redis_pool

Synopsis

minimalistic redis client for erlang.

Derived from:

https://github.com/bmizerany/redis-erl/

Active forks:

https://github.com/dialtone/redis_pool

Overview

redis_pool is a redis client implementation that only uses redis's multi bulk command protocol.

It implements connection pools, timeout management, automatic reconnect on errors and a very simple sharding mechanism.

Raw examples

0> redis_pool:start_link().  % connect to default redis port on localhost
{ok,<0.33.0>}
1> redis_pool:expand(10).
ok
2> redis:q(["set", "foo", "bar"]).
{ok,<<"OK">>}
3> redis:q(["get", "foo"]).
{ok,<<"bar">>}
4> redis:q(["sadd", "bar", "a"]).
{ok,1}
5> redis:q(["sadd", "bar", "b"]).
{ok,1}
6> redis:q(["sadd", "bar", "c"]).
{ok,1}
7> redis:q(["lrange", "0", "-1"]).
{error,<<"ERR wrong number of arguments for 'lrange' command">>}
7> redis:q(["lrange", "bar", "0", "-1"]).
{error,<<"ERR Operation against a key holding the wrong kind of value">>}
8> redis:q(["smembers", "bar"]).
[{ok,<<"c">>},{ok,<<"a">>},{ok,<<"b">>}]
9> redis:q(["incr", "counter"]).
{ok,1}

Multiple Pools

1> redis_pool:start_link(one).
{ok,<0.33.0>}
2> redis_pool:start_link(two).
{ok,<0.36.0>}
3> redis_pool:start_link(three).
{ok,<0.39.0>}
4> redis_pool:start_link(four). 
{ok,<0.42.0>}
5> redis_pool:expand(one, 10).
ok
6> redis_pool:expand(two, 10).
ok
7> redis_pool:expand(three, 10).
ok
8> redis_pool:expand(four, 10). 
ok
9> redis_shard:start_link(shard, [one, two, three, four]).
{ok,<0.91.0>}
10> redis:q(redis_shard:pool(shard, "foo"), ["set", "foo", "bar"]).
{ok, <<"OK">>}

Multiple Pools with Manager

1> redis_sup:start_link().
{ok,<0.33.0>}
2> redis_shard:start_link(main, []).
{ok,<0.35.0>}
3> redis_manager:start_link(main, [{one, [10]}, {two, [10]}, {three, [10]}, {four, [10]}]).
{ok,<0.37.0>}
4> redis:q(redis_shard:pool(main, "foo"), ["set", "foo", "bar"]).
{ok,<<"OK">>}

Pub/Sub

%% the standard query function can be used to publish data
redis:q([<<"PUBLISH">>, Key, Value])

  Types:

    Key, Value = binary()

%% the connect method starts a supervised redis child pid with a
%% socket open to the redis server in active mode
redis_subscribe:connect(Key, Opts, Callback) -> {ok, Pid}

  Types:

    Key = binary()
    Opts = list()
    Callback = {M,F,A} | fun() | {fun(), A}
    M = atom()
    F = atom()
    A = list()
    Pid = pid()

1> redis_pool:add_pool(1).
ok
2> redis:q([<<"PUBLISH">>, <<"foo">>, <<"sandwich">>]).
{ok,0}
3> redis_subscribe:connect(<<"foo">>, [], fun(Val) -> io:format("recv'd ~p~n", [Val]) end).
{ok,<0.65.0>}
4> redis:q([<<"PUBLISH">>, <<"foo">>, <<"sandwich">>]).
recv'd {message,<<"foo">>,<<"sandwich">>}
{ok,1}
5> redis_subscribe:connect(<<"bar">>, [], {fun(Arg, Val) -> io:format("recv'd ~p, additional arg: ~p~n", [Val, Arg]) end, [myarg]}).
{ok,<0.68.0>}
6> redis:q([<<"PUBLISH">>, <<"bar">>, <<"birthday cake">>]).
recv'd {message,<<"bar">>,<<"birthday cake">>}, additional arg: myarg
{ok,1}

Breakdance/Breakdown

redis_pool:start_link() -> ok | {error, Reason}
redis_pool:start_link(Name) -> ok | {error, Reason}
redis_pool:start_link(Name, Opts) -> ok | {error, Reason}
redis_pool:start_link(Name, Opts, MaxRestarts, Interval) -> ok | {error, Reason}

  Types:
    
    Name = atom()
    MaxRestarts = Interval = number()
    Options = [Opt]
    Opt = {atom(), Value}
    Value = term()

  Connects to redis server

  The available options are:

  {ip, Ip}

    If the host has several network interfaces, this option
    specifies which one to use.  Default is "127.0.0.1".

