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

Rest of TPCH #459

Merged
merged 4 commits into from
Jul 16, 2024
Merged

Rest of TPCH #459

merged 4 commits into from
Jul 16, 2024

Conversation

gatesn
Copy link
Contributor

@gatesn gatesn commented Jul 15, 2024

No description provided.

@gatesn gatesn requested a review from a10y July 15, 2024 17:11
@a10y
Copy link
Contributor

a10y commented Jul 16, 2024

image

Added some coloration: cells will have yellow background if they're 10-50% slower than arrow, red background if they are >50% slower, green otherwise

@a10y a10y force-pushed the ngates/tpch-benchmarks branch from 1f5c517 to 9c1f7b1 Compare July 16, 2024 08:49
@gatesn gatesn enabled auto-merge (squash) July 16, 2024 08:56
@gatesn gatesn merged commit d2ebfa2 into develop Jul 16, 2024
2 checks passed
@gatesn gatesn deleted the ngates/tpch-benchmarks branch July 16, 2024 08:57
@a10y
Copy link
Contributor

a10y commented Jul 16, 2024

For comparison, here's what it looks like if you run all of them single-threaded

image

It seems for Q2, Q11, Q19, doing single-threaded VS multi-threaded makes a large difference

gatesn pushed a commit that referenced this pull request Jul 16, 2024
In #459 , we used an 8-thread pool and a single tokio runtime to run all
of the queries in parallel.

There was huge variability on the values returned by those benchmarks.
This PR

* Switches from using tokio global runtime for parallelism to using a
Rayon pool
* Each query executes in a current-thread tokio runtime, which should be
a bit more apples-to-apples
* Removed CSV measurements, and now have each query+format pair do 3
warmup runs, followed by 20 measured runs, then take the min over
measurements
* Fixed the coloration logic for table (oops)
* A few small drive-bys
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants