erlang的热更新

erlang作为一个为电信级别而出现的语言，热更新是其最重要的特性之一

 热代码升级-Erlang允许程序代码在运行系统中被修改。旧代码能被逐步淘汰而后被新代码替换。在此过渡期间，新旧代码是共存的。

下面我们以最典型的gen_server为例子，讲解一下这个BT的功能

 -module(tt13).
 -behaviour(gen_server).

 -export([test/0]).
 -export([start_link/0, stop/0, init/1, handle_call/3, handle_cast/2, handle_info/2, terminate/2, code_change/3]).

 -record(state, {cnt}).

 -define(SERVER, ?MODULE).

 %%--------------------------------------------------------------------
 %% Function: start_link() -> {ok,Pid} | ignore | {error,Error}
 %% Description: Starts the server
 %%--------------------------------------------------------------------
 start_link() ->
     gen_server:start_link({local, ?MODULE}, ?MODULE, [], [{debug, [trace]}]).

 test() ->
     gen_server:call(?SERVER, test). 

 stop() ->
     gen_server:cast(?SERVER, stop). 

 %%--------------------------------------------------------------------
 %% Function: init(Args) -> {ok, State} |
 %%                         {ok, State, Timeout} |
 %%                         ignore               |
 %%                         {stop, Reason}
 %% Description: Initiates the server
 %%--------------------------------------------------------------------
 init(_) -> {ok, #state{cnt=1}}.
 %%--------------------------------------------------------------------
 %% Function: %% handle_call(Request, From, State) -> {reply, Reply, State} |
 %%                                      {reply, Reply, State, Timeout} |
 %%                                      {noreply, State} |
 %%                                      {noreply, State, Timeout} |
 %%                                      {stop, Reason, Reply, State} |
 %%                                      {stop, Reason, State}
 %% Description: Handling call messages
 handle_call(test, _From, #state{cnt=Cnt} = State) ->
     {reply, {ok, Cnt}, State#state{cnt=Cnt+1}};

 handle_call(stop, _From, State) ->
     {stop, normal, ok, State};

 handle_call(_Unrec, _From, State) ->
     {reply, {error, invalid_call}, State}.

 %%--------------------------------------------------------------------
 %% Function: handle_cast(Msg, State) -> {noreply, State} |
 %%                                      {noreply, State, Timeout} |
 %%                                      {stop, Reason, State}
 %% Description: Handling cast messages
 %%--------------------------------------------------------------------
 handle_cast(_Msg, State) ->
     {noreply, State}.

 %%--------------------------------------------------------------------
 %% Function: handle_info(Info, State) -> {noreply, State} |
 %%                                       {noreply, State, Timeout} |
 %%                                       {stop, Reason, State}
 %% Description: Handling all non call/cast messages
 %%--------------------------------------------------------------------

 handle_info(_Info, State) ->
     {noreply, State}.

 %%--------------------------------------------------------------------
 %% Function: terminate(Reason, State) -> void()
 %% Description: This function is called by a gen_server when it is about to
 %% terminate. It should be the opposite of Module:init/1 and do any necessary
 %% cleaning up. When it returns, the gen_server terminates with Reason.
 %% The return value is ignored.
 %%--------------------------------------------------------------------

 terminate(_Reason, _State) ->
     io:format("hello gen server: terminating~n").

 %%--------------------------------------------------------------------
 %% Func: code_change(OldVsn, State, Extra) -> {ok, NewState}
 %% Description: Convert process state when code is changed
 %%--------------------------------------------------------------------

 code_change(_OldVsn, State, _Extra) ->
     {ok, State}.

 %%====================================================================
 %%other fun
 %%====================================================================

 编译运行结果

 1> c(tt13).
 {ok,tt13}
 2> tt13:start_link().
 {ok,<0.39.0>}
 3> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,1} to <0.32.0>, new state {state,2}
 {ok,1}
 4> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,2} to <0.32.0>, new state {state,3}
 {ok,2}
 5> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,3} to <0.32.0>, new state {state,4}
 {ok,3}

如果修改了函数，可以直接运行

 -module(tt13).
 -version("1.1").
 -behaviour(gen_server).

 %...........
 %...........省略若干行
 %................

 handle_call(test, _From, #state{cnt=Cnt} = State) ->
     {reply, {ok, Cnt}, State#state{cnt=Cnt*2}};

 %...........
 %...........省略若干行
 %................

　可以看到我们修改了计数的方法，而且修改了版本号，然后我们继续运行

 6> c(tt13).          　　　　　　　　　　　　　　　　　　　　　　　　　　%编译新的代码
 {ok,tt13}
 7> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,4} to <0.32.0>, new state {state,8}
 {ok,4}
 8> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,8} to <0.32.0>, new state {state,16}
 {ok,8}
 9> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,16} to <0.32.0>, new state {state,32}
 {ok,16}

　可以看到代码就直接替换了，注意编译的时候会用新的代码替换下一次运行的结果，正在运行还是old code，所以不要编译多次（一般在测试环境先进行热更新测试）。

   如果要替换init/1里面的代码？这个方法肯定是不行的，因为init/1代码只运行一次,比如我要修改state结构体，那要怎么弄呢

 -module(tt13).
 -version("2.0").
 -behaviour(gen_server).

 -record(state, {testcheck, cnt}).
 %...........
 %...........省略若干行
 %................

 init(_) -> {ok, #state{testcheck='chk', cnt=1}}.
 %%--------------------------------------------------------------------
 %% Function: %% handle_call(Request, From, State) -> {reply, Reply, State} |
 %%                                      {reply, Reply, State, Timeout} |
 %%                                      {noreply, State} |
 %%                                      {noreply, State, Timeout} |
 %%                                      {stop, Reason, Reply, State} |
 %%                                      {stop, Reason, State}
 %% Description: Handling call messages
 handle_call(test, _From, #state{cnt=Cnt} = State) ->
     {reply, {ok, Cnt}, State#state{cnt=Cnt+1}};

 %...........
 %...........省略若干行
 %................

26 code_change("1.1", {state, Cnt}, _Extra) ->
27     {ok, {state, chk, Cnt}};

 code_change(_OldVsn, State, _Extra) ->
     {ok, State}.

 %...........
 %...........省略若干行
 %................

　　我们修改了state结构体，修改了init/1函数，而且重写了code_change/3，下面我们运行如下

 1 10> compile:file(tt13).　　　　　　　　　　/*编译代码
 2 {ok,tt13}
 3 11> sys:suspend(tt13).　　　　　　　　　　 /*暂停服务
 4 ok
 5 12> code:purge(tt13).　　　　　　　　　　　/*清除旧的代码，如果有运行的化
 6 false
 7 13> code:load_file(tt13).　　　　　　　　/*加载新的代码
 8 {module,tt13}
 9 14> sys:change_code(tt13,tt13,"1.1",[]).　　　　/*修改state状态
10 ok
11 15> sys:resume(tt13).　　　　　　　　　　　　　　　　/*恢复服务
12 ok
 16> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,32} to <0.32.0>, new state {state,chk,64}
 {ok,32}
 17> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,64} to <0.32.0>, new state {state,chk,128}
 {ok,64}
 18> tt13:test().
 *DBG* tt13 got call test from <0.32.0>
 *DBG* tt13 sent {ok,128} to <0.32.0>, new state {state,chk,256}
 {ok,128}

　　整个替换过程是10-15步，注意这个过程中的tt13服务是hang住的，如果这时候使用服务，会出现timeout，所以一般这5步都是同时执行。

       后面可以看到state状态已经改变