-
Notifications
You must be signed in to change notification settings - Fork 26
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
feat(core/df-repr): better upper bound pruning to reduce plan space #265
Conversation
skyzh
commented
Dec 14, 2024
•
edited
Loading
edited
- Remove cross join from the repr
- Implement upper bound pruning correctly
- Two-stage cascades, also use the new task graph instead of refactoring the current one
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
b3288bc
to
7fef116
Compare
Signed-off-by: Alex Chi Z <iskyzh@gmail.com> tune plan space budget Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
7fef116
to
b3cff3c
Compare
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
b3cff3c
to
fc187cf
Compare
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
f225728
to
cac5d3f
Compare
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
bf25604
to
7f7c84d
Compare
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
7f7c84d
to
5325531
Compare
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com> tune plan space budget Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
5325531
to
e5f7c50
Compare
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
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.
Reviewed:
- cross join to inner join changes
- main upper bound pruning logic in
optimize_inputs_inner
I need another look at task2.rs
and potentially doc more aggressively like #234. I actually like this version of the task graph a lot (no state machine logic).
Is there any other part of the logic I need pay special attention to?
I think part of this PR is tuning to ensure the plan is what we want (i.e., filter selectivity and behavior on reaching the budget). The tasks2.rs logic is also something that we should fully understand as we are going to bring more logic into it (phys prop, etc.) |
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.
Needs to doc the task graph a little bit more but LGTM
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com> tune plan space budget Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
e5f7c50
to
023d47e
Compare
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com> tune plan space budget Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>
Signed-off-by: Alex Chi Z <iskyzh@gmail.com>