Amazon CloudSearch Cheat Sheet

Amazon CloudSearch is a managed service in the AWS Cloud that makes it simple and cost-effective to set up, manage, and scale a search solution for your website or application.

Core Concepts

• Search Domain: A search domain encapsulates a collection of data you want to make searchable, along with the configuration settings for that search solution. It includes your searchable data and the search instances that handle indexing and search requests.

• Documents: These are the individual items you want to search. You upload data to your search domain as a collection of documents in either JSON or XML format. Each document has a unique ID and one or more fields that contain the data you want to search and return in results.

• Index Fields: You configure index fields to specify how you want to search and display your data. An index field can be a text, literal, date, or numeric type.

• Search Instances: These are the underlying EC2 instances that index your data and process search requests. CloudSearch automatically scales the number and type of search instances based on your data volume and traffic.

Key Features

• Fully Managed: CloudSearch handles the provisioning, setup, configuration, patching, and monitoring of the search infrastructure.

• Simple to Use: You can create a search domain and upload your data through the AWS Management Console, CLI, or SDKs.

• Auto-Scaling: Automatically scales your search domain's resources up or down to meet demand, ensuring high performance and availability.

• Rich Search Functionality: Supports a wide range of search features, including:

  • Faceted search

  • Free text, Boolean, and structured queries

  • Customizable relevance ranking

  • Autocomplete suggestions (suggesters)

  • Highlighting

  • Geospatial search

• High Availability: Offers a high availability option by distributing your search domain across two Availability Zones (AZs). If one AZ becomes unavailable, CloudSearch automatically fails over to the other.

How It Works

1. Create a Search Domain: Define a name for your search domain and configure its settings.

2. Configure Indexing Options: Define the fields you want to include in your index. For each field, you specify its type and how it can be used (e.g., searchable, facetable, returnable).

3. Upload Data: Format your data as a batch of documents in JSON or XML and upload it to your search domain for indexing.

4. Search Your Data: Send search requests to your domain's unique search endpoint. You can specify query criteria, filters, facets, and sorting options.

5. Process Results: CloudSearch returns search results in JSON or XML format, which you can then process and display in your application.

Indexing and Data Management

• Analysis Schemes: Control how the content of text fields is processed during indexing. This includes language-specific text processing like tokenization, stopword removal, and stemming.

• Facets: A facet is an index field that represents a category you want to use to refine and filter search results. For example, if you're searching for products, you might use facets for brand, price range, and color.

• Suggesters: Provide autocomplete functionality for your search bar. As a user types, a suggester provides a list of potential matching phrases from your data.

• Expressions: You can define custom numeric expressions to use for sorting search results. This allows for fine-grained control over relevance ranking.

Scaling and Availability

• Automatic Scaling: CloudSearch handles scaling for both data volume and request traffic.

  • When your data volume grows, CloudSearch can increase the partition count or upgrade the instance type.

  • When request traffic increases, CloudSearch can add more search instances.

• Partitions: A search domain's index is divided into partitions. Each partition is stored on a separate search instance. Splitting the index across multiple instances enables CloudSearch to process requests in parallel for faster performance.

• High Availability (Multi-AZ): To increase fault tolerance, you can enable the Multi-AZ option. This creates a redundant set of search instances in a second Availability Zone. CloudSearch automatically syncs the index between the AZs and will fail over if the primary AZ becomes unavailable.

Monitoring

You can monitor the activity and health of your search domains using several tools:

• CloudSearch Console: The console dashboard provides key metrics and operational status.

• Amazon CloudWatch: CloudSearch automatically sends metrics to CloudWatch, allowing you to track performance, usage, and set alarms. Key metrics include NumberOfDocuments, SearchableDocuments, SuccessfulRequests, and ThrottledRequests.

• AWS CLI/SDKs: You can retrieve status and metrics programmatically.

Pricing

Amazon CloudSearch pricing is based on several components:

• Search Instances: An hourly rate based on the type and number of search instances running.

• Document Batch Uploads: A rate per batch of documents uploaded for indexing.

• IndexDocuments Requests: Older pricing model fee, less common now.

• Data Transfer: Standard AWS data transfer fees apply.