Process Monitor HOW-TO for Linux Al Dev (Alavoor Vasudevan) alavoor@yahoo.com v11.0, 14 June 2001 This document describes how to monitor Linux/Unix processes and to re- start them automatically if they die without any manual intervention. This document also has URLs for "Unix Processes" FAQs. ______________________________________________________________________ Table of Contents 1. Linux or Unix Processes 2. Unix/Linux command - procautostart 3. File procautostart.cpp 4. File debug.cpp 5. File debug.h 6. Makefile 7. Testing the program - monitor_test 8. Other Monitoring Tools 8.1 Unix init command 8.2 OpenSource Monitoring Tools 8.3 Monitoring Tool - "daemontools" 8.4 Commercial Monitoring Tools 9. Related URLs 10. Other Formats of this Document 10.1 Acrobat PDF format 10.2 Convert Linuxdoc to Docbook format 10.3 Convert to MS WinHelp format 10.4 Reading various formats 11. Copyright Notice ______________________________________________________________________ 1. Linux or Unix Processes Processes are the "heart" of the Linux/Unix processes. It is very important to monitor the application processes to ensure 100% availability and reliability of the computer system. For example, processes of databases, web-server etc.. need to be up and running 24 hours a day and 365 days a year. Use the tools described in this document to the monitor important application processes. See also the following related topics on Linux/Unix processes. · Unix Programming FAQ - Chapter 1 Unix Processes · Other FAQs on Unix are at 2. Unix/Linux command - procautostart Use the program procautostart (say "Prok-Auto-Start" or Process AutoStart) to monitor and automatically re-start any Unix/Linux process if they die. This tiny program is very powerful and is comparable to big commercial products which costs about $80,000US. Procautostart can be used for controlling following applications: · For real-time control of process industries like chemical, manufacturing, power generation and others. Use nano-seconds in program to get fine control. · For controlling processes of software applications like Web servers, database servers, mission critical unix processes, etc.. · As an alarm system for any general monitoring software system. The program can fire a pager or call cell phone or flash red lights on the computer screen. For calling a phone line you may need to use a Telephone card on PCI slot of the computer. The program listing is given in following sections in this document. procautostart -n < delay_seconds > -c "< command_line >" nohup & This starts the unix process procautostart and also command_line process. The procautostart process will re-start command_line process if it dies. The -n option is the time delay in seconds before procautostart checks the running process started by command_line. It is advisable to start the procautostart as background process with no- hangup using "nohup &". See 'man nohup'. The procautostart is written in "C" so that it is very fast and efficient, since the program is called every n seconds. Amount of resources consumed by procautostart is very minute and is negligible since the program size is small and is highly optimized with -o3 compiler option. For example - ______________________________________________________________________ procautostart -n 12 -c "monitor_test -d $HOME -a dummy_arg " nohup & ______________________________________________________________________ Here procautostart will be checking the process monitor_test every 12 seconds. The program will output log files in 'mon' sub-directory which has datetime stamp of when the processes died and re-started. These files gives info on how often the processes are dying. You can also use micro-seconds option '-m' or nano-seconds option '-o', edit the source code file procautostart.cpp and uncomment appropriate lines. 