Flux.skip and Flux.take

[KarlHakanNordgren]

Hi! Do the above two methods push the work down to the repository? (Or do the items get pulled from the repository and then get skipped in the caller?).

Thanks!

[KarlHakanNordgren]

@dstepanov : Sorry to pester you always. Do you know the answer to this one?

[KarlHakanNordgren]

Gotcha. Thanks!

[KarlHakanNordgren]

Sorry one more (potentially a bit silly) question: If I do Flux.take(4), it does not fetch records 5, 6, 7, etc. correct? (Is there documentation for this or can you point me to the place in the source so I can have a look?)

[dstepanov]

I think it will fetch everything or at least some internal batch of records. Are we talking about R2DBC backend, right?
Technically, Reactor might ask just for 4 records and then close the subscriber, then its up to how upstream is implemented.

[KarlHakanNordgren]

The backend is MongoDB. Does the above hold true for MongoDB?

[dstepanov]

It depends on how the driver is implemented, but I would assume it will be the same. You should use skip and limit when fetching results instead of reactive skip operators.