-
Notifications
You must be signed in to change notification settings - Fork 19
added broadcast join operator #229
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
added broadcast join operator #229
Conversation
2b0f8ad to
1a31b0d
Compare
b2292dd to
718808b
Compare
|
In CI, test tpcd queries 10, 35, and 69 seemed to fail but pass repetitively locally. Are these known as flaky tests in CI? |
| if !partition_group.contains(&i) { | ||
| style = "[style=invis]"; | ||
| } | ||
| } else if let Some(partition_group) = &isolator_partition_group |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
clippy wouldn't leave me alone 😄
| if let Some(stage) = plan.input_stage() { | ||
| result.push(stage); | ||
| } | ||
| if let Some(plan) = plan.as_network_boundary() |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
same here
They are flaky yes #267, although I the failures I see in this PR look different. I think the issue is that sometimes the queries in those tests get distributed and sometimes not, as the number of files produced by the tpc-ds generator is random |
|
With the current setup, almost no tpc-ds query is actually getting distributed. I actually needed to rework those tests in #264. It looks like at random points they do get distributed though |
Draft for broadcast join
Closes #223
Benches are volatile, doesn't seem like drastic improvements due to tpch data
Follow up could be some more benching