3. File procautostart.cpp // From your browser save this file as text-file named as 'procautostart.cpp'. ______________________________________________________________________ // Author: Al Dev alavoor@yahoo.com // // Program to monitor the unix processes // and automatically re-start them if they die // //************************************************************************ // NOTE: This program uses the Al Dev's String class library. Download string // class from http://linuxdoc.org/HOWTO/C++Programming-HOWTO.html //************************************************************************ #include #include // C strings #include // for getopt #include // for free #include // for kill() - error numbers command extern int errno; #ifdef Linux #include // for kill() - error numbers command #endif #include // for kill() command #include // for kill() command #include // for wait() #include // for setenv() #include // for strftime() #include // for basename() //#include // for logging #include "debug.h" #include "String.h" #include "StringTokenizer.h" #define BUFF_HUN 100 #define BUFF_THOU 1024 #define PR_INIT_VAL -10 #define WAIT_FOR_SYS 5 // wait for process to start up #define DEF_SL_SECS 6 // default sleep time #define SAFE_MEM 10 // to avoid any possible memory leaks #define LOG_NO false // do not output to logfile #define LOG_YES true // do output to logfile #define STD_ERR_NO false // do not print to std err #define STD_ERR_YES true // do print to std err #define DATE_NO false // do not print date #define DATE_YES true // do print date int start_process(char *commandline, char *args[], char **envp, pid_t proc_pid); int fork2(pid_t parent_pid, unsigned long tsecs); inline void error_msg(char *mesg_out, char *lg_file, bool pr_lg, bool std_err, bool pr_dt); ////////////////////////////////////////////// // To test this program use -- // procautostart -n 5 -c 'monitor_test dummy1 -a dummy2 -b dummy3 ' & ////////////////////////////////////////////// void usage(char **argv) { printf("%s:\n", argv[0]); printf("\ninterval specification:\n"); printf(" -n # -- seconds\n"); printf(" -m # -- microseconds\n"); printf(" -o # -- nanoseconds\n"); printf("\nprocess specification:\n"); printf(" -c 'cmdline'\n"); printf (" -p pidfile -- if specified reads process pid from this file\n"); printf("\n"); printf("\nUsage : %s -n -m -o -c ''\n", argv[0]); printf("\nExample: procautostart -n 5 -c 'monitor_test dummy1 -a dummy2 -b dummy3 ' \n"); exit(-1); } int main(int argc, char **argv, char **envp) { unsigned long sleep_sec, sleep_micro, sleep_nano; int ch; pid_t proc_pid; int pr_no = PR_INIT_VAL; char mon_log[40]; char *pr_name = NULL, **cmdargs = NULL; String cmdline; char *pidfile = NULL; // you can turn on debug by editing Makefile and put -DDEBUG_PRT in gcc debug_("test debug", "this line"); debug_("argc", argc); // Use getpid() - man 2 getpid() proc_pid = getpid(); // get the Process ID of procautostart debug_("PID proc_pid", (int) proc_pid); // Create directory to hold log, temp files system("mkdir mon 1>/dev/null 2>/dev/null"); sleep_sec = DEF_SL_SECS ; // default sleep time sleep_micro = 0; // default micro-sleep time sleep_nano = 0; // default nano-sleep time optarg = NULL; while ((ch = getopt(argc, argv, "n:m:o:h:c:")) != -1) // needs trailing colon : { switch (ch) { case 'n': debug_("scanned option n ", optarg); sleep_sec = atoi(optarg); debug_("sleep_sec", sleep_sec); break; case 'm': debug_("scanned option m ", optarg); sleep_micro = atoi(optarg); debug_("sleep_micro", sleep_micro); break; case 'o': debug_("scanned option o ", optarg); sleep_nano = atoi(optarg); debug_("sleep_nano", sleep_nano); break; case 'c': debug_("scanned option c ", optarg); cmdline = optarg; //cmdline = strdup(optarg); // does auto-malloc here debug_("cmdline", cmdline.val()); break; case 'h': debug_("scanned option h ", optarg); usage(argv); break; case 'p': pidfile = strdup(optarg); break; default: debug_("ch", "default"); usage(argv); break; } } if (cmdline.length() == 0) //if (cmdline == NULL) usage(argv); // detach from the main process if (fork()) // 0 returned in child process exit(0); // exit parent process - non-zero child PID got here... //openlog(argv[0], LOG_PID, LOG_DAEMON); // trim the trailing blanks -- otherwise problem in grep command //cmdline.trim(true); //cmdline.chopall('&'); // trim trailing ampersand debug_("cmdline", cmdline.val()); // Start the process { // Find the command line args StringTokenizer strtokens(cmdline.val()); // string tokenizer is borrowed from Java cmdargs = (char **) calloc(strtokens.countTokens() + SAFE_MEM, sizeof(char *)); debug_("countTokens()", strtokens.countTokens()); for (int tmpii = 0; strtokens.hasMoreTokens(); tmpii++) { cmdargs[tmpii] = strdup(strtokens.nextToken().val()); debug_("tmpii", tmpii); debug_("cmdargs[tmpii]", (char *) cmdargs[tmpii]); } // In case execve you MUST NOT have trailing ampersand & in the command line!! //pr_no = start_process(cmdline, NULL, NULL, proc_pid); // Using execlp ... pr_no = start_process(cmdargs[0], & cmdargs[0], envp, proc_pid); // Using execve .... //cmdpid = start_command(cmdargs, envp, pidfile); // You