions in order to manipulate what goes into the S3 bucket 

 

Resolution

Common Issues

Below are a number of common questions and details on how to resolve them:

 

How up to date is my S3 storage/data transfer information?

• Information on S3, Glacier, Volume, and Data transfer in/out is updated every 24 hours. 

How to fix an account warning that S3 Billing information cannot be found?

• If you have recently configured the bucket, CloudHealth needs some time to pinpoint the detailed billing information from the S3 bucket and upload it.  Check back on your account status in 24 hours. If no other problems persist, it should be updated to "Healthy" status.

Why is my data missing?

• If a new S3 bucket has been created, it will need to be fully configured and billing files need to be transferred over to the new bucket.

Where is the historical data for the bucket I just configured?

• CloudHealth does not pull historical data from CloudTrail buckets, we only report events starting after the CloudTrail configuration has been enabled.

When is bill stitching appropriate?

• In situations where files were swapped and/or new buckets were configured or removed mid-month, it may result in partial data being shown in CloudHealth.  The billing files can be stitched back together but best practice is to wait until the end of the month before commencing the process.  This will ensure that all appropriate data will be pulled together for the month without any risk of additional data splits/loss.

How do organizations play a role sending sub-org reports to S3 buckets?

• You can only send sub-org reports to S3 buckets that belong to accounts assigned to that sub-org.  The platform can only steer reports to buckets in accounts within the organization unit.  However you can assign one account to multiple organizations, so a workaround is that you can create another account with an S3 bucket in it and then assign that account to each organization.  This will allow you to send reports from every organization to a central bucket.  

What are "RDS - Charged Backup Usage" charges?

• These charges are associated to RDS snapshot costs for instances that have been shut down.  After an RDS instance is terminated, backup storage costs associated to those RDS instances will be billed at a per GB-month fee and fall under the "RDS - Charged Backup Usage" service cost.

Why do some of my AWS S3 charges show as indirect?

• S3 - API is regarded by AWS as a direct asset because the API action can be linked with S3 resources.  Certain API actions cannot be directly linked to a particular asset and thus are categorized as indirect and placed under that categorization.