Redisk

Redisk is based off of antirez's weekend idea of a redis ugly brother - BigDis (check out antirez' repos). Redisk uses Redis's protocol, and saves the data via a sha1 hash to disk.

Usage

budle exec bin/redisk

The default options are

• port 6380
• localhost
• 2 level hashing
• /tmp DB prefix

You can change these arguments via a config file, or passing options in the CLI. Check out lib/redisk/config.rb for what options are valid. Passing a CLI option will override a config file that is passed as well.

Example

• open up one terminal and run the (eventmachine) server:

shair@comp ~/projects/redisk: $ bundle exec bin/redisk

Starting Redisk on port 6380

• Then use the redisk server to store files on disk:

shair@comp ~/projects/redisk: $ echo "set foo bar" | redis-cli -p 6380

OK

shair@comp ~/projects/redisk: $ cat /tmp/0b/ee/0beec7b5ea3f0fdbc95d0dd47f3c5bc275da8a33

bar

shair@comp ~/projects/redisk: $ echo "get foo" | redis-cli -p 6380

bar

shair@comp ~/projects/redisk: $ echo "exists foo" | redis-cli -p 6380

(integer) 1

shair@comp ~/projects/redisk: $ echo "del foo" | redis-cli -p 6380

(integer) 1

shair@comp ~/projects/redisk: $ echo "del foo" | redis-cli -p 6380

(integer) 0

shair@comp ~/projects/redisk: $ cat /tmp/0b/ee/0beec7b5ea3f0fdbc95d0dd47f3c5bc275da8a33

cat: /tmp/0b/ee/0beec7b5ea3f0fdbc95d0dd47f3c5bc275da8a33: No such file or directory

shair@comp ~/projects/redisk: $ echo "exists foo" | redis-cli -p 6380

(integer) 0

shair@comp ~/projects/redisk: $ echo "flushdb" | redis-cli -p 6380

OK

Serving files via HTTP

Nginx is known for it's blazing speed for serving static files:

See: http://wiki.nginx.org/NginxHttpRedis

It can be used with simple HTTP req's like GET /key etc.

The content type can be determined with another key.contenttype and be used as the Content Type header.

Nothing here

Any questions?

Enjoy!