can also use syslog if you do not like above logging //syslog(LOG_NOTICE, "Started process: %s", cmdline.val()); debug_("The child pid", pr_no); if (pr_no < 0) { fprintf(stderr, "\nFatal Error: Failed to start the process\n"); exit(-1); } sleep(WAIT_FOR_SYS); // wait for the process to come up // Get process name - only the first word from cmdline pr_name = strdup(basename(cmdargs[0])); // process name, does auto-malloc here } // generate log file names { char aa[21]; strncpy(aa, pr_name, 20); aa[20] = '\0'; // Define mon file-names - make it unique with combination of // process name and process id sprintf(mon_log, "mon/%s%d.log", aa, (int) proc_pid); } // Print out pid to log file if (pr_no > 0) { char aa[200]; sprintf(aa, "Process ID of %s is %d", pr_name, pr_no); error_msg(aa, mon_log, LOG_YES, STD_ERR_NO, DATE_YES); } // monitors the process - restarts if process dies... char print_log[200]; while (1) // infinite loop - monitor every 6 seconds { //debug_("Monitoring the process now...", "."); switch (kill(pr_no, 0)) { case 0: break; default: case ESRCH: // process died !! // ERSRCH means - No process can be found corresponding to pr_no // hence process had died !! sprintf(print_log, "Error ESRCH: No process or process group can be found for %d", pr_no); error_msg(print_log, mon_log, LOG_YES, STD_ERR_YES, DATE_YES); // You can also use syslog if you do not like above logging //syslog(LOG_NOTICE, "PROCESS DIED: %s", cmdline.val()); //pr_no = start_process(cmdline, NULL, NULL, proc_pid); // Using execlp .... pr_no = start_process(cmdargs[0], & cmdargs[0], envp, proc_pid); // Using execve .... // You can also use syslog if you do not like below logging //syslog(LOG_NOTICE, "Started process: %s", cmdline.val()); sprintf(print_log, "Starting process %s", pr_name); error_msg(print_log, mon_log, LOG_YES, STD_ERR_NO, DATE_NO); sleep(WAIT_FOR_SYS); // wait for the process to come up break; } sprintf(print_log, "Process ID of %s is %d", pr_name, pr_no); error_msg(print_log, mon_log, LOG_YES, STD_ERR_NO, DATE_NO); //debug_("Sleeping now ......", "."); sleep(sleep_sec); // Uncomment these to use micro-seconds // For real-time process control use micro-seconds or nana-seconds sleep functions // See 'man3 usleep', 'man 2 nanasleep' // If you do not have usleep() or nanosleep() on your system, use select() or poll() // specifying no file descriptors to test. //usleep(sleep_micro); // To sleep nano-seconds ... Uncomment these to use nano-seconds //struct timespec *req = new struct timespec; //req->tv_sec = 0; // seconds //req->tv_nsec = sleep_nano; // nanoseconds //nanosleep( (const struct timespec *)req, NULL); /* You can use select instead of sleep for portability struct timeval interval; interval.tv_sec += tmp; interval.tv_usec += (long int) 1e3 *tmp; select(1, NULL, NULL, NULL, & interval); */ } //closelog(); if using syslog } inline void error_msg(char *mesg_out, char *lg_file, bool pr_lg, bool std_err, bool pr_dt) { if (pr_lg) // (pr_lg == true) output to log file { char tmp_msg[BUFF_THOU]; if (pr_dt == true) // print date and message to log file 'lg_file' { sprintf(tmp_msg, "date >> %s; echo '\n%s\n' >> %s\n ", lg_file, mesg_out, lg_file); system(tmp_msg); } else { sprintf(tmp_msg, "echo '\n%s\n' >> %s\n ", mesg_out, lg_file); system(tmp_msg); } } if (std_err) // (std_err == true) output to standard error fprintf(stderr, "\n%s\n", mesg_out); //debug_("mesg_out", mesg_out); } // start a process and returns PID or -ve value if error // The main() function has envp arg as in - main(int argc, char *argv[], char **envp) int start_process(char *commandline, char *args[], char **envp, pid_t parent_pid) { int ff; unsigned long tsecs; tsecs = time(NULL); // time in secs since Epoch 1 Jan 1970 debug_("Time tsecs", tsecs); // Use fork2() instead of fork to avoid zombie child processes switch (ff = fork2(parent_pid, tsecs)) // fork creates 2 process each executing the following lines { case -1: fprintf(stderr, "\nFatal Error: start_process() - Unable to fork process\n"); _exit(errno); break; case 0: // child process debug_("\nStarting the start child process\n", " "); // For child process to ignore the interrupts (i.e. to put // child process in "background" mode. // Signals are sent to all processes started from a // particular terminal. Accordingly, when a program is to be run non-interactively // (started by &), the shell arranges that the program will ignore interrupts, so // it won't be stopped by interrupts intended