英语不好，请见谅！！！！ I use the poolboy as my database connection pools,i have read the README.md on the github:https://github.com/devinus/poolboy But at last i do not konw where i have started the poolboy when i want it to start,then i got an error:already_started
My project's files:http://pastebin.com/zus6dGdz I use the cowboy to be my http server,but you can ignore it.
I start the program like this: 1.I use the rebar to compile $rebar clean & make 2.then i use the erl to run my program $ erl -pa ebin/ -pa deps/*/ebin -s start server_start But i got the errors as follows:
=CRASH REPORT==== 3-Feb-2015::17:47:27 ===
crasher:
initial call: poolboy:init/1
pid: <0.171.0>
registered_name: []
exception exit: {{badmatch,{error,{already_started,<0.173.0>}}},
[{poolboy,new_worker,1,
[{file,"src/poolboy.erl"},{line,260}]},
{poolboy,prepopulate,3,
[{file,"src/poolboy.erl"},{line,281}]},
{poolboy,init,3,[{file,"src/poolboy.erl"},{line,143}]},
{gen_server,init_it,6,
[{file,"gen_server.erl"},{line,306}]},
{proc_lib,init_p_do_apply,3,
[{file,"proc_lib.erl"},{line,237}]}]}
in function gen_server:init_it/6 (gen_server.erl, line 330)
ancestors: [hello_erlang_sup,<0.66.0>]
messages: []
links: [<0.172.0>,<0.173.0>,<0.170.0>]
dictionary: []
trap_exit: true
status: running
heap_size: 610
stack_size: 27
reductions: 205
neighbours:
neighbour: [{pid,<0.173.0>},
{registered_name,db_mongo_handler},
{initial_call,{db_mongo_handler,init,['Argument__1']}},
{current_function,{gen_server,loop,6}},
{ancestors,[<0.172.0>,mg_pool1,hello_erlang_sup,<0.66.0>]},
{messages,[]},
{links,[<0.172.0>,<0.174.0>,<0.171.0>]},
{dictionary,[]},
{trap_exit,false},
{status,waiting},
{heap_size,233},
{stack_size,9},
{reductions,86}]
Please help to solve the problem!Ths!
You are starting a pool of 10 workers with the same registered name. When a process is registered with a name and another process tries to register with the same name, you get the error already_started.
In your example code, the worker module for poolboy is db_mongo_handler. Poolboy tries to start 10 workers by calling db_mongo_handler:start_link/1 which is implemented as
start_link(Args) ->
gen_server:start_link({local, ?SERVER}, ?MODULE, Args, []).
The first worker can start but when the second worker starts it crashes with already_started.
Normally the workers of a pool of many similar workers should not have a registered name. Instead, only the pool has a name and when you need a worker, you ask poolboy to deliver a pid() of one of the workers using poolboy:checkout(mg_pool1).
To fix the code, change gen_server:start_link({local, ?SERVER}, ?MODULE, Args, []) to gen_server:start_link(?MODULE, Args, []). Then it will not be registered with a name.