  {port, Port}

    Specify which local port number to use.  Default is 6379.

  {pass, Pass}

    Specify to password to auth with upon a successful connect.
    Default is <<>>.

  {db, DbIndex}

    Sepcify the db to use.  Default is 0.

  Name represents the name of the pool you are creating, by default
  it's redis_pool.
  
  MaxRestarts and Interval stop the restart cycle when the
  connections in the pool die too often, where often is defined as
  MaxRestarts in Interval.

redis_pool:expand(NewSize) -> ok
redis_pool:expand(Name, NewSize) -> ok

  Types:
  
    NewSize = number()
    
      Set the connection pool to the given size.
    
    Name = atom()

      Send the call to a specific pool. By default redis_pool.

redis_pool:pool_size() -> number()
redis_pool:pool_size(Name) -> number()

  Types
  
    Name = atom()
  
  Returns the current number of live connections in the pool.

redis_pool:pid() -> pid() | {error, term()}
redis_pool:pid(Name) -> pid() | {error, term()}

  Types
  
    Name = atom()

  Return a connection from the specified pool or from the default one.

redis_pool:cycle(NewOpts) -> ok
redis_pool:cycle(Name, NewOpts) -> ok

  Types
  
    NewOpts = [NewOpt]
    NewOpt = {atom(), Value}
    Value = term()
    Name = atom()

  Recreate every connection in the pool using the provided NewOpts.
  NewOpts follows the same rules as in redis_pool:start_link/4.
  
  Name specifies the name of the connection pool or redis_pool if not
  provided.

redis_pool:info() -> term()
redis_pool:info(Name) -> term()

  Types
  
    Name = atom()

  Returns the current state of the connection pool.

redis_shard:start_link(Servers) -> ok | {error, Reason}
redis_shard:start_link(Name, Servers) -> ok | {error, Reason}

  Types
  
    Servers = [Server]
    Server = term()
    Name = atom()

    Starts a shard manager that splits evenly an md5 interval
    across the given terms in order.
    
    The suggested way of using this is by having each Server be the
    name of a connection pool. But you can also store additional
    information. You'll however need to extract the label to the
    connection pool in order to access it.

redis_shard:pool(Key) -> atom() | {error, Reason}
redis_shard:pool(Name, Key) -> atom() | {error, Reason}

  Types
  
    Key = number() | list()
    Name = atom()
    
    Returns from the given shard the label that contains the
    specified key or index.

redis_shard:update_shards(Servers) -> ok
redis_shard:update_shards(Name, Servers) -> ok

  Types
  
    Servers = [Server]
    Server = term()
    Name = atom()

  Updates the initial configuration and redistributes the shards to
  new servers. No consistency checks are done at this point.

redis:q(Parts) -> {ok, binary()} | {ok, int()} | {error, binary()}
redis:q(Name, Parts) -> {ok, binary()} | {ok, int()} | {error, binary()}
redis:q(Name, Parts, Timeout) -> {ok, binary()} | {ok, int()} | {error, binary()}

  Types

    Parts = [Part]

      The sections, in order, of the command to be executed

    Part = binary() | list()

      A section of a redis command

  Tell redis to execute a command.
  
    Name = atom()
    
      Execute a command in a specific pool of connections
    
    Timeout = number()
    
      Specify a timeout in ms after which the call can be considered
      failed and the connection can be recycled.

LICENSE

Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation
files (the "Software"), to deal in the Software without
restriction, including without limitation the rights to use,
copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following
conditions:

The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
OTHER DEALINGS IN THE SOFTWARE.

jkvor/redis_pool