Write Through and Read Through caching library inspired by CacheMoney and cache_fu, support ActiveRecord 4 and 5.
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.



Gem Version Dependency Status Build Status Code Climate

SecondLevelCache is a write-through and read-through caching library inspired by Cache Money and cache_fu, support ActiveRecord 4.

Read-Through: Queries by ID, like current_user.articles.find(params[:id]), will first look in cache store and then look in the database for the results of that query. If there is a cache miss, it will populate the cache.

Write-Through: As objects are created, updated, and deleted, all of the caches are automatically kept up-to-date and coherent.


In your gem file:

ActiveRecord 5.2:

gem 'second_level_cache', '~> 2.4.0'

ActiveRecord 5.0.x, 5.1.x:

gem 'second_level_cache', '~> 2.3.0'

For ActiveRecord 4:

gem "second_level_cache", "~> 2.1.9"

For ActiveRecord 3:

gem "second_level_cache", "~> 1.6"


For example, cache User objects:

class User < ActiveRecord::Base
  second_level_cache expires_in: 1.week

Then it will fetch cached object in this situations:

User.where(status: 1).find(1)
User.where(id: 1).first # or .last

Cache key:

user = User.find(1)
user.second_level_cache_key  # We will get the key looks like "slc/user/1/0"

Expires cache:

user = User.find(1)

or expires cache using class method:


Disable SecondLevelCache:

User.without_second_level_cache do
  user = User.find(1)
  # ...

Only SELECT * query will be cached:

# this query will NOT be cached
User.select("id, name").find(1)


  • SecondLevelCache cache by model name and id, so only find_one query will work.
  • Only equal conditions query WILL get cache; and SQL string query like User.where("name = 'Hooopo'").find(1) WILL NOT work.
  • SecondLevelCache sync cache after transaction commit:
# user and account's write_second_level_cache operation will invoke after the logger.
ActiveRecord::Base.transaction do
  Rails.logger.info "info"
end # <- Cache write

# if you want to do something after user and account's write_second_level_cache operation, do this way:
ActiveRecord::Base.transaction do
end # <- Cache write
Rails.logger.info "info"
  • If you are using SecondLevelCache with database_cleaner, you should set cleaning strategy to :truncation:
DatabaseCleaner.strategy = :truncation


In production env, we recommend to use Dalli as Rails cache store.

config.cache_store = [:dalli_store, APP_CONFIG["memcached_host"], { namespace: "ns", compress: true }]


  • When you want to clear only second level cache apart from other cache for example fragment cache in cache store, you can only change the cache_key_prefix:
SecondLevelCache.configure.cache_key_prefix = "slc1"
  • SecondLevelCache was added model schema digest as cache version, this means when you add/remove/change columns, the caches of this Model will expires.
  • When your want change the model cache version by manualy, just add the version option like this:
class User < ActiveRecord::Base
  second_level_cache version: 2, expires_in: 1.week
  • It provides a great feature, not hits db when fetching record via unique key (not primary key).
# this will fetch from cache
user = User.fetch_by_uniq_keys(nick_name: "hooopo")
post = Post.fetch_by_uniq_keys(user_id: 2, slug: "foo")

# this also fetch from cache
user = User.fetch_by_uniq_keys!(nick_name: "hooopo") # this will raise `ActiveRecord::RecordNotFound` Exception when nick name not exists.
  • You can use Rails's Eager Loading feature as normal. Even better, second_level_cache will transform the IN query into a Rails.cache.multi_read operation. For example:
Answer.includes(:question).limit(10).order("id DESC").each{|answer| answer.question.title}
Answer Load (0.2ms)  SELECT `answers`.* FROM `answers` ORDER BY id DESC LIMIT 10 # Only one SQL query and one Rails.cache.read_multi fetching operation.

Details for read_multi feature.

Original design by:


Contributor List


MIT License