Implement a backup technique for Amazon Fast Sight BI property

Amazon Fast Sight is a core characteristic inside Amazon Fast — an agentic, AI-powered digital workspace designed to maximise end-user productiveness— that gives AI-powered BI capabilities by pure language queries, interactive dashboards, and embedded analytics from trusted enterprise knowledge sources.

Amazon Fast Sight property comparable to dashboards, analyses, datasets, and knowledge sources will be backed up utilizing the AssetsAsBundle APIs described on this put up. A backup technique helps defend in opposition to unintended deletions, unintended modifications, and regional disruptions. For groups that depend on Fast Sight to help essential enterprise selections, a well-designed backup plan is really helpful.

This put up is the primary in a two-part collection overlaying backup and restore for Amazon Fast Sight BI property:

• Half 1 (this put up): Covers methods to design and implement a backup technique, together with asset choice, the APIs accessible for export, and a ready-to-use pattern automation software.
• Half 2: Covers the restore course of. You should utilize the backups created in Half 1 to recuperate property after unintended deletion, unintended modifications, or as a part of a broader catastrophe restoration plan.

An efficient backup technique is particularly essential for organizations in closely regulated industries comparable to monetary companies, healthcare, and vitality, for a number of causes:

• Information loss prevention protects in opposition to human errors, unintended deletions, and occasions like ransomware.
• Assembly restoration aims helps organizations obtain their Restoration Level Goals (RPO) and Restoration Time Goals (RTO), minimizing knowledge loss throughout incidents.
• Audit and reporting helps monitoring and reporting on property all through their lifecycle (creation, updates, and deletion).
• Elevated workload resiliency permits fast restoration of programs to earlier states, decreasing downtime and enhancing reliability. This aligns with the Reliability pillar of the AWS Nicely-Architected Framework.
• Catastrophe restoration (DR) preparedness gives a basis for implementing a DR course of that anticipates technology-related disasters and contributes to your group’s enterprise continuity plan (BCP).

For extra details about the catastrophe restoration capabilities of Fast, and methods to assess them in opposition to organizational necessities, see the Amazon Fast catastrophe restoration and resiliency information.

On this put up, we cowl finest practices for implementing an efficient backup technique for BI property in Fast Sight. We begin by overlaying the choices for choosing the property to incorporate in your backup, then clarify the high-level APIs accessible for that function, and finalize with pattern code that can assist you get began shortly.

Backup practices for enterprise intelligence

BI programs current distinctive enterprise continuity challenges due to their position in supporting decision-making processes and key stakeholders. You could defend them in opposition to service disruptions by implementing an efficient backup plan. Earlier than constructing this plan, it’s necessary to know the structure and the scale to contemplate as a part of your DR plan.

The previous diagram exhibits that Fast Sight depends on AWS’s world infrastructure throughout a number of AWS Areas to present excessive availability for Fast Sight property, together with knowledge sources, datasets, analyses, and dashboards.

The Tremendous-fast, Parallel, In-memory Calculation Engine (SPICE) shops and encrypts imported knowledge with excessive availability (HA) by redundant copies throughout a number of Availability Zones (AZs) throughout the Fast Sight Area.

With this regional design, you may keep assets in a number of Areas and use a secondary Area within the unlikely occasion of a regional outage affecting your main BI assets.

For person and id administration, Fast Sight makes use of a single Area that you simply outline through the preliminary account subscription course of. The diagram exhibits that this Area hosts person and group id info and have to be accessible for customers to entry Fast Sight.

For instance, if a person desires to entry a dashboard within the eu-west-1 Area however the Fast Sight most important Area is us-east-1, each Areas have to be accessible to complete the person entry circulation. Fast Sight makes use of regional structure with AZs for redundancy. Nonetheless, if your online business wants safety in opposition to the unlikely occasion of a regional outage, you need to design your catastrophe restoration (DR) technique accordingly.

Tip: In case you’re not sure of your Fast Sight most important Area, you may retrieve this info by working the next command:

aws quicksight describe-account-settings --aws-account-id XXXXXXXXXXXX --region us-east-1

Be aware: This aws quicksight describe-account-settings command specifies us-east-1 because the endpoint Area. In case you obtain a 200 standing, your id Area is us-east-1. In any other case, you obtain an error like the next, which instructs you to level to your present id Area (for instance, eu-west-1):

An error occurred (AccessDeniedException) when calling the DescribeAccountSettings operation: Operation is being referred to as from endpoint us-east-1, however your id area is us-east-1. Please use the eu-west-1 endpoint.