for foreground processes. // Hence if previous value of signal is not IGN than set it to IGN. // Note: Signal handlers cannot be set for SIGKILL, SIGSTOP if (signal(SIGINT, SIG_IGN) == SIG_ERR) fprintf(stderr, "\nSignal Error: Not able to set signal to SIGINT\n"); else if (signal(SIGINT, SIG_IGN) != SIG_IGN) // program already run in background signal(SIGINT, SIG_IGN); // ignore interrupts if (signal(SIGHUP, SIG_IGN) == SIG_ERR) fprintf(stderr, "\nSignal Error: Not able to set signal to SIGHUP\n"); else if (signal(SIGHUP, SIG_IGN) != SIG_IGN) // program already run in background signal(SIGHUP, SIG_IGN); // ignore hangups if (signal(SIGQUIT, SIG_IGN) == SIG_ERR) fprintf(stderr, "\nSignal Error: Not able to set signal to SIGQUIT\n"); else if (signal(SIGQUIT, SIG_IGN) != SIG_IGN) // program already run in background signal(SIGQUIT, SIG_IGN); // ignore Quit if (signal(SIGABRT, SIG_IGN) == SIG_ERR) fprintf(stderr, "\nSignal Error: Not able to set signal to SIGABRT\n"); else if (signal(SIGABRT, SIG_IGN) != SIG_IGN) // program already run in background signal(SIGABRT, SIG_IGN); // ignore ABRT if (signal(SIGTERM, SIG_IGN) == SIG_ERR) fprintf(stderr, "\nSignal Error: Not able to set signal to SIGTERM\n"); else if (signal(SIGTERM, SIG_IGN) != SIG_IGN) // program already run in background signal(SIGTERM, SIG_IGN); // ignore TERM // sigtstp - Stop typed at tty. Ignore this so that parent process // be put in background with CTRL+Z or with SIGSTOP if (signal(SIGTSTP, SIG_IGN) == SIG_ERR) fprintf(stderr, "\nSignal Error: Not able to set signal to SIGTSTP\n"); else if (signal(SIGTSTP, SIG_IGN) != SIG_IGN) // program already run in background signal(SIGTSTP, SIG_IGN); // ignore TSTP // You can use debug_ generously because they do NOT increase program size! debug_("before execve commandline", commandline); debug_("before execve args[0]", args[0]); debug_("before execve args[1]", args[1]); debug_("before execve args[2]", args[2]); debug_("before execve args[3]", args[3]); debug_("before execve args[4]", args[4]); debug_("before execve args[5]", args[5]); debug_("before execve args[6]", args[6]); debug_("before execve args[7]", args[7]); debug_("before execve args[8]", args[8]); debug_("before execve args[9]", args[9]); execve(commandline, args, envp); // execlp, execvp does not provide expansion of metacharacters // like <, >, *, quotes, etc., in argument list. Invoke // the shell /bin/sh which then does all the work. Construct // a string 'commandline' that contains the complete command //execlp("/bin/sh", "sh", "-c", commandline, (char *) 0); // if success than NEVER returns !! // If execlp returns than there is some serious error !! And // executes the following lines below... fprintf(stderr, "\nFatal Error: Unable to start child process\n"); ff = -2; exit(127); break; default: // parent process // child pid is ff; if (ff < 0) fprintf(stderr, "\nFatal Error: Problem while starting child process\n"); { char buff[BUFF_HUN]; FILE *fp1; sprintf(buff, "mon/%d%lu.out", (int) parent_pid, tsecs); // tsecs is unsigned long fp1 = fopen(buff, "r"); if (fp1 != NULL) { buff[0] = '\0'; fgets(buff, BUFF_HUN, fp1); ff = atoi(buff); } fclose(fp1); debug_("start process(): ff - ", ff); #ifndef DEBUG_PRT sprintf(buff, "rm -f mon/%d%lu.out", (int) parent_pid, tsecs); system(buff); #endif // DEBUG_PRT } // define wait() to put child process in foreground or else put in background //waitpid(ff, & status, WNOHANG || WUNTRACED); //waitpid(ff, & status, WUNTRACED); //wait(& status); break; } return ff; } /* fork2() -- like fork, but the new process is immediately orphaned * (won't leave a zombie when it exits) * Returns 1 to the parent, not any meaningful pid. * The parent cannot wait() for the new process (it's unrelated). */ /* This version assumes that you *haven't* caught or ignored SIGCHLD. */ /* If you have, then you should just be using fork() instead anyway. */ int fork2(pid_t parent_pid, unsigned long tsecs) { pid_t mainpid, child_pid = -10; int status; char buff[BUFF_HUN]; if (!