Add a single repartition join diagram

pull/7226/head
Marco Slot 2023-09-18 13:53:45 +02:00 committed by Önder Kalacı
parent d667c22f70
commit 659683a4c7
2 changed files with 1 additions and 5 deletions

Binary file not shown.

After

Width:  |  Height:  |  Size: 168 KiB

View File

@ -1672,11 +1672,7 @@ Note that these stages are run in parallel for all dependent jobs (read: all tab
Once all dependent jobs are finished, the main Job is executed via the regular adaptive executor code path. The main job will include calls to read_intermediate_result that concatenate all the intermediate results for a particular hash range. Once all dependent jobs are finished, the main Job is executed via the regular adaptive executor code path. The main job will include calls to read_intermediate_result that concatenate all the intermediate results for a particular hash range.
A diagram of a machine ![Example of single hash repartition join](../../../images/single-repartition-join.png)
Description automatically generated
repartitioning.pptx
Dependent jobs have similarities with subplans. A Job can only be distributed query without a merge step, which is what allows the results to be repartitioned, while a subplan can be any type of plan, and is always broadcast. One could imagine a subplan also being repartitioned if it is subsequently used in a join with a distributed table. The difference between subplans and jobs in the distributed query plans are one of the most significant technical debts. Dependent jobs have similarities with subplans. A Job can only be distributed query without a merge step, which is what allows the results to be repartitioned, while a subplan can be any type of plan, and is always broadcast. One could imagine a subplan also being repartitioned if it is subsequently used in a join with a distributed table. The difference between subplans and jobs in the distributed query plans are one of the most significant technical debts.