ianko / nebulex_redis_adapter

Nebulex adapter for Redis

NebulexRedisAdapter

Nebulex adapter for Redis (including Redis Cluster support).

This adapter uses Redix; a Redis driver for Elixir.

The adapter supports different configurations modes which are explained in the next sections.

See also online documentation and Redis cache example.

Installation

Add nebulex_redis_adapter to your list of dependencies in mix.exs:

defp deps do
  [
    {:nebulex_redis_adapter, "~> 2.1"},
    {:crc, "~> 0.10"},    #=> Needed when using Redis Cluster
    {:jchash, "~> 0.1.2"} #=> Needed when using consistent-hashing
  ]
end

In order to give more flexibility and loading only needed dependencies, this adapter makes all its dependencies optional. For example:

• :crc - Required when using the adapter in mode :redis_cluster. See Redis Cluster.
• :jchash - Required if you want to use consistent-hashing when using the adapter in mode :client_side_cluster.

Then run mix deps.get in your shell to fetch the dependencies.

Usage

After installing, we can define our cache to use Redis adapter as follows:

defmodule MyApp.RedisCache do
  use Nebulex.Cache,
    otp_app: :nebulex,
    adapter: NebulexRedisAdapter
end

The rest of Redis configuration is set in our application environment, usually defined in your config/config.exs:

config :my_app, MyApp.RedisCache,
  conn_opts: [
    # Redix options
    host: "127.0.0.1",
    port: 6379
  ]

Since this adapter is implemented by means of Redix, it inherits the same options, including regular Redis options and connection options as well. For more information about the options, please check out NebulexRedisAdapter module and also Redix.

See also Redis cache example.

Distributed Caching

There are different ways to support distributed caching when using NebulexRedisAdapter.

To learn more about the available config options for the different cluster alternatives below, check out the [online documentation][nebulex_redis_adapter].

Redis Cluster

Redis Cluster is a built-in feature in Redis since version 3, and it may be the most convenient and recommendable way to set up Redis in a cluster and have a distributed cache storage out-of-box. This adapter provides the :redis_cluster mode to set up Redis Cluster from the client-side automatically and be able to use it transparently.

First of all, ensure you have Redis Cluster configured and running.

Then we can define our cache which will use Redis Cluster:

defmodule MyApp.RedisClusterCache do
  use Nebulex.Cache,
    otp_app: :nebulex,
    adapter: NebulexRedisAdapter
end

The config:

config :my_app, MyApp.RedisClusterCache,
  # Enable redis_cluster mode
  mode: :redis_cluster,

  # Master nodes. There must be at least one, in order the adapter be able to
  # get the cluster slots and configure the client side automatically.
  # If one fails, the adapter retries with the next in the list.
  master_nodes: [
    [
      host: "127.0.0.1",
      port: 7000
    ],
    [
      url: "redis://127.0.0.1:7001"
    ],
    # Maybe more master nodes ...
  ],

  # Redix options, except `:host` and `:port`; unless we have a cluster
  # of nodes with the same host and/or port, which doesn't make sense.
  conn_opts: [
    # Maybe Redix options
  ]

The pool of connections against the different master nodes is automatically configured by the adapter once it gets the cluster slots info.

This one could be the easiest and recommended way for distributed caching using Redis and NebulexRedisAdapter.

Client-side Cluster based on Sharding

NebulexRedisAdapter also brings with a simple client-side cluster implementation based on Sharding distribution model.

We define our cache normally:

defmodule MyApp.ClusteredCache do
  use Nebulex.Cache,
    otp_app: :nebulex,
    adapter: NebulexRedisAdapter
end

The config:

config :my_app, MyApp.ClusteredCache,
  # Enable client-side cluster mode
  mode: :client_side_cluster,

  # Nodes config (each node has its own options)
  nodes: [
    node1: [
      # Node poll size
      pool_size: 10,

      # Redix options to establish the pool of connections against this node
      conn_opts: [
        host: "127.0.0.1",
        port: 9001
      ]
    ],
    node2: [
      pool_size: 4,
      conn_opts: [
        url: "redis://127.0.0.1:9002"
      ]
    ],
    node3: [
      conn_opts: [
        host: "127.0.0.1",
        port: 9003
      ]
    ]
    # Maybe more ...
  ]

By default, the adapter uses NebulexRedisAdapter.ClientCluster.Keyslot for the keyslot. Besides, if :jchash is defined as dependency, the adapter will use consistent-hashing automatically.

NOTE: It is highly recommended to define the :jchash dependency when using the adapter in :client_side_cluster mode.