(mainpid = fork())) { switch (child_pid = fork()) { case 0: //child_pid = getpid(); //debug_("At case 0 fork2 child_pid : ", child_pid); return 0; case -1: _exit(errno); /* assumes all errnos are <256 */ default: debug_("fork2 child_pid : ", (int) child_pid); sprintf(buff, "echo %d > mon/%d%lu.out", (int) child_pid, (int) parent_pid, tsecs); system(buff); _exit(0); } } //debug_("fork2 pid : ", pid); if (mainpid < 0 || waitpid(mainpid, & status, 0) < 0) return -1; if (WIFEXITED(status)) if (WEXITSTATUS(status) == 0) return 1; else errno = WEXITSTATUS(status); else errno = EINTR; /* well, sort of :-) */ return -1; } // // char respawn[1024]; // strcpy(respawn, cmdline); // For "C" program use kill(pid_t process, int signal) function. // #include // See 'man 2 kill' // Returns 0 on success and -1 with errno set. // kill -0 $pid 2>/dev/null || respawn // To get the exit return status do -- // kill -0 $pid 2>/dev/null | echo $? // Return value 0 is success and others mean failure // Sending 0 does not do anything to target process, but it tests // whether the process exists. The kill command will set its exit // status based on this process. // // Alternatively, you can use // ps -p $pid >/dev/null 2>&1 || respawn // To get the exit return status do -- // ps -p $pid >/dev/null 2>&1 | echo $? // Return value 0 is success and others mean failure //*********************************************************************** // You can use pidfile to get the process id //********************************************************************** /* void poll_pidfile(char *pidfile) { struct stat buf; struct timeval interval = { tv_sec:0, tv_usec:(long int) 1e4 }; // 10 miliseconds while (stat(pidfile, & buf) ? errno == ENOENT : buf.st_size == 0) { struct timeval i = interval; select(1, NULL, NULL, NULL, & i); } } int start_command(char **args, char **envp, char *pidfile) { pid_t cmdpid; if (pidfile != NULL) { switch (unlink(pidfile)) { case ENOENT: case 0: break; default: return -1; } } switch (cmdpid = fork2()) { case 0: // child execve(args[0], args, envp); exit(-1); case -1: // error return -1; default: // parent break; } if (pidfile != NULL) { FILE *pf; poll_pidfile(pidfile); if ((pf = fopen(pidfile, "r")) == NULL) { syslog(LOG_ERR, "failed to read pidfile, using fork2 pid"); } else { char textpid[1024]; pid_t pid; fgets(textpid, sizeof(textpid), pf); if ((pid = atoi(textpid)) != -1) { cmdpid = pid; } else syslog(LOG_ERR, "failed to find pid in pidfile, using fork2 pid"); fclose(pf); } } return cmdpid; } */ ______________________________________________________________________ 4. File debug.cpp // From your browser save this file as text-file named as 'debug.cpp'. ______________________________________________________________________ #ifdef DEBUG_PRT #include "debug.h" // Variable value[] can be char, string, int, unsigned long, float, etc... void local_dbg(char name[], char value[], char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %s\n", fname, lineno, name, value ); } void local_dbg(char name[], int value, char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %d\n", fname, lineno, name, value ); } void local_dbg(char name[], unsigned int value, char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %d\n", fname, lineno, name, value ); } void local_dbg(char name[], long value, char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %ld\n", fname, lineno, name, value ); } void local_dbg(char name[], unsigned long value, char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %ld\n", fname, lineno, name, value ); } void local_dbg(char name[], short value, char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %d\n", fname, lineno, name, value ); } void local_dbg(char name[], unsigned short value, char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %d\n", fname, lineno, name, value ); } void local_dbg(char name[], float value, char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %f\n", fname, lineno, name, value ); } void local_dbg(char name[], double value, char fname[], int lineno, bool logfile) { printf("\nDebug %s Line: %d %s is = %f\n", fname, lineno, name, value ); } // You add many more here - value can be a class, ENUM, datetime, etc... #endif // DEBUG_PRT ______________________________________________________________________ 5. File debug.h // From your browser save this file as text-file named as 'debug.h'. ______________________________________________________________________ #ifdef DEBUG_PRT #include //#include //#include // assert() macro which is also used for debugging // Debugging code // Use debug2_ to output result to a log file #define debug_(NM, VL) (void) ( local_dbg(NM, VL, __FILE__, __LINE__) ) #define debug2_(NM, VL, LOG_FILE) (void) ( local_dbg(NM, VL, __FILE__, __LINE__, LOG_FILE) ) void local_dbg(char name[], char value[], char fname[], int lineno, bool logfile= false); void local_dbg(char name[], int value, char fname[], int lineno, bool logfile= false); void