Defining Fast Sight property to incorporate within the backup plan

With a clearer understanding of Fast Sight structure, the following step is deciding on the property to incorporate in your backup plan, for this you may comply with two methods:

Again up particular property:

This feature is appropriate once you outline a backup or DR technique targeted on defending essential property for your online business operations which you can conveniently restore after a catastrophe or unintended deletion. This consists of particular dashboards (and their dependent property) that key stakeholders use to make enterprise selections or that working groups (finance, logistics, procurement, and so forth) use to help continued enterprise operation.

This feature is really helpful once you require a simple backup plan and when the BI property which are key to enterprise continuity are a subset of all of the property accessible in your Fast Sight occasion.

Again up all property:

This technique is really helpful once you wish to outline a backup technique that covers each versioning and potential catastrophe restoration. By backing up all property, you may carry out in-place rollback of any asset to a earlier state if a human error causes an unintended modification or deletion. Moreover, as a result of you’ve got a backup of all property in your account, you may choose particular property to revive as a part of your DR plan.

This method provides you most protection but additionally requires extra advanced orchestration and automation. This put up focuses on this technique and gives pattern code which you can adapt to reduce time to manufacturing.

After you choose your technique, select the kind of BI property to export. Fast Sight affords the next asset sorts:

• Dashboards: Learn-only property focused at reader customers, printed from an evaluation. You may as well save a dashboard to an evaluation to make edits.
• Analyses and dashboards: An evaluation is an editable model of a dashboard. Solely the authors you select can entry it.
• Information sources: An information supply implements the connection to your knowledge, which might come from analytic sources comparable to databases or knowledge warehouses, AWS companies comparable to Amazon Easy Storage Service (Amazon S3), or third-party software program as a service (SaaS) knowledge suppliers comparable to Jira and ServiceNow.
• Datasets: An asset kind that makes use of a knowledge supply to entry exterior knowledge that you need to use to organize and construction the info that powers your analyses and dashboards.
• VPC connections: A characteristic that you need to use to combine together with your VPC assets comparable to databases and knowledge warehouses which are situated in that VPC or reachable from it (peered VPCs or networks related by VPN or AWS Direct Join).
• Themes: A group of styling and look settings which you can apply to a number of analyses and dashboards to match an aesthetic commonplace that meets your product or company branding wants.

All these property have dependencies between one another, with analyses and dashboards on the prime of this dependency chain, as the next diagram illustrates.

While you select the asset sorts to again up, pay attention to these dependencies so you may totally restore property from the backup. For instance, once you again up a dashboard, you additionally must again up its dependencies, which could embrace datasets, knowledge sources, VPC connections, and a theme. The following sections clarify how Fast Sight export APIs deal with these dependencies.

Backup course of overview

The mechanism we cowl on this put up makes use of the AssetsAsBundle APIs accessible in Fast Sight. AssetsAsBundle APIs (additionally referenced as AAB APIs) are a set of high-level APIs designed to help programmatic export and import of Fast Sight assets. They cowl a spread of use instances comparable to launch administration, backup and restore, cross-account migration, and steady integration and steady supply (CI/CD) workflows.

This set of APIs consists of the next operations:

• StartAssetBundleExportJob: Creates a package deal (bundle) that comprises the property exported as a part of the operation. The package deal is a zipper file with textual content information. The format will be both JSON or AWS CloudFormation relying on the worth specified within the ExportFormat parameter. Relying on the format, you may import these property utilizing the AAB APIs straight or use CloudFormation infrastructure as code (IaC) for provisioning. After the asynchronous operation finishes, the system uploads the bundle to a brief S3 location for downloading.
• StartAssetBundleImportJob: Takes a beforehand exported bundle and restores the property packed in it. You should utilize the import operation to outline overrides for a large set of parameters comparable to asset names and knowledge supply connection parameters (host, port, workgroup, and extra).
• DescribeAssetBundleImportJob and DescribeAssetBundleExportJob: Each AssetBundle operations are asynchronous. You should utilize these APIs to explain the operation, ballot for its standing, and act after it finishes. While you carry out an export job, you need to use DescribeAssetBundleExportJob to retrieve the DownloadUrl for the bundle, which is legitimate for five minutes. You may renew the URL with additional calls to DescribeAssetBundleExportJob.

