Since all contexts receive clock ticks, timers work in all contexts. However, it is possible for some contexts to run behind others. That is, a timer in a particular monitor for which there are monitor instances in multiple contexts might fire at different points in real time. In each context, the timer can process the series of clock ticks at a speed that is different from the other contexts.

A context that is running a monitor instance in a very long running loop might not remove entries from its input queue for a long time. If a context has a full input queue the clock tick distributer thread does not block. Instead, the correlator quashes clock ticks onto the end of the context’s input queue. This means that the correlator unpacks the clock tick event when the context input queue either drains or accepts a new event. There is no perceptible difference between normally received clock ticks and quashed clock ticks.