local_dbg(char name[], unsigned long value, char fname[], int lineno, bool logfile= false); void local_dbg(char name[], float value, char fname[], int lineno, bool logfile= false); #else #define debug_(NM, VL) ((void) 0) #define debug2_(NM, VL, LOG_FILE) ((void) 0) #endif // DEBUG_PRT ______________________________________________________________________ 6. Makefile # From your browser save this file as text-file named as 'Makefile'. ______________________________________________________________________ #//***************************************************************** #// Copyright policy is GNU/GPL and it is requested that #// you include author's name and email on all copies #// Author : Al Dev Email: alavoor@yahoo.com #//***************************************************************** .SUFFIXES: .pc .cpp .c .o HOSTFLAG=-DLinux #HOSTFLAG=-DSunOS CC=gcc CXX=g++ MAKEMAKE=mm #LIBRARY=libString.a DEST=/home/myname/lib # Note: You should set only ONE value of MYCFLAGS below, that is only # one line is uncommented and others are commented. # Use options -Wall (all warning msgs) -O3 (optimization) MYCFLAGS=-DDEBUG_PRT -g3 -Wall #MYCFLAGS=-O3 -Wall #PURIFY=purify -best-effort SRCS=procautostart.cpp debug.cpp #HDR=my_malloc.h String.h StringTokenizer.h File.h debug.h string_multi.h #LIBOBJS=my_malloc.o String.o StringTokenizer.o File.o debug.o OBJS=procautostart.o debug.o EXE=procautostart # For generating makefile dependencies.. SHELL=/bin/sh CPPFLAGS=$(MYCFLAGS) $(OS_DEFINES) CFLAGS=$(MYCFLAGS) $(OS_DEFINES) # # If the libString.a is in the current # directory then use -L. (dash L dot) MYLIBDIR=-L$(MY_DIR)/libmy -L. ALLLDFLAGS= $(LDFLAGS) $(MYLIBDIR) COMMONLIBS=-lstdc++ -lm MYLIBS=-lString LIBS=$(COMMONLIBS) $(MYLIBS) all: $(LIBRARY) $(EXE) $(MAKEMAKE): @rm -f $(MAKEMAKE) $(PURIFY) $(CXX) -M $(INCLUDE) $(CPPFLAGS) *.cpp > $(MAKEMAKE) $(EXE): $(OBJS) $(LIBRARY) @echo "Creating a executable " $(PURIFY) $(CC) -o $(EXE) $(OBJS) $(ALLLDFLAGS) $(LIBS) #$(LIBRARY): $(LIBOBJS) # @echo "\n***********************************************" # @echo " Loading $(LIBRARY) ... to $(DEST)" # @echo "***********************************************" # @ar cru $(LIBRARY) $(LIBOBJS) # @echo "\n " .cpp.o: $(SRCS) $(HDR) # @echo "Creating a object files from " $*.cpp " files " $(PURIFY) $(CXX) -c $(INCLUDE) $(HOSTFLAG) $(CPPFLAGS) $*.cpp .c.o: $(SRCS) $(HDR) # @echo "Creating a object files from " $*.c " files " $(PURIFY) $(CC) -c $(INCLUDE) $(HOSTFLAG) $(CFLAGS) $*.c clean: rm -f *.o *.log *~ *.log.old *.pid core err a.out lib*.a afiedt.buf *.class tags rm -f $(EXE) rm -f $(MAKEMAKE) ln -s ../cpphowto/libString.a . #%.d: %.c # @echo "Generating the dependency file *.d from *.c" # $(SHELL) -ec '$(CC) -M $(CPPFLAGS) $< | sed '\''s/$*.o/& $@/g'\'' > $@' #%.d: %.cpp # @echo "Generating the dependency file *.d from *.cpp" # $(SHELL) -ec '$(CC) -M $(CPPFLAGS) $< | sed '\''s/$*.o/& $@/g'\'' > $@' # Must include all the c flags for -M option #$(MAKEMAKE): # @echo "Generating the dependency file *.d from *.cpp" # $(CXX) -M $(INCLUDE) $(CPPFLAGS) *.cpp > $(MAKEMAKE) include $(MAKEMAKE) #include $(SRCS:.cpp=.d) #include $(SRCS:.c=.d) ______________________________________________________________________ 7. Testing the program - monitor_test From your browser save this file as text-file named as 'monitor_test'. Use this program for testing the 'procautostart' program. For example - ______________________________________________________________________ procautostart -n 12 -c "monitor_test -d $HOME -a dummy_arg " nohup & ______________________________________________________________________ Here procautostart will be checking the process monitor_test every 12 seconds. ______________________________________________________________________ #!/bin/ksh # Program to test the procautostart echo "Started the monitor_test ...." date > monitor_test.log while : do date >> monitor_test.log sleep 2 done ______________________________________________________________________ Then do a tail command to monitor the output. And simulate the fail­ ures of monitor_test programs. ______________________________________________________________________ bash$ tail -f monitor_test.log bash$ ps -ef | grep monitor_test See the PID of monitor_test and kill it.. bash$ kill -9 < PID of monitor_test > ______________________________________________________________________ Once you kill the process, you will notice that it immediately comes alive due to procautostart ! 