Supported property and present limitations of AssetsAsBundle APIs

AssetsAsBundle APIs help an inventory of Fast Sight property together with analyses, dashboards, knowledge sources, datasets, shared folders, restricted folders, refresh schedules, themes, and VPC connections. Nonetheless, some asset sorts have limitations.

Unsupported knowledge sources: Adobe Analytics, File, GitHub, Jira, Salesforce, ServiceNow, Amazon S3 (with domestically uploaded manifest information), and Twitter.

Unsupported datasets: Datasets that comprise machine studying (ML) columns generated utilizing inference by related SageMaker ML fashions.

You could exclude these property out of your backup plan to keep away from an InvalidParameterValueException error once you situation the StartAssetBundleExportJob operation.

To work round this, you may change unsupported knowledge sources and datasets by following these procedures.

For Amazon S3 knowledge sources with native manifest information:

1. Create a brand new Amazon S3 knowledge supply.
2. Add the manifest file to Amazon S3.
3. Reference the manifest file out of your knowledge supply.
4. Change the info supply within the dependent datasets utilizing the UpdateDataSet API.

For different unsupported knowledge sources and datasets:

Observe this process to rework your incompatible dataset right into a suitable one:

1. Create an evaluation related to the info supply you wish to help in your backup.
2. Create a desk visible that shows all dataset columns.
3. Export the info as a CSV file.
4. Create an Amazon S3 dataset utilizing a manifest uploaded to Amazon S3.
5. Replace your analyses and dashboards with the brand new dataset utilizing the change dataset performance.

Different property to contemplate as a part of your backup

Though Fast Sight assets are the important thing property to again up, you want to embrace some further assets and configurations in your backup plan for potential restore or catastrophe restoration conditions.

You may export Fast Sight property together with their permissions, together with the customers and teams which have entry to them. You management this by setting the IncludePermissions flag to true.

As a result of every Fast Sight asset is owned by a person, you want to again up customers and teams to have a full and restorable backup.

AssetsAsBundle APIs don’t cowl customers and teams, however you need to use DescribeUser, DescribeGroup, and DescribeGroupMembership to incorporate this info within the backup.

Along with customers and teams, contemplate backing up account settings comparable to account customization (the DescribeAccountCustomization API), personalized manufacturers (the DescribeBrand API), and folders (the ListFolders, DescribeFolder, and DescribeFolderPermissions APIs).

Technical implementation

On this part, we cowl methods to create an automation that orchestrates the invocation of the Fast Sight APIs wanted to carry out an efficient backup implementation. We offer pattern code on the finish of this part that implements each customers and teams backup and Fast Sight property backup.

Backup orchestration circulation

The automation software helps three modes of operation: person backup solely, property backup solely, and each. This gives most flexibility once you carry out your backup plan. The next diagram exhibits the circulation that the software follows relying on the chosen operation mode.

Customers and group backup

The person and teams backup service makes use of the Fast Sight person and group APIs to learn your account’s present state and retailer the retrieved person and group knowledge in Amazon DynamoDB. The service makes use of date-based suffixes for DynamoDB desk names to protect historic backup knowledge and stop overwrites. This permits point-in-time restoration and backup historical past monitoring. This design additionally simplifies restore operations since you don’t must filter by date suffixes once you question knowledge inside a particular backup.

Instance for a backup run on 2025-10-19:

• Customers: quicksight-users-backup-2025-10-19
• Teams: quicksight-groups-backup-2025-10-19
• Consumer-Group Memberships: quicksight-users-groups-backup-2025-10-19

Customers Desk Schema:

{
  "user_name": "string (partition key)",
  "arn": "string",
  "electronic mail": "string",
  "position": "string",
  "identity_type": "string",
  "lively": "boolean",
  "principal_id": "string",
  "backup_timestamp": "string (ISO 8601)",
  "custom_permissions_name": "string"
}

Teams Desk Schema:

{
  "group_name": "string (partition key)",
  "arn": "string",
  "description": "string",
  "principal_id": "string",
  "members": ["list of user names"],
  "backup_timestamp": "string (ISO 8601)"
}

Customers-Teams Membership Desk Schema:

{
  "membership_id": "string (partition key, format: username#groupname)",
  "user_name": "string",
  "group_name": "string",
  "user_arn": "string",
  "group_arn": "string",
  "backup_timestamp": "string (ISO 8601)"
}

Be aware: The person and group backup service implements dual-Area help. Consumer and group operations use the identity_region configuration parameter, whereas backup asset operations use the usual aws_region. This design addresses enterprise eventualities the place Fast Sight id administration is configured in a special Area than asset storage.

