Optimiser les performances d’un BaaS

When building mobile apps with Backend-as-a-Service (BaaS) platforms like Firebase or Supabase, performance is key. An efficient backend ensures your app is fast and responsive.

• Faster User Experience: Quick data loading keeps users engaged.
• Reduced Costs: BaaS platforms often charge based on usage (reads, writes, bandwidth).
• Scalability: Optimized systems handle more users without breaking a sweat.

Let's dive into techniques to make your BaaS backend shine!

BaaS platforms typically bill you for:

• Document Reads/Writes: Each time your app fetches or saves data.
• Storage: The amount of data stored in your database.
• Bandwidth: Data transferred to and from your app.
• Cloud Function Invocations: Each time your serverless function runs.

Inefficient data access directly increases your bill. Optimizing means doing more with less!

In traditional relational databases, you often normalize data to avoid redundancy. With NoSQL BaaS databases, denormalization is often preferred for performance.

Denormalization means duplicating data across multiple documents or collections to reduce the need for complex, costly queries or 'joins'.

Imagine a social app with posts and users. Instead of fetching a post, then separately fetching the author's details, you can denormalize:

• Normalized: posts/{postId} (contains authorId) and users/{userId} (contains name, avatar). Requires two reads.
• Denormalized: posts/{postId} (contains authorId, authorName, authorAvatar). Requires one read.

This speeds up common queries at the cost of slight data redundancy.

Just like a book's index helps you find information faster, database indexes speed up your BaaS queries.

• How it works: Indexes are special lookup tables that allow the database to find rows matching your query conditions more quickly.
• BaaS Indexes: Platforms like Firestore automatically create single-field indexes, but you often need to create composite indexes for queries involving multiple fields (e.g., where('category', '==', 'tech').orderBy('timestamp', 'desc')).

Always check your BaaS documentation for index management.

Fetching an entire database collection can be slow and expensive. Instead, use limiting and pagination:

• Limit: Use .limit(N) to fetch only a specific number of documents (e.g., 10 posts at a time).
• Pagination: For long lists, fetch data in chunks. Use cursors (e.g., .startAfter() or .endBefore()) to retrieve the 'next page' of results based on the last document fetched.

This drastically reduces reads and speeds up UI loading.

Be precise with your queries. Use filters (.where() clauses) to retrieve only the data that matches your exact criteria.

• Combine Filters: You can chain multiple .where() clauses. Remember that some combinations might require a composite index.
• Order Data: Use .orderBy() to sort your results. This is often combined with .limit() and pagination to provide a consistent user experience.

Efficient filtering and ordering minimize the data processed by your BaaS.

Even with optimized queries, repeated requests for the same data are inefficient. Implement client-side caching:

• In-memory cache: Store recently fetched data directly in your app's memory for quick access during a session.
• Persistent cache: For data that doesn't change often, store it locally using mechanisms like Shared Preferences, SQLite, or secure storage on the device.

This reduces BaaS reads and allows your app to work offline or load faster.

Many modern BaaS platforms offer powerful built-in offline capabilities. For example, Firebase Firestore automatically caches data locally and synchronizes it when the device comes online.

• Seamless Offline Experience: Users can continue interacting with previously fetched data even without an internet connection.
• Automatic Synchronization: Changes made offline are uploaded automatically when connectivity is restored.
• Reduced Reads: Your app reads from the local cache first, only hitting the server for new data or updates.

Always enable and configure these features if your BaaS platform supports them.

Test your knowledge on optimizing BaaS performance.

Great job! In this lesson, we explored crucial techniques for optimizing your BaaS backend:

• Smart Data Structuring: Use denormalization to reduce complex queries.
• Efficient Querying: Leverage indexes, limit results, paginate, and filter precisely.
• Strategic Caching: Implement client-side caching and utilize BaaS offline features.

By applying these strategies, your indie app will be faster, more cost-effective, and provide a smoother user experience. Keep building awesome apps!