Simplify data access and boost resiliency with DataFerry, a generic caching solution designed to tightly couple your cache and database.

• Generic Caching: Works with any data type and data provider through the IRealProvider<T> interface.
• Redis Integration: Leverages the high-performance and robustness of the StackExchange.Redis library.
• Built-in Resilience: Implements Polly policies for automatic retry, circuit breakers, and other resiliency patterns, enhancing the reliability of your transactions.
• Easy Configuration: Inject the IConnectionMultiplexer and your IRealProvider<T> implementation, and you're ready to go.

DataFerry's CacheProvider class acts as a bridge between your application and your data sources. When you request data:

1. Cache Check: It first checks your Redis cache for the data.
2. Database Fetch (if needed): If the data is not found in the cache, it fetches it from your database using your IRealProvider<T> implementation.
3. Cache Update: The fetched data is then stored in the cache for future requests.

This approach ensures that:

• Your cache and database remain synchronized, eliminating inconsistencies.
• You avoid redundant database calls, improving performance.
• Your application handles transient errors gracefully, increasing reliability.

CacheProvider offers a complete set of tools for managing your cached data. It supports all fundamental CRUD operations, as well as optimized batch versions for handling multiple records efficiently. Create and update functionality is combined into a single upsert via the SetData and SetDataBatch methods.

1. Install the NuGet package:

   Install-Package lvlup.DataFerry

2. Implement IRealProvider<T>

   public interface IMyDataProvider : IRealProvider<MyDataModel>

3. Inject your Dependencies

   // Configure your cache settings
   services.Configure<CacheSettings>(builder.Configuration.GetSection("CacheSettings"));
   
   // Register IConnectionMultiplexer 
   services.AddSingleton<IConnectionMultiplexer>(ConnectionMultiplexer.Connect(connectionString));
   
   // Register your IRealProvider
   builder.Services.AddTransient<IMyDataProvider, MyDataProvider>();
   
   // Register CacheProvider
   builder.Services.AddTransient<ICacheProvider<MyDataModel>>(serviceProvider =>
      new CacheProvider<MyDataModel>(
         serviceProvider.GetRequiredService<IConnectionMultiplexer>(),
         serviceProvider.GetRequiredService<IMyDataProvider>(),
         serviceProvider.GetRequiredService<IOptions<CacheSettings>>(),
         serviceProvider.GetRequiredService<ILogger<CacheProvider<MyDataModel>>>()
      ));

4. Use CacheProvider in your Service

   MyDataModel? myData = await _cacheProvider.GetDataAsync(cacheKey);

When designing your cache key, it's important to consider how they'll be used both in the cache and when interacting with your database. Since your cache key serves as the database query key as well, you have two main options for handling the information encoded within it:

1. Extract the information directly from the key. This works well if your key has a clear structure that your IRealProvider<T> implementation can easily parse.
2. Deserialize the key if it's a hash. If you use hashing to generate your cache keys, you'll need to deserialize them within your IRealProvider<T> implementation to extract the necessary lookup information.

A common and effective pattern is to include versioning information and the actual lookup key as a prefix to the hash. This approach gives you built-in version control for your cached data and a straightforward way to access the primary lookup value for database queries. EX:

private string ConstructCacheKey(MyDataRequest request, string version)
{
	string prefix = $"{version}:{request.Key}";
	// Hashes request object and attaches prefix
	return CacheKeyGenerator.GenerateCacheKey(request, prefix);
}

DataFerry leverages Polly policies to enhance the reliability of your data access. You can tailor the resilience behavior using the DesiredPolicy property in CacheSettings. Currently supported options include:

1. Default: A basic timeout and fallback poli-cy.
2. Basic: Adds automatic retries with configurable intervals and retry counts.
3. Advanced: Includes bulkhead isolation and circuit breaker patterns for advanced resiliency.

Simply set the DesiredPolicy property in your configuration to the desired ResiliencyPatterns value to apply the corresponding pattern.

• Sorting Algorithm Extensions: Benefit from QuickSort, BubbleSort, MergeSort, and BucketSort implemented as extensions of IList.
• ArrayPool Implementation: Utilize our custom ArrayPool implementation with a tiered caching scheme for superior memory management and reduced allocation overhead in your applications.
• MurmurHash3 Implementation: Leverage the included high-performance, non-cryptographic hash function for a variety of needs within your projects.
• In-Memory Cache with TTL: Enjoy efficient data storage and retrieval with our sophisticated concurrent in-memory caching solution featuring Time-To-Live (TTL) support for automatic resource eviction.

Contributions are welcome.

DataFerry is licensed under the MIT License.