However, you can also provide your own implementation by implementing the Nebulex.Adapter.Keyslot and set it into the :keyslot option. For example:

defmodule MyApp.ClusteredCache.Keyslot do
  use Nebulex.Adapter.Keyslot

  @impl true
  def hash_slot(key, range) do
    # your implementation goes here
  end
end

And the config:

config :my_app, MyApp.ClusteredCache,
  # Enable client-side cluster mode
  mode: :client_side_cluster,

  # Provided Keyslot implementation
  keyslot: MyApp.ClusteredCache.Keyslot,

  # Nodes config (each node has its own options)
  nodes: [
    ...
  ]

Using Nebulex.Adapters.Dist

Another simple option is to use the Nebulex.Adapters.Partitioned and set as local cache the NebulexRedisAdapter. The idea here is each Elixir node running the distributed cache (Nebulex.Adapters.Partitioned) will have as local backend or cache a Redis instance (handled by NebulexRedisAdapter).

This example shows how the setup a distributed cache using Nebulex.Adapters.Partitioned and NebulexRedisAdapter:

defmodule MyApp.DistributedCache do
  use Nebulex.Cache,
    otp_app: :nebulex,
    adapter: Nebulex.Adapters.Partitioned,
    primary_storage_adapter: NebulexRedisAdapter
end

Using a Redis Proxy

The other option is to use a proxy, like twemproxy on top of Redis. In this case, the proxy does the distribution work, and from the adparter's side (NebulexRedisAdapter), it would be only configuration. Instead of connect the adapter against the Redis nodes, we connect it against the proxy nodes, this means, in the config, we just setup the pools with the host and port for each proxy.

Using the cache for executing a Redis command or pipeline

Since NebulexRedisAdapter works on top of Redix and provides features like connection pools and "Redis Cluster" support, it may be seen also as a sort of Redis client, but it is meant to be used mainly with the Nebulex cache API. However, Redis API is quite extensive and there are a lot of useful commands we may want to run taking advantage of the NebulexRedisAdapter features. Therefore, the adapter injects two additional/extended functions to the defined cache: command!/3 and pipeline!/3.

iex> MyCache.command!("mylist", ["LPUSH", "mylist", "world"])
1
iex> MyCache.command!("mylist", ["LPUSH", "mylist", "hello"])
2
iex> MyCache.command!("mylist", ["LRANGE", "mylist", "0", "-1"])
["hello", "world"]

iex> cache.pipeline!("mylist", [
...>   ["LPUSH", "mylist", "world"],
...>   ["LPUSH", "mylist", "hello"],
...>   ["LRANGE", "mylist", "0", "-1"]
...> ])
[1, 2, ["hello", "world"]]

NOTE: The key is required when used the adapter in mode :client_side_cluster or :redis_cluster, for :standalone no needed (optional). And the name is in case you are using a dynamic cache and you have to pass the cache name explicitly.

Testing

To run the NebulexRedisAdapter tests you will have to have Redis running locally. NebulexRedisAdapter requires a complex setup for running tests (since it needs a few instances running, for standalone, cluster and Redis Cluster). For this reason, there is a docker-compose.yml file in the repo so that you can use Docker and docker-compose to spin up all the necessary Redis instances with just one command. Make sure you have Docker installed and then just run:

$ docker-compose up

Since NebulexRedisAdapter uses the support modules and shared tests from Nebulex and by default its test folder is not included in the Hex dependency, the following steps are required for running the tests.

First of all, make sure you set the environment variable NEBULEX_PATH to nebulex:

export NEBULEX_PATH=nebulex

Second, make sure you fetch :nebulex dependency directly from GtiHub by running:

mix nbx.setup

Third, fetch deps:

mix deps.get

Finally, you can run the tests:

mix test

Running tests with coverage:

mix coveralls.html

You will find the coverage report within cover/excoveralls.html.

Benchmarks

Benchmarks were added using benchee; to learn more, see the benchmarks directory.

To run the benchmarks:

$ MIX_ENV=test mix run benchmarks/benchmark.exs

Benchmarks use default Redis options (host: "127.0.0.1", port: 6379).

Contributing

Contributions to Nebulex are very welcome and appreciated!

Use the issue tracker for bug reports or feature requests. Open a pull request when you are ready to contribute.

When submitting a pull request you should not update the CHANGELOG.md, and also make sure you test your changes thoroughly, include unit tests alongside new or changed code.

Before to submit a PR it is highly recommended to run mix check and ensure all checks run successfully.

Copyright and License

Copyright (c) 2018, Carlos Bolaños.

NebulexRedisAdapter source code is licensed under the MIT License.