path: root/py_cmd/tsg_software_reboot.py

#coding=utf-8
import os
import sys
import syslog
import argparse
import subprocess
import time
import re
import logging
import logging.handlers
from sys import path
path.append('../py_common') #将存放module的路径添加进来
path.append('./py_common') #将存放module的路径添加进来
#from common_logger import *
from common_modules_deploy import *
from common_modules_operator import *
from common_args import *
from common_whoami import *

##define KERN_EMERG    "<0>"  /* system is unusable               */
##define KERN_ALERT    "<1>"  /* action must be taken immediately */
##define KERN_CRIT     "<2>"  /* critical conditions              */
##define KERN_ERR      "<3>"  /* error conditions                 */
##define KERN_WARNING  "<4>"  /* warning conditions               */
##define KERN_NOTICE   "<5>"  /* normal but significant condition */
##define KERN_INFO     "<6>"  /* informational                    */
##define KERN_DEBUG    "<7>"  /* debug-level messages             */

G_SYS_LOG_STRING = ['EMERG', 'ALERT', 'CRIT', 'ERR', 'WARNING', 'NOTICE', 'INFO', 'DEBUG']
G_SYSLOG_EVENT_SRC = "CLI"
G_SOFT_REBOOT_NAME = "tsg_software_reboot"
G_SOFT_REBOOT_VERSION = "1.0.3 20190827"

class CommandException(Exception):
    pass
    
def tsg_restart_err_log(error_num, user_msg):   
    syslog_msg = "[%s] %s:%s" %(G_SYSLOG_EVENT_SRC, "tsg software reboot error", user_msg)
    print_msg = "%s:%s" %("tsg software reboot error", user_msg)
    syslog.syslog(syslog.LOG_ERR, syslog_msg)
    print (print_msg)
    
    sys.exit(error_num)

def tsg_restart_succ_log():
    syslog_msg = "[%s] %s" %(G_SYSLOG_EVENT_SRC, "tsg software reboot success")
    syslog.syslog(syslog.LOG_NOTICE, syslog_msg)
    print("tsg software reboot success")
    sys.exit(0)

#return exitcode value + output message:
#    0: succ
#    1: error
def system_cmd_run(cmd_str):
    dangerous_cmd = {"rm", "mv", "poweroff", "shutdown"}
    
    for cmd in dangerous_cmd:
        pattern = "\s*%s" %(cmd)
        match_str = re.match(pattern, cmd_str)
        if not match_str is None:
            print("can't run this cmd:%s" %(cmd_str))
            sys.exit(1)
            
    try:
        exitcode, output = subprocess.getstatusoutput(cmd_str)
    except Exception as e:  
        print(e)    
        print("###### %s" %(e.message))
        #if exitcode != 0:
        #    output = ""
        return 1, e.message

    return exitcode, output
    
#return value:
# 1: progcess of prog_name is exist
# 0: progcess of prog_name is not exist  
def tsg_check_process_health_by_ps(module_name):
    cmd_str = "ps -afx | grep -w %s$ | grep -v grep" %(module_name)
    exitcode, output = system_cmd_run(cmd_str)
    if exitcode == 0:
        return 1
    
    return 0

#return value:
# 1: progcess of prog_name is exist
# 0: progcess of prog_name is not exist    
def tsg_check_process_health_by_systemctl_status(module_name):
    print("systemctl_status check method TODO!")
    sys.exit(1)
    return 0

def tsg_kill_app_process_by_killall(module_name, extra_progs):
    #todo , stop sapp, xxx, check process exist or not, maybe zombie, maybe very slow
    command = "killall -9 %s %s" %(module_name, extra_progs)
    try:    
        exitcode, output = subprocess.getstatusoutput(command)
        #print("%d" %(exitcode))
    except Exception as e:
        pass
    
    if exitcode != 0:
        return 1
        
    return 0    

def tsg_kill_app_process_by_systemctl_stop(module_name):
    print("tsg_kill_app_process_by_systemctl_stop  TODO!!!")
    sys.exit(1)
    return 0
    
def tsg_stop_app_process(module_name, extra_progs, stop_func, check_func):
    res_code = 0
    running_flag = 0
    
    running_flag = check_func(module_name)    
        
    if running_flag == 0:
        #tsg_logger.debug("%s is not running, ignore stop, try start it..." %(module_name))
        return 0
    #ready to stop progcess,  retry for 3 times
    
    for times in range(5):
        stop_func(module_name, extra_progs)
        #此处不判断stop_func的返回值, 可能程序不存在, 可能守护不存在,等原因
        #直接用check()方法检测stop()的成功
        res_code = tsg_check_process_health_by_ps(module_name)
        if res_code != 0:
            time.sleep(1 + times)
            continue
        else:
            break    
            
    if res_code != 0: 
        errmsg = "can't stop process %s" %(module_name)
        #tsg_logger.error(errmsg)
        tsg_restart_err_log(res_code, errmsg)
    
    return res_code

def tsg_start_app_process_by_exec_call(module_name, module_cwd, module_exe):
    #tsg_logger.debug("try cd to dir:%s" %(module_cwd))
    try:
        os.chdir(module_cwd)
    except Exception as e: 
        print("%s" %(e))
        return 1
    
