Skip to content

Commit e2a1b7f

Browse files
committed
[docs] Update ASOF join section in time-series tutorial
Signed-off-by: Leonid Ryzhyk <ryzhyk@gmail.com>
1 parent bc7c2b7 commit e2a1b7f

File tree

1 file changed

+5
-7
lines changed

1 file changed

+5
-7
lines changed

docs.feldera.com/docs/tutorials/time-series.md

Lines changed: 5 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -655,13 +655,11 @@ ON purchase.customer_id = customer.customer_id;
655655

656656
#### Garbage collection
657657

658-
The incremental as-of join operator stores both of its input relations. Feldera
659-
currently implements GC for the left input only: if both timestamp columns
660-
in the `MATCH_CONDITION` have waterlines, the operator will discard old records
661-
below the smaller of the two waterlines.
662-
663-
GC for the right input is on our [roadmap](https://github.com/feldera/feldera/issues/1850).
664-
[Let us know](https://github.com/feldera/feldera/issues/new/) if you are interested in this feature.
658+
The incremental as-of join operator stores both of its input relations. If both
659+
timestamp columns in the `MATCH_CONDITION` have waterlines, the operator will
660+
- discard records below the smaller of the two waterlines from the left collection
661+
- discard all except the last record below the smaller of the two waterlines for each join key
662+
from the right collection.
665663

666664
### `LAG` and `LEAD`
667665

0 commit comments

Comments
 (0)