A lot of gaps in event-store


We’re using Axon Server as routing mechanism and SQLServer as storage for the domain events. The event stream is tracked using a GapAwareTrackingToken.

In our testing environments, we have the requirement to ‘refresh’ a client.
A client consists of multiple aggregates. Refresh means that we’ve to delete all events related to the client from the event-store and replay all projections we have. Afterwards we load the client again from an external source.
The erase of events causes huge gaps in the event stream. In extreme cases, 10 000 events might have been deleted.

As mentioned, the GapAwareTrackingToken stores all gaps in the token’s payload.
This payload can grow significantly and can lead to long transaction times, even transaction timeouts.

Is there anything we can do to prevent these long transaction times?