Belongings backup

The property bundle backup service coordinates the export of property inside a Area and uploads the generated bundle to an Amazon S3 location for later use. The automation backs up the next property: knowledge sources, datasets, analyses, dashboards, and themes. By default, the backup consists of all dependencies. You may disable this setting if wanted.

At a excessive stage, the service performs the next duties:

• Lists all knowledge sources utilizing the ListDataSources API, filtering out Amazon S3 manifest-based knowledge sources and knowledge sources with invalid VPC connection names. Names should comprise solely alphanumeric characters separated by hyphens.
• Lists all datasets utilizing the ListDataSets API, filtering out FILE datasets by checking the ImportMode area.
• Lists all analyses utilizing the ListAnalyses API.
• Lists all dashboards utilizing the ListDashboards API.
• Teams property by kind for separate export jobs. You may configure the variety of property to incorporate in every bundle, with a most of 100 (the API restrict).
• Checks the export job standing utilizing the DescribeAssetBundleExportJob API and implements exponential backoff to keep away from throttling.
• Uploads the finished asset bundle to Amazon S3 utilizing the next prefix construction.

my-QuickSight-backups/
└── QuickSight-backups/                          # Customized S3 prefix
    ├── 2024/01/15/
    │   ├── datasources/
    │   │   ├── datasources-143022.zip            # Single bundle (≤ max_assets_per_bundle)
    │   │   └── datasources_bundle_1-143045.zip   # A number of bundles when property exceed restrict
    │   ├── datasets/
    │   │   ├── datasets_bundle_1-143045.zip      # A number of bundles when property exceed restrict
    │   │   └── datasets_bundle_2-143045.zip      # Sequential numbering for a number of bundles
    │   ├── analyses/
    │   │   └── analyses-143108.zip               # Single bundle
    │   └── dashboards/
    │       ├── dashboards_bundle_1-143131.zip    # First of a number of dashboard bundles
    │       └── dashboards_bundle_2-143131.zip    # Second dashboard bundle
    └── 2024/01/16/
        ├── datasources/
        │   └── datasources-090015.zip
        ├── datasets/
        │   └── datasets-090030.zip
        └── ...

Be aware: The bundle quantity string is current solely when the variety of property to again up exceeds the configured worth in max_assets_per_bundle.

Finish-to-end software for backup creation

The QuickSight-backup software gives a easy option to export all of your Fast Sight property and their dependencies into sturdy, cheap storage comparable to Amazon S3. The software creates new prefixes for generated bundles, so earlier backups aren’t overwritten. The software additionally exports customers and teams utilizing the identical precept: DynamoDB shops this knowledge, and desk names comprise the date when the backup was generated. With this method, you need to use backups as a supply in your restoration technique and monitor the historical past of modifications to your Fast Sight property and related customers.

The code makes use of the Boto3 Python SDK and consists of packaging help by setuptools for setup and use.

Tooling utilization and configuration

Earlier than utilizing the software, be sure you meet the next conditions:

• Python 3.8 or greater.
• A Fast Sight account with Enterprise version or greater.
• AWS Command Line Interface (AWS CLI) configured with applicable credentials.
• Required AWS permissions. See the Permissions part within the code.

Clone from supply

git clone https://github.com/aws-samples/sample-quicksight-backup-tool.git
cd quicksight-backup-tool

Create a Python venv (really helpful)

python3 -m venv ./.venv
supply .venv/bin/activate

Set up the package deal

Create a configuration file

To get began, check with the config-basic.yaml file within the repo or create one from scratch. This configuration file defines key parameters for the software, together with the next:

• AWS account.
• Area.
• Backup areas (DynamoDB tables and Amazon S3 bucket prefixes).

