atc-net / atc-cosmos-eventstore Goto Github PK

Describe the bug
I have an issue where I have an event that contains a TimeOnly type which was added in .NET 6. Sadly .NET 6 does not come with build in JsonConverter for TimeOnly and DateOnly.

These are easily to implement as a custom JsonConverter however there are no way to add these converter to the Event Store.

A "workaround" is to add the converters to the TimeOnly/DateOnly properties in the event object through JsonConverterAttribute, like so:

[StreamEvent("event-name:v1")]
public record TimeEvent(
    [property: JsonConverter(typeof(JsonTimeOnlyConverter))]
    TimeOnly Time,

This works for simple cases but not if the I have more complex objects that contains TimeOnly further down. Then I would have to make custom converters for these types, only to be able to handle deserialization of a single properties within it.

My code could also end up with a lot of these attributes if a have a lot of different events requiring custom converters.

To Reproduce
Steps to reproduce the behavior:

1. Create an event with a TimeOnly or DateOnly type
2. Write the event to the event store
3. An exception is thrown with the error System.NotSupportedException : Serialization and deserialization of 'System.DateOnly/TimeOnly' instances are not supported

Expected behavior
I am able to add custom JsonConverters to the Event Store client that are used when writing to, or reading from, the Event Store.

LARS og Egil jam

Persisting a large batch of events in a single command throws an exception that doesn't clearly explain why the operation failed

CosmosDb Transactional batch operation limitations
Currently, there are two known limits:

• The Azure Cosmos DB request size limit constrains the size of the TransactionalBatch payload to not exceed 2 MB, and the maximum execution time is 5 seconds.
• There's a current limit of 100 operations per TransactionalBatch to ensure the performance is as expected and within SLAs.

Expected behavior
Throw a more concrete error message in a specialized exception type would be the preferred way to describe why the operation failed