Kafka Couchbase Connector Issue

Kafka Connector
11

@david.nault

I wanted to jump in on

Note that if the connector has had a chance to save its checkpoints, it won’t start over from the beginning unless you rename the connector or set the couchbase.stream.from property to BEGINNING

I am also setting up a pipeline and have this question in this forum.

But this potentially solves it all.

If my connect cluster dies and then comes up again and is set to stream from lastSaved (or beginning) will it receive only the changes since the last checkpoint? If so, how much downtime can it have before the relevant cdc/wal/dcp logs get rolled over and can we configure this?

(I tested this behavior using a standalone connector deployment and simply got from the beginning behavior but the guys at confluence told me that the standalone deployment does not use the same checkpoint offset system the distributed deployment does which gives me hope)

Also, just to clarify fromTheBeginning really only means the current state of the data is ins’t a literal replay of the history (at least that what it looks like from my tests), is that correct?

Thank you so much for your help.