异步gen_server进行port访问时性能严重下降的原因和应对方法...
在进行项目开发时，有这么一个需求：一个生产者不断产生消息，并将消息异步投递给消费者，异步投递不受任何限制，消费者进程内部将消息写入文件或者通过套接字发送该消息。看到这个需求，我想也没想，熟练地写下了如下代码：API:start_link(FileName, DsgHost, DstPort) ->
gen_server:start_link({local, ?SERVER}, ?MODULE, , []).
deliver_to_file(Msg) -> deliver_to(file, Msg).
deliver_to_net(Msg) -> deliver_to(net, Msg).
deliver_to(Type, Msg) ->
gen_server:cast(?SERVER, {deliver_to, Type, Msg}).gen_server callback:init() ->
{ok, FD} = file:open(FileName,
),
{ok, Socket} = gen_tcp:connect(DsgHost, DstPort,
),
{ok, #state{file = FD, socket = Socket}}.
handle_cast({deliver_to, Type, Msg}, State) ->
handle_msg(Type, Msg, State),
{noreply, State};
handle_msg(Type, Msg, #state{file = File, socket = Socket}) when is_binary(Msg) ->
case Type of
file -> file:write(File, Msg);
net -> gen_tcp:send(Socket, Msg)
end;
handle_msg(Type, Msg, State) ->
handle_msg(Type, term_to_binary(Msg), State).
压测函数编写如下：deliver_to_with_len_and_times(Type, Len, Times) ->
Msg = makeup_log_by_len(Len),
deliver_to_times(Type, Msg, Times).
deliver_to_times(Type, Msg, Times) when Times > 0 ->
deliver_to(Type, Msg);
deliver_to_times(_Type, _Msg, 0) ->
ok.
makeup_log_by_len(Len) ->
list_to_binary(lists:duplicate(Len, $a)).
函数主要目的为异步地向consumer投递Times个长度为Len的消息。这段看似简单的代码，却在压测的时候产生了意想不到的效果：当投递消息个数较小时（<1000），handle_msg的处理qps上万，但是随着投递消息数目的增加（>10000），性能却急剧下降，最后仅有300+，使用vmstat、ifstat和top工具观察，磁盘写入和网络发送数据量均不大，但是cpu消耗却不小，这是肿么回事呢？
未完待续...
没后续了？
页:
[1]