8. Other Monitoring Tools 8.1. Unix init command The init command is a cool tool to do simple process monitoring. Add :respawn: entry to your /etc/inittab, if you need procees to be respawned. See the online manual page by typing 'man init' at bash prompt. 8.2. OpenSource Monitoring Tools On linux systems you can find the following packages. If it is not in the main cdrom than you must check in the contrib cdrom : · On contrib cdrom daemontools*.rpm · · · · · · · To use top commands type at unix prompt - ___________________________________________________________________ $ top $ ktop $ gtop ___________________________________________________________________ 8.3. Monitoring Tool - "daemontools" Visit the web site of daemontools at To install the daemontools RPM, do - ______________________________________________________________________ # rpm -i /mnt/cdrom/daemontools*.html # man supervise ______________________________________________________________________ supervise monitors a service. It starts the service and restarts the service if it dies. The companion svc program stops, pauses, or restarts the service on sysadmin request. The svstat program prints a one-line status report. See man page by 'man supervise' svc - control a supervised service. svc changes the status of a supervise-monitored service. dir is the same directory used for supervise. You can list several dirs. svc will change the status of each service in turn. svstat - print the status of a supervised service. svstat prints the status of a supervise-monitored service. dir is the same directory used for supervise. You can list several dirs. svstat will print the status of each service in turn. cyclog writes a log to disk. It automatically synchronizes the log every 100KB (by default) to guarantee data integrity after a crash. It automatically rotates the log to keep it below 1MB (by default). If the disk fills up, cyclog pauses and then tries again, without losing any data. See man page by 'man cyclog' accustamp puts a precise timestamp on each line of input. The timestamp is a numeric TAI timestamp with microsecond precision. The companion tailocal program converts TAI timestamps to local time. See 'man accustamp' usually watches a log for lines that do not match specified patterns, copying those lines to stderr. The companion errorsto program redirects stderr to a file. See 'man usually' setuser runs a program under a user's uid and gid. Unlike su, setuser does not gain privileges; it does not check passwords, and it cannot be run except by root. See 'man setuser' 8.4. Commercial Monitoring Tools There are commercial monitoring tools available. Check out - · BMC Patrol for Unix/Databases · TIBCO corp's Hawk for Unix monitoring · LandMark corporation · Platinum corporation 9. Related URLs Linux goodies main site is at Mirror sites are at - , angelfire , geocities , virtualave , 50megs , theglobe , NBCi , Terrashare , Fortunecity , Freewebsites , Tripod , Spree , Escalix , Httpcity , Freeservers . 10. Other Formats of this Document This document is published in 14 different formats namely - DVI, Postscript, Latex, Adobe Acrobat PDF, LyX, GNU-info, HTML, RTF(Rich Text Format), Plain-text, Unix man pages, single HTML file, SGML (Linuxdoc format), SGML (Docbook format), MS WinHelp format. This howto document is located at - · and click on HOWTOs and search for howto document name using CTRL+f or ALT+f within the web-browser. You can also find this document at the following mirrors sites - · · · · · Other mirror sites near you (network-address-wise) can be found at select a site and go to directory /LDP/HOWTO/xxxxx-HOWTO.html · You can get this HOWTO document as a single file tar ball in HTML, DVI, Postscript or SGML formats from - and · Plain text format is in: and · Single HTML file format is in: Single HTML file can be created with command (see man sgml2html) - sgml2html -split 0 xxxxhowto.sgml · Translations to other languages like French, German, Spanish, Chinese, Japanese are in and Any help from you to translate to other languages is welcome. The document is written using a tool called "SGML-Tools" which can be got from - Compiling the source you will get the following commands like · sgml2html xxxxhowto.sgml (to generate html file) · sgml2html -split 0 xxxxhowto.sgml (to generate a single page html file) · sgml2rtf xxxxhowto.sgml (to generate RTF file) · sgml2latex xxxxhowto.sgml (to generate latex file) 10.1. Acrobat PDF format PDF file can be generated from