Utilizing the software

After set up, you may run the software as follows:

quicksight-backup --config config.yaml --mode full

You solely want to offer the --config parameter. You may omit the remainder. The --mode parameter controls the backup kind (full, users-only, or assets-only), the place full is the default mode. The next checklist describes the arguments the software helps.

Optionally available arguments

• --mode, -m: Backup mode (full, users-only, assets-only); default is full.
• --output-dir, -o: Output listing for studies and manifests.
• --verbose, -v: Allow verbose (DEBUG) logging.
• --log-file: Path to log file.
• --dry-run: Validate configuration with out working the backup.
• --no-progress: Disable progress indicators.
• --generate-manifest: Generate a backup manifest file.
• --generate-report: Generate a human-readable backup report.
• --version: Present model info.

For extra info, see the software README file.

Instrument code

Yow will discover the code for this software within the aws-samples repository. This software helps you get began shortly. Use it as a foundational reference to refine and adapt in your particular backup necessities.

Earlier than you implement a backup answer in your manufacturing surroundings, affirm that you simply:

• Evaluate and adapt the code to align together with your particular infrastructure necessities, safety insurance policies, and compliance requirements.
• Conduct thorough testing in a non-production surroundings to validate performance and efficiency.
• Implement applicable safety controls together with encryption, entry administration, and audit logging required by your group.
• Validate restoration procedures to verify your backup technique meets your outlined Restoration Time Goals (RTO) and Restoration Level Goals (RPO).
• Contemplate value optimization methods and monitoring to maintain the answer inside your operational finances.
• Keep away from concurrent software execution: This software depends on the AssetsAsBundle APIs, which have low throttling thresholds. The pattern software will not be designed to run a number of cases in parallel throughout the identical AWS account. If a number of groups want to make use of the software, contemplate implementing a concurrency management mechanism (for instance, a lock desk in DynamoDB or a database-level lock) to stop concurrent runs that would set off API throttling.

Scheduled execution

The pattern software described within the earlier part is designed for on-demand execution and is nicely fitted to getting began or working ad-hoc backups. For a production-grade backup technique, you would possibly wish to automate backup runs on a recurring schedule in order that your Fast Sight property are constantly protected with out handbook intervention.

This part outlines the high-level structure for a scheduled, totally automated backup answer. Detailed implementation and code for this structure are exterior the scope of this put up.

Structure overview

The scheduled execution structure is constructed on three AWS managed companies that work collectively to offer a dependable, serverless, and cost-effective automation pipeline:

• Amazon EventBridge is the scheduler. It triggers the backup workflow at an outlined cadence, for instance, day by day at midnight. EventBridge guidelines allow you to outline versatile cron-based or rate-based schedules with out managing any underlying infrastructure.
• AWS Step Capabilities is the orchestration layer. It coordinates the run of the person backup steps within the right sequence. Step Capabilities gives built-in error dealing with, retry logic, and execution historical past, which makes it nicely fitted to long-running workflows that span a number of API calls and asynchronous operations.
• AWS Lambda implements every particular person backup step as an impartial, stateless perform. Splitting the backup logic throughout a number of Lambda features addresses the time constraints inherent within the backup course of. Every export job is asynchronous and would possibly take a number of minutes to complete, relying on the quantity and measurement of property being exported.

Workflow steps

As a result of the end-to-end backup course of can take a big period of time, the automation is decomposed into discrete steps, every applied by a devoted Lambda perform. AWS Step Capabilities orchestrates these features in sequence, passing state between them and dealing with retries for transient failures. The workflow consists of the next steps:

1. Customers and teams backup: Retrieves all Fast Sight customers, teams, and group memberships utilizing the Fast Sight id APIs and persists the info to DynamoDB with date-based desk suffixes, as described within the Technical implementation part. This operation can run in parallel with the asset backup operations as a result of it doesn’t have any dependency.
2. Asset backup discovery: Lists all Fast Sight property within the goal Area (knowledge sources, datasets, analyses, and dashboards), applies the required filters to exclude unsupported asset sorts, and teams property into lists of as much as 100 objects every. The output of this step is handed to subsequent steps as enter.
3. Generate bundle: Initiates export jobs for all of the property included within the checklist specified because the enter parameter, polls for job completion, and uploads the ensuing ZIP bundles to the designated Amazon S3 prefix.
4. Test standing: Periodically polls the lively bundle execution and notifies the AWS Step Capabilities state machine when the export finishes.

