Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[BUG] Qualification: Jobs with UDFs may have inaccurate speedups #716

Closed
kuhushukla opened this issue Jan 8, 2024 · 4 comments
Closed
Assignees
Labels
bug Something isn't working core_tools Scope the core module (scala)

Comments

@kuhushukla
Copy link
Collaborator

Describe the bug
SQL queries with UDFs are hard to quantify in terms of what is the actual cost of them falling back to the CPU. Recently some of the customers reported Q tool speedup values of 3x while the actual runtimes were 2x slower.

Expected behavior
We should mark the whole stage as a CPU fallback and penalize UDF SQL queries heavily to indicate them as bad migration candidates.

@kuhushukla kuhushukla added bug Something isn't working ? - Needs Triage labels Jan 8, 2024
@amahussein
Copy link
Collaborator

@nartal1

FWIW, Addition context:

#1271

@amahussein amahussein changed the title [BUG] Qualification: Jobs with UDFs can have considerably inaccurate recommended speed up time. [BUG] Qualification: Jobs with UDFs may have inaccurate speedups Jan 19, 2024
@amahussein amahussein added the core_tools Scope the core module (scala) label Jan 23, 2024
@amahussein
Copy link
Collaborator

@nartal1 is this still an issue or should it be closed with a comment of your final findings?

@nartal1
Copy link
Collaborator

nartal1 commented Feb 9, 2024

Adding penalty for stages with UDF's is done and UDF's are detected correctly in the qualification tool. But it didn't improve the total speedup factor as we do the average of speedups across all stages. I would still keep it open but lower the priority of it to investigate if other improvements can be done.

@amahussein
Copy link
Collaborator

Closing this as speedup has shifted to the QualX now

@amahussein amahussein closed this as not planned Won't fix, can't repro, duplicate, stale Aug 9, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working core_tools Scope the core module (scala)
Projects
None yet
Development

No branches or pull requests

3 participants