Reinvent Event Broadcasters

Like any sane engineering team we use library provided event aggregation to coalesce down events while server fails over nodes.

Rather inconveniently it rate limits to 25 events over a 5 minute period (non configurable), which then results in down and failed over events going missing.

This then has the knock on effect that where tests are relying on particular events for synchronization, they may never happen.  Also this screws with event validation.

Sadly we need to reinvent the wheel and make our own event caching and aggregation layer to avoid rate limiting and event filtering.

TestLogCollectWithDefaultRetentionAndSize and TestLogCollectWithCustomRetentionAndSize are particularly affected by this so re-enable event checking and use that for your test driven development.