HouseofCat.Compression 3.1.3

RabbitMQ & Dataflows

A RabbitMQ library of .NET tools to help rapidly develop well performant services or to just help manage durable connectivity with the RabbitMQ.Client!

Why Make A RabbitMQ Powered Dataflow?

Dataflows have configurable concurrency, serialization, monitoring, compression, and encryption all as first class citizens. This paradigm allows developers to just focus on the important stuff - getting work done

Here are some features ready with RabbitMQ and Dataflows today!

Queueing

• Async Processing, batch processing, consumer cloning and more!
• Queue-based Retriability via Ack/Nack.
• Async Error Handling (simplify functional error handling by allowing it to throw.)
• A RabbitMQ ConnectionPool and ChannelPool (connection durability) provided by namespace HouseofCat.RabbitMQ.Pools;.

Built-Ins

• Supports ILogger<T> via LogHelpers.
• Configurable concurrency/parallelism, no code changes required.
• Contracted IWorkState simplifies functional generic returns and integration.
• Has Json (System.Text.Json and Newtonsoft) and MessagePack serialization providers.
• Allows seamless encryption/decryption steps.
• Allows seamless compression/decompression steps.
• Async Error Handling with Predicate triggers and an actionable callback.

Core Interchangeability

• Allows you to replace serialization provider with ISerializationProvider and have basic implementations.
• Allows you to replace encryption provider with IEncryptionProvider and have basic implementations.
• Allows you to replace compression provider with ICompressionProvider and have basic implementations.

Business Logic

• All steps process in the order provided allowing you to still control order of execution.
• All automatically subscribed to Async Error handling by WorkState.IsFaulted flag.

Testing

• All built-in steps will have integration tests that should remove concerns from end-user developer.
• Future case will include much more complex abstract UnitTesting as time allows.
• The developer should only need to unit test their functional business code.

Implicit Benefits

The benefits of a dataflow pattern extend beyond fancy machine learning and Tensorflows or high throughput GCP Dataflow for mass computation. When brought to the service level, it helps organize your code into more manageable blocks. You can still write monolithic functions, but you would be hamstringing yourself and scarificing concurrency and parallelism. By designing code into small functional steps, you always write better, cleaner, code reduced with cyclomatic complexity. That very same code is easier to UnitTest. The orchestration of the function calls are the order they are added allowing you extend the original functionality infinitely. You don't have to write deserialization or post-processing encryption/compression as they all baked in. Designing from the ground up with concurrency and parallelism, you stay nimble and fast - able to scale up internally, before horizontally and vertically, saving costs. All without needing code changed or refactored.

Lastly, after everything is said and done, all your business code is re-usable. Should you decide to abandon this workflow (:worried:) for a different mechanism, engine, or what not, all of your code will happily port to whatever other project / flow you are working with and so will all your testing making it a win win.

Help & Guides

• Getting started with a RabbitMQ.Dataflow.ConnectionPool.
• Getting started with a RabbitMQ.Dataflow.ChannelPool.
• Getting started with a ChannelPool and IModel.BasicPublish.
• Getting started with a ChannelPool and IModel.BasicGet.
• Getting started with a ChannelPool and IModel.BasicConsume.
• Getting started with RabbitMQ.Dataflow Serialization.

You can also find various library examples inside the tests/UnitTests or the tests/RabbitMQ.Console.Test project.

Status

Test Server: Windows 11
RabbitMQ Server: v3.13
Erlang: v26.2.3

Main RabbitMQ Library

HouseofCat.RabbitMQ

A library that focuses on RabbitMQ connection and channel management to create fault tolerant Publishers and Consumers.
Formerly called CookedRabbit.Core/Tesseract.

DataFlow Library

HouseofCat.Dataflows

A library that provides the base magic Dataflows for RabbitMQ.Dataflows.

• Custom TPL Block - ChannelBock used as a Channel-based BufferBlock<TIn>
• Has DataFlowEngine and ChannelBlockEngine.
• Has Pipelines (Dataflow alternative).

Core Productivity Libraries

These libraries are here to help you build powerful Dataflows for your messages.

HouseofCat.Serialization

A library that has a collection of .NET ISerializationProvider or the interface to make your own.

• Supports MessagePack and System.Text.Json and Newtonsoft.Json.

HouseofCat.Compression

A library that has a collection of .NET ICompressionProvider or the interface to make your own.

• Supports LZ4, Gzip, Brotli, and Deflate.
• Supports RecyclableMemoryStream variants.

HouseofCat.Hashing

A library that focuses on implementing hashing.

• Supports Argon2.

HouseofCat.Encryption

A library that provides encryption contracts and the base AesGCM/AesCBC .NET IEncryptionProvider as well as the interface to make your own.

• Supports AesCbc via CryptoStream (good for encrypted file/memorystreams).
• Supports .NET AesGcm 128, 192, 256 (non-streams).
• Supports BouncyCastle AesGcm 128/192/256.
• Supports RecyclableMemoryStream variants.

HouseofCat.Metrics

A library that has a collection of .NET IMetricsProvider or the interface to make your own.

• Supports Prometheus.Net.

Non-Critical Library Integrations

HouseofCat.Data

A library that provides the provides helper classes for data manipulation and transformation.

Also provides Database abstractions, a simple Dapper integration, and SqlKata integration SQL query generation.

Database Connection Factory Support

• System.Data.SqlClient
• Microsoft.Data.SqlClient
• MySql.Data.MySqlClient
• Npgsq
• MySql.Data
• Oracle
• SQLite

HouseofCat.Utilities

A library that focuses on general purpose utilities and functions that simplify the coding experience.

HouseofCat.io