Event Store retry best practices.


We are sending events from multiple sources to Event Store using an ESB. We are using tcp to send the events via the ESB.

I am interested in what is considered best practice for handling retry operations.

In our solution we have a unique event id for every event. The event stream is based on a unique ID tied to an artifact in the solution data model.

The ESB will not know if the stream exists and so uses ExpectedVersion.Any.

I read Optimistic concurrency and idempotence in the documentation. It appears to me that a small retry time windows should be acceptable to

avoid duplicates.

Is there anything I should be concerned about in retrying for timeouts or other connectivity issues?

Thanks for any input.