[Feature Request]: Automatic Hash for upsert (Insert/Update) component
- Dominant language
- Java
- Stars
- 1.5k
- Forks
- 476
- Avg merge
- 19h
- Merged PRs (30d)
- 227
Description
### What would you like to happen?
Hello
As you know, upsert is super time consuming, i have simple example 200k rows takes approx 49 sec (capture 1)
the best practice for huge table is to go through hash and just update when hash is different instead of comparing all fields see capture 2
we go from 49 to 2 sec for the exact same table and the exact same fields , this is as you know is widely common practice , but as you can see in capture 2 it take longer to devellop and all pipeline builder do this manually because of performance
it would be super nice to have a tab or any way to simplify this the UI woulds ask
do you want to use hash for change validation?
if yes ? what hash type ?
whats the column name source (optional , if any calculated on the fly if not provided)
whats the destination column (not optional)
this would be a killing feature ! 3 option to fill up instead of a entire dev
all the necessary code is already in the checksum component, so, I dont think it would take much effort to implement this
nice isn't it ?
### Issue Priority
Priority: 2
### Issue Component
Component: Transforms
Contributor guide
Research direction
Start by reviewing the existing checksum component and the upsert (Insert/Update) transform, since the issue identifies both as relevant entry points. Determine how the UI could expose hash type, optional source column, and required destination column, then verify that hash-based change validation supports the requested performance improvement.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- java
- Domain
- data-engineering
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 38/100