postscript file using either acrobat distill or Ghostscript. And postscript file is generated from DVI which in turn is generated from LaTex file. You can download distill software from . Given below is a sample session: ______________________________________________________________________ bash$ man sgml2latex bash$ sgml2latex filename.sgml bash$ man dvips bash$ dvips -o filename.ps filename.dvi bash$ distill filename.ps bash$ man ghostscript bash$ man ps2pdf bash$ ps2pdf input.ps output.pdf bash$ acroread output.pdf & ______________________________________________________________________ Or you can use Ghostscript command ps2pdf. ps2pdf is a work-alike for nearly all the functionality of Adobe's Acrobat Distiller product: it converts PostScript files to Portable Document Format (PDF) files. ps2pdf is implemented as a very small command script (batch file) that invokes Ghostscript, selecting a special "output device" called pdfwrite. In order to use ps2pdf, the pdfwrite device must be included in the makefile when Ghostscript was compiled; see the documentation on building Ghostscript for details. 10.2. Convert Linuxdoc to Docbook format This document is written in linuxdoc SGML format. The Docbook SGML format supercedes the linuxdoc format and has lot more features than linuxdoc. The linuxdoc is very simple and is easy to use. To convert linuxdoc SGML file to Docbook SGML use the program ld2db.sh and some perl scripts. The ld2db output is not 100% clean and you need to use the clean_ld2db.pl perl script. You may need to manually correct few lines in the document. · Download ld2db program from or from Al Dev site · Download the cleanup_ld2db.pl perl script from from Al Dev site The ld2db.sh is not 100% clean, you will get lots of errors when you run ___________________________________________________________________ bash$ ld2db.sh file-linuxdoc.sgml db.sgml bash$ cleanup.pl db.sgml > db_clean.sgml bash$ gvim db_clean.sgml bash$ docbook2html db.sgml ___________________________________________________________________ And you may have to manually edit some of the minor errors after run­ ning the perl script. For e.g. you may need to put closing tag < /Para> for each < Listitem> 10.3. Convert to MS WinHelp format You can convert the SGML howto document to Microsoft Windows Help file, first convert the sgml to html using: ______________________________________________________________________ bash$ sgml2html xxxxhowto.sgml (to generate html file) bash$ sgml2html -split 0 xxxxhowto.sgml (to generate a single page html file) ______________________________________________________________________ Then use the tool HtmlToHlp . You can also use sgml2rtf and then use the RTF files for generating winhelp files. 10.4. Reading various formats In order to view the document in dvi format, use the xdvi program. The xdvi program is located in tetex-xdvi*.rpm package in Redhat Linux which can be located through ControlPanel | Applications | Publishing | TeX menu buttons. To read dvi document give the command - xdvi -geometry 80x90 howto.dvi man xdvi And resize the window with mouse. To navigate use Arrow keys, Page Up, Page Down keys, also you can use 'f', 'd', 'u', 'c', 'l', 'r', 'p', 'n' letter keys to move up, down, center, next page, previous page etc. To turn off expert menu press 'x'. You can read postscript file using the program 'gv' (ghostview) or The ghostscript program is in ghostscript*.rpm package and gv program is in gv*.rpm package in Redhat Linux which can be located through ControlPanel | Applications | Graphics menu buttons. The gv program is much more user friendly than ghostscript. Also ghostscript and gv are available on other platforms like OS/2, Windows 95 and NT, you view this document even on those platforms. · Get ghostscript for Windows 95, OS/2, and for all OSes from To read postscript document give the command - gv howto.ps ghostscript howto.ps You can read HTML format document using Netscape Navigator, Microsoft Internet explorer, Redhat Baron Web browser or any of the 10 other web browsers. You can read the latex, LyX output using LyX a X-Windows front end to latex. 11. Copyright Notice Copyright policy is GNU/GPL as per LDP (Linux Documentation project). LDP is a GNU/GPL project. Additional restrictions are - you must retain the author's name, email address and this copyright notice on all the copies. If you make any changes or additions to this document than you should intimate all the authors of this document.