sqlcache

sqlcache is a caching middleware for database/sql that enables existing Go programs to add caching in a declarative way. It leverages APIs provided by the handy sqlmw project and is inspired from slonik-interceptor-query-cache.

This liberates your Go program from maintaining imperative code that repeatedly implements the cache-aside pattern. Your program will perceive the database client/driver as a read-through cache.

Tested with PostgreSQL database with pgx as the underlying driver.

Cache backends supported:

• ristretto (in-memory)
• redis

It's easy to add other caching backends by implementing the cache.Cacher interface.

Usage

Create a backend cache instance and install the interceptor:

import (
	"database/sql"

	redis "github.com/go-redis/redis/v7"
	"github.com/jackc/pgx/v4/stdlib"
	"github.com/prashanthpai/sqlcache"
)

func main() {
	...
	rc := redis.NewUniversalClient(&redis.UniversalOptions{
		Addrs: []string{"127.0.0.1:6379"},
	})

	// create a sqlcache.Interceptor instance with the desired backend
	interceptor, err := sqlcache.NewInterceptor(&sqlcache.Config{
		Cache: sqlcache.NewRedis(rc, "sqc"),
	})
	...

	// wrap pgx driver with cache interceptor and register it
	sql.Register("pgx-with-cache", interceptor.Driver(stdlib.GetDefaultDriver()))

	// open the database using the wrapped driver
	db, err := sql.Open("pgx-with-cache", dsn)
	...

Caching is controlled using cache attributes which are SQL comments starting with @cache- prefix. Only queries with cache attributes are cached.

Cache attributes:

Example query:

rows, err := db.QueryContext(context.TODO(), `
	-- @cache-ttl 30
	-- @cache-max-rows 10
	SELECT name, pages FROM books WHERE pages > $1`, 100)

See example/main.go for a full working example.

References

• A declarative way to cache PostgreSQL queries using Node.js: a blog post by the author of Slonik.
• Declarative Caching with Postgres and Redis: Kyle Davis's talk on Slonik + Redis.