Google DataStore and Google Firestore

 

 

Google Datastore and Google Firestore are both NoSQL, scalable database services provided by Google Cloud, but they differ in features, use cases, and underlying architecture. Here's a breakdown of the key differences between the two:

1. Architecture & API

• Google Datastore:

  • Originally designed as a highly scalable, NoSQL document database, it uses a schema-less data model.
  • Legacy product: Datastore is built on older technology, designed for large-scale applications with strong eventual consistency.
  • API: Uses the older Datastore API.

• Google Firestore:

  • Firestore is the newer, next-generation version of Datastore, offering improved features and flexibility. It is part of the Firebase and Google Cloud platforms.
  • Modern database: It uses a more structured, document-oriented data model, making it easier to use, with support for more advanced queries and real-time capabilities.
  • API: Uses a more modern Firestore API, which is more user-friendly and flexible compared to Datastore.

2. Consistency Model

• Google Datastore:

  • Eventual Consistency: Datastore operates with eventual consistency for queries (except ancestor queries, which are strongly consistent). This means there may be some delay before all replicas of data reflect recent changes.
  • Strong Consistency: Ancestor queries (queries that follow a specific entity group or hierarchy) in Datastore are strongly consistent.

• Google Firestore:

  • Strong Consistency by Default: Firestore supports strong consistency for both document reads and queries. This is true across global, multi-region setups, meaning you can always expect to get the latest data immediately.

3. Data Model

• Google Datastore:
  • Datastore stores data in the form of entities (key-value pairs) that can have multiple properties and belong to an entity group.
  • Entity Group Model: Data is structured hierarchically with entity groups, which help with strong consistency and transactions.
• Google Firestore:
  • Firestore stores data in the form of documents, which are JSON-like objects, organized in collections.
  • Hierarchical Structure: Firestore uses a collection/document model, where each document can contain sub-collections. This nested structure provides more flexibility in how you organize your data.

4. Real-time Updates

• Google Datastore:

  • Datastore does not natively support real-time updates. Applications need to poll the database for changes.

• Google Firestore:

  • Firestore provides real-time synchronization out of the box. Changes to the database can be instantly pushed to all connected clients, making it great for real-time applications like chat apps or live dashboards.

5. Transactions

• Google Datastore:

  • Supports ACID transactions, but they are limited to entity groups. Complex transactions that span multiple entity groups can be more challenging to implement.

• Google Firestore:

  • Firestore also supports ACID transactions but across any set of documents. Transactions in Firestore are more flexible, allowing multiple reads and writes to different parts of the database.

6. Query Capabilities

• Google Datastore:
  • Datastore supports basic queries like filtering, ordering, and projections. However, its query capabilities are somewhat limited compared to Firestore.
• Google Firestore:
  • Firestore supports more advanced querying capabilities, such as:
    • Compound queries (multiple filters combined).
    • In/array membership filters.
    • Range queries.
    • Real-time listeners for query results.

7. Indexing

• Google Datastore:
  • Datastore automatically indexes your data, but you have to manually configure composite indexes for complex queries.
• Google Firestore:
  • Firestore automatically handles a broader set of index types and supports composite indexes without much manual configuration.

8. Pricing

• Google Datastore:
  • Pricing is based on the number of reads, writes, deletes, and storage used. It typically operates on a pay-per-use basis, with costs related to the size and frequency of operations.
• Google Firestore:
  • Firestore’s pricing is also based on operations, but it includes additional charges for real-time synchronization. Firestore can be more expensive for high-frequency operations, especially when using real-time updates.

9. Integration with Firebase

• Google Datastore:

  • Datastore is part of Google Cloud, but it is not integrated with Firebase directly.

• Google Firestore:

  • Firestore is a core part of Firebase, offering seamless integration with Firebase features such as Firebase Authentication, Firebase Cloud Functions, and Firebase Hosting, making it a strong choice for mobile and web app development.

10. Migration Path

• Google Datastore:
  • Datastore is still supported by Google Cloud, but Google encourages migration to Firestore. Firestore has a Datastore mode, which retains the Datastore API while running on Firestore’s improved infrastructure.
• Google Firestore:
  • Firestore in Datastore mode provides backward compatibility for Datastore users who want to take advantage of Firestore’s infrastructure without changing their application code.

---

Summary of Differences:

Feature	Google Datastore	Google Firestore
API	Older Datastore API	Modern Firestore API
Consistency	Eventual & strong for some	Strong consistency by default
Real-time Updates	No	Yes, built-in
Data Model	Entity-based	Document-based (collections)
Query Capabilities	Basic	Advanced & real-time queries
Transactions	Limited to entity groups	More flexible transactions
Pricing	Pay-per-use (reads/writes)	Pay-per-use + real-time cost
Firebase Integration	No	Yes, native integration

In general, Firestore is a more modern, feature-rich, and versatile option compared to Datastore. Firestore is better suited for applications requiring real-time updates, flexible queries, and Firebase integration, while Datastore is still a good choice for legacy applications and simpler use cases.