This is the official repository for Leveraging Applicaton Data Constraints to Optimize Database-Backed Web Applications by Xiaoxuan Liu, Shuxian Wang, Mengzhu Sun, Sicheng Pan, Ge Li, Siddharth Jha, Cong Yan, Junwen Yang, and Alvin Chueng.
Here we present Coco, the first tool that identifies data relationships by analyzing the programs that generate and maintain the persist data. Once identified,Coco leverages the found constraints to optimize the application's physical design and query execution by rewriting queries. Instead of developing a fixed set of predefined rewriting rules, Coco employs an enumerate-test-verify technique to automatically exploit the discovered data constraints to improve query execution.
requirements: we need a linux system to run Coco
- Install nix, the package manager
sh <(curl -L https://nixos.org/nix/install) --daemon
And enable the Nix flake feature which we rely on
mkdir -p ~/.config/nix
echo "experimental-features = nix-command flakes" > ~/.config/nix/nix.conf
- Run the end-to-end pipeline
nix run . -- <DATA> <APP>
DATA is the data directory that includes application source code (data/app_source_code), sqls that define app scheme (data/app_create_sql/), and app queries (data/queries/) to be optimized. In the repo, the DATA should be set to ./data.
APP is the application name, which can be set to redmine, forem, mastodon, openproject, openstreetmap, spree.
| folder | description |
|---|---|
| extractor | Constraint Extraction. |
| rewriter | Rewrite queries based on input constraints. Rewriting queries is performed in the following steps: enumerate rewrites with rules, estimate cost, run tests to filter out incorrect rewrites, collect the queries that pass the test and prepare into inputs for prover, and formally verify the equivalence of query pairs. |
| data | Input and output data. |