对PostgreSQL 后台进程名称的再认知
看 Postmaster 中如下的代码
/* StartChildProcess -- start an auxiliary process for the postmaster * * xlop determines what kind of child will be started. All child types * initially go to AuxiliaryProcessMain, which will handle common setup. * * Return value of StartChildProcess is subprocess' PID, or 0 if failed * to start subprocess. */ static pid_t StartChildProcess(AuxProcType type) { pid_t pid; char *av[10]; int ac = 0; char typebuf[32]; /* * Set up command-line arguments for subprocess */ av[ac++] = "postgres"; #ifdef EXEC_BACKEND av[ac++] = "--forkboot"; av[ac++] = NULL; /* filled in by postmaster_forkexec */ #endif snprintf(typebuf, sizeof(typebuf), "-x%d", type); av[ac++] = typebuf; av[ac] = NULL; Assert(ac < lengthof(av)); #ifdef EXEC_BACKEND pid = postmaster_forkexec(ac, av); #else /* !EXEC_BACKEND */ pid = fork_process(); if (pid == 0) /* child */ { IsUnderPostmaster = true; /* we are a postmaster subprocess now */ /* Close the postmaster's sockets */ ClosePostmasterPorts(false); /* Lose the postmaster's on-exit routines and port connections */ on_exit_reset(); /* Release postmaster's working memory context */ MemoryContextSwitchTo(TopMemoryContext); MemoryContextDelete(PostmasterContext); PostmasterContext = NULL; AuxiliaryProcessMain(ac, av); ExitPostmaster(0); } #endif /* EXEC_BACKEND */ …… /* * in parent, successful fork */ return pid; }
AuxiliaryProcessMain 是各个子进程的入口点。
那么为何 ps -ef | grep post 可以看到各个子进程的不同的名字呢：
这和 AuxiliaryProcessMain的实现有关， 其代码在 bootstrap.c 中，下面缩略一下，只考虑我关心的代码：
/* * AuxiliaryProcessMain * * The main entry point for auxiliary processes, such as the bgwriter, * walwriter, walreceiver, bootstrapper and the shared memory checker code. * * This code is here just because of historical reasons. */ void AuxiliaryProcessMain(int argc, char *argv[]) { …… /* If no -x argument, we are a CheckerProcess */ MyAuxProcType = CheckerProcess; while ((flag = getopt(argc, argv, "B:c:d:D:Fr:x:-:")) != -1) { switch (flag) { …… case 'x': MyAuxProcType = atoi(optarg); break; …… } } …… /* * Identify myself via ps */ if (IsUnderPostmaster) { const char *statmsg; switch (MyAuxProcType) { case StartupProcess: statmsg = "startup process"; break; case BgWriterProcess: statmsg = "writer process"; break; case CheckpointerProcess: statmsg = "checkpointer process"; break; case WalWriterProcess: statmsg = "wal writer process"; break; case WalReceiverProcess: statmsg = "wal receiver process"; break; default: statmsg = "??? process"; break; } init_ps_display(statmsg, "", "", ""); } …… /* * XLOG operations */ SetProcessingMode(NormalProcessing); switch (MyAuxProcType) { case CheckerProcess: /* don't set signals, they're useless here */ CheckerModeMain(); proc_exit(1); /* should never return */ case BootstrapProcess: bootstrap_signals(); BootStrapXLOG(); BootstrapModeMain(); proc_exit(1); /* should never return */ case StartupProcess: /* don't set signals, startup process has its own agenda */ StartupProcessMain(); proc_exit(1); /* should never return */ case BgWriterProcess: /* don't set signals, bgwriter has its own agenda */ BackgroundWriterMain(); proc_exit(1); /* should never return */ case CheckpointerProcess: /* don't set signals, checkpointer has its own agenda */ CheckpointerMain(); proc_exit(1); /* should never return */ case WalWriterProcess: /* don't set signals, walwriter has its own agenda */ InitXLOGAccess(); WalWriterMain(); proc_exit(1); /* should never return */ case WalReceiverProcess: /* don't set signals, walreceiver has its own agenda */ WalReceiverMain(); proc_exit(1); /* should never return */ default: elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType); proc_exit(1); } }
可以看到，会根据调用的时候，所传递的flag，来设定各子进程的 显示名称：
主要是这一段：
switch (MyAuxProcType)
{
case StartupProcess:
statmsg = "startup process";
break;
case BgWriterProcess:
statmsg = "writer process";
break;
case CheckpointerProcess:
statmsg = "checkpointer process";
break;
case WalWriterProcess:
statmsg = "wal writer process";
break;
case WalReceiverProcess:
statmsg = "wal receiver process";
break;
default:
statmsg = "??? process";
break;
}
init_ps_display(statmsg, "", "", "");
其中 ，init_ps_display 起到了关键的作用。但是具体在合适才导致 ps 时可以看到各个不同的名称，还有待进一步的观察。
可以先作一个实验来看：
在bootstrap 的 init_ps_dispalya(statmsg,"","","")； 完毕后，加一句：
//added by gaojian , sleep 3 minutes
fprintf(stderr,"sleeping...\n");
sleep(180);
让它停顿三秒种，此时用ps 命令看， 然后，三秒过后，再用ps 命令看，看到的结果如下：
启动 Postgres ，看到 sleeeping...信息后，立即执行如下命令：
[root@localhost ~]# ps -ef|grep post
root 2928 2906 0 13:42 pts/1 00:00:00 grep post
[root@localhost ~]# ps -ef|grep post
root 2953 2932 0 13:42 pts/2 00:00:00 su - postgres
postgres 2954 2953 0 13:42 pts/2 00:00:00 -bash
postgres 2989 2954 0 13:42 pts/2 00:00:00 /usr/local/pgsql/bin/postgres -D /usr/local/pgsql/data
postgres 2990 2989 0 13:42 pts/2 00:00:00 postgres: startup process
root 2992 2906 0 13:42 pts/1 00:00:00 grep post
可以看到有一个叫 postgres: startup process 的子进程出现。后面这个进程不见了。这是很有趣的。
然后等过了几秒钟，再用ps 命令来看：
[root@localhost ~]# ps -ef|grep post
root 2953 2932 0 13:42 pts/2 00:00:00 su - postgres
postgres 2954 2953 0 13:42 pts/2 00:00:00 -bash
postgres 2989 2954 0 13:42 pts/2 00:00:00 /usr/local/pgsql/bin/postgres -D /usr/local/pgsql/data
postgres 3000 2989 0 13:45 ? 00:00:00 postgres: checkpointer process
postgres 3001 2989 0 13:45 ? 00:00:00 postgres: writer process
postgres 3002 2989 0 13:45 ? 00:00:00 postgres: wal writer process
postgres 3003 2989 0 13:45 ? 00:00:00 postgres: autovacuum launcher process
postgres 3004 2989 0 13:45 ? 00:00:00 postgres: stats collector process
root 3018 2906 0 13:53 pts/1 00:00:00 grep post
--------------------------------------------------------------------------------------------------
对此暂时先放一放。我想，更重要的是如下这样的代码：
switch (MyAuxProcType)
{
......
case BgWriterProcess:
/* don't set signals, bgwriter has its own agenda */
BackgroundWriterMain();
proc_exit(1);
......
}
可以打个比方： 一个父亲有几个儿子，有一天他问几个儿子，你们长大想做什么呢？
老大说，我想当将军，保家卫国。老二说，我想当医生，救死扶伤。 老三说，我想当教师，桃李满天下。
于是父亲说，好啊，他给每个儿子作了一定帽子，写着 将军、医生、教师。在人们的嘲笑声中，三个儿子一个去了部队，一个去了医学院，一个去了师范学院。走上了各自不同的道路，后来真的成就了各自的梦想。
这就是不同的子进程拥有不同的名称的类比了。
