I've been looking into using triggers for a streaming job, but the differences between Continuous trigger vs processingTime trigger are not clear to me.

As far as I've read on different sites:

1. continuous is just an attempt to make the streaming almost real-time instead of micro-batch based (using much lower latency of 1ms).
2. As of the time of writing this question, only supports a couple of sources and sinks like Kafka.

Are these two points the only differences between the two triggers?

CodePudding user response：

You are pretty much right Structured Streaming continuous got added in order to respond to low latency needs by achieving near-real-time processing using a continuous query, unlike the old batch way where the latency is depending on processing time and the batch job duration (aka micro-batch query)

the docs are pretty useful to get more in-depth.