Skip to content

Domain modeling with TypeScript and FModel library, utilising Deno runtime

Notifications You must be signed in to change notification settings

fraktalio/fmodel-deno-demo

Repository files navigation

fmodel-deno-demo

TECHNOLOGY: Fmodel, TypeScript, Deno, Deno KV, Deno Deploy

A demo/example project for the imaginary restaurant and order management.

It demonstrates how to run our unique domain and orchestrate it in an EventSourced/EventDriven way.

event model image this blueprint is an outcome of the event-modeling process

Fmodel

This project is using Fmodel - TypeScript library.

Fmodel is:

  • enabling functional, algebraic and reactive domain modeling with Typescript programming language.
  • inspired by DDD, EventSourcing and Functional programming communities, yet implements these ideas and concepts in idiomatic TypeScript, which in turn makes our code
    • less error-prone,
    • easier to understand,
    • easier to test,
    • and type-safe.
  • enabling illustrating requirements using examples
    • the requirements are presented as scenarios.
    • a scenario is an example of the system’s behavior from the users’ perspective,
    • and they are specified using the Given-When-Then structure to create a testable/runnable specification
      • Given < some precondition(s) / events >
      • When < an action/trigger occurs / commands>
      • Then < some post condition / events >

First steps

Run the tests

We are using the Given-When-Then structure to create a testable specification:

  • Given < some precondition(s) >
  • When < an action/trigger occurs >
  • Then < some post condition >
deno test

Run the application (HTTP Server)

deno run --unstable-kv --allow-net --allow-read --allow-write server.ts

Run the client

This is a simple client that sends a create restaurant command and a change restaurant menu command to the server.

deno run --allow-net client.ts

Deno KV as an event store

Deno KV is a key-value database built directly into the Deno runtime, available in the Deno.Kv namespace. It can be used for many kinds of data storage use cases. Deno KV is available in the Deno CLI and on Deno Deploy (Hassle-free platform for serverless JavaScript/TypeScript applications).

Deno KV is built on FoundationDB, capable of handling millions of operations per second. You know what else is built on FoundationDB? iCloud, Snowflake, and more.

Modeling event sourcing

To model event sourcing in Deno KV properlly, we are going to use a very simple Key schema eventsByStreamId.<streamId>.<eventId> for our Events.

  • In this approach, each event is uniquely identified by a combination of a stream ID and an event ID. The stream ID represents the stream or aggregate/decider to which the event belongs, while the event ID is a unique identifier for the event within that stream.
  • This format allows events to be grouped by stream and ordered within each stream based on the event ID.

Optimistic locking

Optimistic locking is a concurrency control mechanism used to prevent conflicts between multiple processes that are attempting to modify the same data concurrently. In the context of event sourcing, optimistic locking ensures that updates to an aggregate (or stream) are performed atomically and consistently, even in the presence of concurrent writes.

The Deno KV store utilizes optimistic concurrency control transactions rather than interactive transactions like many SQL systems like PostgreSQL or MySQL. This approach employs versionstamps, which represent the current version of a value for a given key, to manage concurrent access to shared resources without using locks. When a read operation occurs, the system returns a versionstamp for the associated key in addition to the value.

When introducing a new key schema like lastStreamEvent.<streamId>, you can leverage it to implement optimistic locking by tracking the version of each stream. Here's how you can do it:

  1. Incrementing Stream Version:
  • Each time an event is appended to a stream, the version of that stream is incremented. This version represents the last event that have been appended to the stream.
  • When appending a new event to a stream, you include the current version of the stream in the event's metadata.
  1. Checking Stream Version during Update:
  • When updating a stream, you retrieve the current version of the stream from the event store.
  • Before appending a new event to the stream, you compare the retrieved version with the version included in the event to be appended.
  • If the versions match, it indicates that no other updates have occurred since the version was retrieved, and it's safe to append the event.
  • If the versions don't match, it indicates that another update has occurred concurrently, and you may need to handle the conflict (e.g., by retrying the operation, merging changes, or notifying the user).

Modeling event streaming

When appending events to the event store, in addition to appending them to their respective streams (eventsByStreamId.<streamId>.<eventId>), you can also append them to the global stream. The Key schema for the global stream might look like this: events.<eventId>. To read all events ordered by event ID, you simply query the global stream. As all events are appended to this stream, you get a comprehensive view of all events in the system.

Events are enqueued into the Deno KV queue to ensure reliable delivery to interested query-side components, like Materialized Views.

  1. Advantages:
  • Simplified Querying: You can query events across all streams in a single operation, simplifying event retrieval and processing.
  • Comprehensive View: The global stream provides a comprehensive view of all events in the system, facilitating analysis and reporting.
  1. Considerations:
  • Storage Overhead: Maintaining a global stream requires additional storage space to store redundant copies of events.
  • Consistency: Ensuring consistency between individual streams and the global stream may require additional synchronization mechanisms to prevent data inconsistencies (Deno suports this in a nice way with atomic transactions).

Created with ❤️ by Fraktalio

Excited to launch your next IT project with us? Let's get started! Reach out to our team at info@fraktalio.com to begin the journey to success.

About

Domain modeling with TypeScript and FModel library, utilising Deno runtime

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published