StreamStore.Storage.EventFlow 0.1.0

StreamStorage for EventFlow

Adapter of StreamStore's storage backends to use as EventFlow's event store.

What

Implements IEventPersistence interface uses IStreamStorage interface implementation of any StreamStore's storage backends.

Why

Why would you need this?

Despite the fact that EventFlow solution already contains many implementations of persistence, such as SQL, EventStore, ElasticSearch etc., there is not so many NoSQL backends provided.

So, if you are using or decided to use EventFlow want to high available distributed storage, such as Cassandra, you should consider to use this component.

How

How to use it? It is pretty easy.

First, if you just decided to use EventFlow, you must install EventFlow package:

  dotnet add package EventFlow

Then install adapter:

  dotnet add package StreamStore.Storage.EventFlow

After that install particular StreamStore storage backend, for instance Cassandra:

dotnet add package StreamStore.NoSql.Cassandra

Finally, register and configure EventFlow in DI container:

using var serviceCollection = new ServiceCollection()
    // ...
    .AddEventFlow(e => 
    // ...
    // Register adapter
      e => e.UseStreamStorageEventStore(services =>
      // Register and configure particular storage backend, for instance Cassandra
      // How to configure storage backend, you can find in documentation of particular one.
      // For instance here:
      // https://github.com/kostiantyn-matsebora/streamstore/tree/master/src/StreamStore.NoSql.Cassandra#configuration-options
        services.UseCassandra(...);
     )
    )
    // ...
    .BuildServiceProvider();

More information about EventFlow configuration you can find in official documentation.

Multitenancy

Despite the fact, that EventFlow itself does not have explicit support of multitenancy, this component provides such ability in a tricky way.

To be able to use adapter in multitenant mode, you must create and register implementation if ITenantIdResolver interface, responsible for resolution of tenant identifier based on current context, for example HtttpContextTenantIdResolver resolves tenant identifier based on HTTP request:

using var serviceCollection = new ServiceCollection()
    // ...
    .AddEventFlow(e => 
    // ...
    // Register adapter
      e => e.UseStreamStorageEventStore<HtttpContextTenantIdResolver>(services =>
      // Register and configure particular storage backend, for instance Cassandra
      // How to configure storage backend, you can find in documentation of particular one.
      // For instance here:
      // https://github.com/kostiantyn-matsebora/streamstore/tree/master/src/StreamStore.NoSql.Cassandra#configuration-options
        services.UseCassandraWithMultitenancy(...);
     )
    )
    // ...
    .BuildServiceProvider();

Schema Provisioning

Automatic schema provisioning is a part of StreamStore solution implemented as background service and not available as part of component.

However, you can still use schema provisioning implementation of particular storage.

For single tenant mode:

var serviceProvider = serviceCollection.BuildServiceProvider();

// Get provisioner from DI container
var provisioner = serviceProvider.GetRequiredService<ISchemaProvisioner>();

// Run schema provisioning
await provisioner.ProvisionSchemaAsync(token);

For multitenancy:

var serviceProvider = serviceCollection.BuildServiceProvider();

// Get provisioner factory from DI container
var provisionerFactory = serviceProvider.GetRequiredService<ITenantSchemaProvisionerFactory>();

// Iterate tenants
foreach (var tenantId in tenants) {
  // Get provisioner for particular tenant
  var provisioner = provisionerFactory.Create(tenantId);
  // Run schema provisioning for tenant
  await provisioner.ProvisionSchemaAsync(token);
}