Stockpile is a simple cache written in Ruby backed by Redis. It has built in cache-stampede (also known as dog-piling) protection and support for multiple Redis servers.
Can be used with any Ruby or Ruby on Rails project. Can be used as a replacement for existing Ruby on Rails cache.
Intended as a heavy usage cache to prevent concurrent execution of code when cache is expired that will lead to congestion collapse of your systems.
Upon caching serializes cached value using Oj gem. While reading value from cache will deserialize value from cache using same gem.
When perform_cached
method is invoked with a key and a block of code as
arguments Stockpile will attempt to fetch value from cache using given key. If
no value is returned it will set a lock deferring all other requests for given
key (for specified amount of time) and run provided block of code and storing
it's return value at the key. After that a lock will be released allowing other
requests to fetch their values from cache.
In case there is a cache miss and an active execution lock for a given key is
present request will go into slumber for 2 seconds (configurable by
STOCKPILE_SLUMBER
environment variable or by calling slumber
method on
configuration object). During slumber request will keep trying to read value
from cache and if no result is returned during that time cache will be bypassed
and value will be computed by executing passed in block.
Add the following line to your Gemfile:
gem 'stockpile_cache'
And run bundle
from your shell.
To install gem manually run from your shell:
gem install stockpile_cache
Only requirement to run this gem is Redis. Other than that it is not dependant on any other framework or system.
The only thing you need to set up is URL of your Redis server. You can do this
by either setting STOCKPILE_REDIS_URL
environment variable or by executing
following code during runtime. For Ruby on Rails create
config/initializers/stockpile.rb
file and put the following code in there:
Stockpile.configure do |configuration|
configuration.redis_url = <REDIS_URL>
end
There are two ways to configure Stockpile: using environment variables or invoking configuration block during runtime.
Following settings are supported:
Variable | Method | Settings |
---|---|---|
STOCKPILE_CONNECTION_POOL |
connection_pool |
Redis connection pool size to share amongst the fibers or threads in your Ruby. Defaults to 100 . |
STOCKPILE_CONNECTION_TIMEOUT |
connection_timeout |
How long to wait for a connection from connection pool to become available (in seconds). Defaults to 3 . |
STOCKPILE_LOCK_EXPIRATION |
lock_expiration |
Time to keep execution lock alive (in seonds). Defaults to 10 . |
STOCKPILE_REDIS_URL |
redis_url |
URL of your Redis server that will be used for caching. Defaults to redis://localhost:6379/1 . |
STOCKPILE_REDIS_SENTINELS |
sentinels |
(optional) Comma separated list of Sentinels IPs for Redis. Defaults to nil . Example value: 8.8.8.8:42,8.8.4.4:42 . |
STOCKPILE_SLUMBER |
slumber |
Timeout (in seconds) for stampede protection lock. After timeout passed in code will be executed instead of reading a value from cache. Defaults to 2 . |
STOCKPILE_CONFIGURATION_FILE |
configuration_file |
(optional) .yml configuration file to read connection information from. See Multiple Database. |
To use simply wrap your code into perform_cached
block:
Stockpile.perform_cached(key: 'meaning_of_life', ttl: 42) do
21 + 21
end
perform
method accepts 4 named arguments:
Argument | Meaning |
---|---|
key |
Pointer in cache by which a value will be either looked up or stored in cache once code provided in block is executed. |
ttl |
(optional) Time in seconds for which a cached value will be stored. Defaults to 300 seconds (5 minutes). |
db |
(optional) Name of the Redis database to cache value in. Defaults to :default |
&block |
Block of code to execute; it's return value will be stored in cache. |
To expire your cache immediately run:
Stockpile.expire_cached(key: 'meaning_of_life')
Stockpile comes with a support for multiple databases. A word of caution: unless you have very good reason to run multiple databases within single instance of Redis server you probably should avoid doing so as you will not see any performance improvements in doing so.
To allow multi-database support you have to do two things. First you have to set
configuration_file
setting to point at .yml
containing your configuration.
You can do so by either setting a STOCKPILE_CONFIGURATION_FILE
environment
variable or by executing a configuration block during runtime (for Rails create
config/initializers/stockpile.rb
with following content):
Stockpile.configure do |configuration|
configuration.configuration_file = <PATH/TO/FILE>
end
Second thing to do is to create a .yml
configuration file. It has to have at
least one database definition. Providing sentinels
is optional. Everything
else is mandatory:
---
master:
url: 'redis://redis-1-host:6379/1'
sentinels: '8.8.8.8:42,8.8.4.4:42'
pool_options:
size: 5
timeout: 5
commander:
url: 'redis://redis-2-host:6379/1'
pool_options:
size: 5
timeout: 5
To query different databases provide a corresponding db:
param with
perform_cached
method:
Stockpile.perform_cached(db: :master, key: 'meaning_of_life', ttl: 42) do
21 + 21
end
Stockpile.perform_cached(db: :commander, key: 'meaning_of_life', ttl: 21) do
21
end
If you do not provide a db:
param then a :default
database will be used; if
you do not define it in a configuration file your request will error out.
Using configuration_file
setting will make Stockpile ignore all other
Redis connection related settings and it will read configuration from .yml
file instead.
Stockpile optionally supports compression of cached content; you will not see
much benefit from compressing small strings but once you start caching bigger
payloads like fragments of HTML you could see some improvements by using
compression. To use compression you will have to use configuration file set by
STOCKPILE_CONFIGURATION_FILE
.
To enable compression you have to do two things. First you have to set
configuration_file
setting to point at .yml
containing your configuration.
You can do so by either setting a STOCKPILE_CONFIGURATION_FILE
environment
variable or by executing a configuration block during runtime (for Rails create
config/initializers/stockpile.rb
with following content):
Stockpile.configure do |configuration|
configuration.configuration_file = <PATH/TO/FILE>
end
Second thing to do is to create a .yml
configuration file. It has to have at
least one database definition. Providing sentinels
and compression
is
optional. Everything else is mandatory:
---
master:
url: 'redis://redis-1-host:6379/1'
sentinels: '8.8.8.8:42,8.8.4.4:42'
compression: true
pool_options:
size: 5
timeout: 5
From that point everything that will be cached in master
database will be
compressed.
There is no timeout or rescue set for code you will be running through the cache. If you need to do either you have to handle it outside of Stockpile.
Locks are never set indefinitely and by default will expire after 10 seconds
allowing next request to trigger cache recalculation. Lock duration is
configurable by either setting STOCKPILE_LOCK_EXPIRATION
environment variable
or by calling slumber
method on configuration object.
While there is an active lock for the key each request trying to read that key
will wait in slumber for 2 seconds (configurable by STOCKPILE_SLUMBER
environment variable or by calling slumber
method on configuration object) and
will bypass cache after that if no value will be set in that time.
After checking out the repo, run bin/setup
to install dependencies. Then, run
rspec
to run the tests. You can also run bin/console
for an interactive
prompt that will allow you to experiment.
To install this gem onto your local machine, run bundle exec rake install
. To
release a new version, update the version number in version.rb
, and then run
bundle exec rake release
, which will create a git tag for the version, push
git commits and tags, and push the .gem
file to
rubygems.org.
Bug reports and pull requests are welcome on GitHub at https://github.com/ConvertKit/stockpile_cache. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the Contributor Covenant code of conduct.
The gem is available as open source under the terms of the [Apache License Version 2.0] (http://www.apache.org/licenses/LICENSE-2.0).
Everyone interacting in the Stockpile project’s codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.