The next diagram illustrates the high-level circulation of the scheduled execution structure.

Key design concerns

• Asynchronous polling: The check-status Lambda perform polls the job initiated by the generate-bundle Lambda perform utilizing the DescribeAssetBundleExportJob API till the job reaches a terminal state (SUCCESSFUL or FAILED). The check-status Lambda perform runs in a loop with a ready situation (for instance, 30 seconds) between calls.
• Parallelism: Configure an sufficient stage of parallelism to regulate the quantity of API calls carried out by the steps in your workflow, particularly on the generate-bundle step that calls the DescribeAssetBundleExportJob and StartAssetBundleExportJob APIs, which have low concurrent fee limits. You should utilize the inline map state MaxConcurrency area to restrict the variety of concurrent runs of the generate-bundle step.
• Error dealing with: Step Capabilities permits you to outline catch blocks and retry insurance policies at every stage. A failure in a single step (for instance, an unsupported asset kind) doesn’t abort all the backup run.
• Value: When scheduling is enabled, prices scale with backup frequency and retention interval. For steering on estimating storage prices, see the Value estimation part.

Value estimation

The next sections estimate the prices of working the backup software on Amazon S3 (for asset bundles) and DynamoDB (for person and group metadata).

Amazon S3: asset bundle storage

Asset bundles are compressed ZIP information uploaded to Amazon S3 after every export job. Based mostly on the answer design, every bundle of as much as 100 property averages roughly 500 KB when compressed.

Key takeaway: Amazon S3 storage prices for asset bundles are minimal. Even for very massive Fast Sight deployments with hundreds of property, the compressed bundle measurement stays within the low megabytes vary, leading to a month-to-month storage value nicely beneath $0.01.

Amazon DynamoDB: person and group metadata storage

Consumer and group info is saved in DynamoDB tables with date-based suffixes to protect backup historical past. DynamoDB storage is priced at roughly $0.25 per GB monthly (Normal desk class, on-demand mode).

Every merchandise saved in DynamoDB represents a single person or group definition (together with all related attributes comparable to ARN, electronic mail, position, group memberships, and backup timestamp). Based mostly on the schema described on this put up, the typical merchandise measurement is roughly 256 KB.

You should utilize this method to estimate the scale of your DynamoDB tables:

Desk measurement estimate = Variety of objects × Common merchandise measurement (256 KB)

Key takeaway: For small and medium organizations, DynamoDB storage prices stay minimal (below $0.10 monthly per backup snapshot). For giant organizations with tens of hundreds of customers, prices are nonetheless modest, within the low single-digit greenback vary per snapshot.

Abstract

For a single, unscheduled backup run, the full AWS value is successfully close to zero, dominated by a number of cents of Amazon S3 and DynamoDB storage at most. In case you implement scheduled backups (lined within the Scheduled execution part), prices scale linearly with backup frequency and retention interval. Even with day by day backups retained for 90 days, complete storage prices stay within the low single-digit greenback vary for many deployments. Think about using Amazon S3 Lifecycle insurance policies and DynamoDB Normal-IA to optimize prices as your backup historical past grows.

Conclusion

On this put up, we lined methods to design and implement a complete backup technique for Amazon Fast Sight property so you may keep enterprise continuity, meet regulatory necessities, and defend in opposition to knowledge loss.

We lined methods to use AssetsAsBundle APIs to programmatically export and protect essential BI property, together with dashboards, analyses, datasets, and knowledge sources, together with their dependencies and permissions. That will help you get began, this put up features a pattern automation software which you can take a look at and adapt to your group’s wants. The code orchestrates these APIs, shops asset bundles in Amazon S3, and preserves person and group info in DynamoDB for point-in-time restoration.

Prepared to guard your Fast Sight BI property? Get began as we speak by cloning the pattern backup software from the AWS Samples repository and testing it in your non-production surroundings. Start with a easy configuration to again up your most important dashboards, then broaden to a production-ready backup technique as you validate the method. To be taught extra about Amazon Fast Sight, see the Amazon Fast Sight Consumer Information.

In regards to the creator