    #tsg_logger.debug("try call exec :%s" %(module_exe))
    cmd_str = "./%s" %(module_exe)
    ret_code, output = system_cmd_run(cmd_str)
    if ret_code != 0:
        errmsg = "start program %s error, call %s/%s failed" %(module_name, module_cwd, module_exe)
        tsg_restart_err_log(ret_code, errmsg)

    return 0
    
def tsg_start_app_process_by_systemctl_start(module_name, module_cwd, module_exe, check_method):
    print("tsg_start_app_process_by_systemctl_start  TODO!!!!!")
    sys.exit(1)

def tsg_start_app_process(module_name, module_cwd, module_exe, start_func, check_func):
    ret = start_func(module_name, module_cwd, module_exe)        
    if ret != 0:
        #tsg_logger.debug("call start cmd for %s failed" %(module_name))
        return 1

    running_flag = check_func(module_name)    
    if running_flag == 0:
        #tsg_logger.debug("call cmd succ, but %s process is not exist" %(module_name))
        errmsg = "start process %s error" %(module_name)
        return 1
 
    return 0
    
#参数说明:
#
# module_name: 模块名称
# extra_progs: 其他需要kill的附加程序, 如sapp的r3守护, 需要杀掉, 否则后台可能会重复启动sapp
# module_cwd:  应用的绝对路径
# module_exe: 启动应用的名称, 可能跟module_name不一样, 比如用r2启动sapp
# stop_method: 停止应用方法  
# start_method: 启动应用方法  
# check_method: 检测应用是否运行方法  
#  
def tsg_restart_app_process(module_name, extra_progs, module_cwd, module_exe, start_func, stop_func, check_func):
    res_code = 0
    
    #tsg_logger.debug("into func: tsg_restart_app_process()")
    
    res_code = tsg_stop_app_process(module_name, extra_progs, stop_func, check_func)
    if res_code != 0:
        return res_code
        
    res_code = tsg_start_app_process(module_name, module_cwd, module_exe, start_func, check_func)
    if res_code != 0:
        return res_code
    
    return 0

#根据配置文件的参数, 选择用那种操作函数继续下一步
#返回值, 函数指针:
#ret_code, start_fun, stop_fun, check_health_fun    
def tsg_get_operator_by_config(module_operator):    
    if module_operator[TSG_OP_MODULE_START_INDEX] == 'exec':
        start_func = tsg_start_app_process_by_exec_call
    elif module_operator[TSG_OP_MODULE_START_INDEX] == 'systemctl_start':
        start_func = tsg_start_app_process_by_systemctl_start
    else:
        errmsg = "not support start method:%s, only be [exec, systemctl_start]" %(module_operator[TSG_OP_MODULE_START_INDEX])
        tsg_restart_err_log(1, errmsg)

    if module_operator[TSG_OP_MODULE_STOP_INDEX] == 'killall':
        stop_func = tsg_kill_app_process_by_killall
    elif module_operator[TSG_OP_MODULE_STOP_INDEX] == 'systemctl_stop':
        stop_func = tsg_kill_app_process_by_systemctl_stop
    else:
        errmsg = "not support stop method:%s, only be [killall, systemctl_stop]" %(module_operator[TSG_OP_MODULE_STOP_INDEX])
        tsg_restart_err_log(1, errmsg)

    if module_operator[TSG_OP_MODULE_STATUS_INDEX] == 'ps':
        check_func = tsg_check_process_health_by_ps
    elif module_operator[TSG_OP_MODULE_STATUS_INDEX] == 'systemctl_status':
        check_func = tsg_check_process_health_by_systemctl_status
    else:
        errmsg = "not support check method:%s, only be [ps, systemctl_status]" %(module_operator[TSG_OP_MODULE_STATUS_INDEX])
        tsg_restart_err_log(1, errmsg)
    
    return 0, start_func, stop_func, check_func
    
def tsg_software_reboot():
    #根据IP地址, 获取本机的sled, id, name
    sled_type,sled_id,sled_name = tsg_whoami()
    if sled_name == "":
        tsg_restart_err_log(1, "can't get local sled name")
        sys.exit(1)
    
    log_handle = syslog.openlog(sled_name)
    
    #获取当前计算板运行的所有模块, 数组形式
    module_array = tsg_get_local_sled_modules(sled_name)    
    if len(module_array) <= 0:
        tsg_restart_err_log(1, "can't get local sled modules")
        sys.exit(1)
    
    #tsg_logger.debug("len(module_array) = %d" %(len(module_array)))  
    for module_name in module_array:
        module_operator = tsg_get_module_opertor(module_name)
        if len(module_operator) <= 0:
            tsg_restart_err_log(1, "can't get local sled module operator for %s" %(module_name))
            sys.exit(1)
        
        ret, start_func, stop_func, check_func =  tsg_get_operator_by_config(module_operator)    
        if ret != 0:
            tsg_restart_err_log(1, "can't get operator for %s" %(module_operator[TSG_OP_MODULE_NAME_INDEX]))
            sys.exit(1)
        
        tsg_restart_app_process(module_operator[TSG_OP_MODULE_NAME_INDEX], module_operator[TSG_OP_MODULE_EXTRA_INDEX], 
                                module_operator[TSG_OP_MODULE_CWD_INDEX], 
                                module_operator[TSG_OP_MODULE_EXE_INDEX],   
                                start_func, stop_func, check_func)

    #执行到此处肯定是成功了, 写syslog记录
    tsg_restart_succ_log()

def setup_cmd_args():
    parser = argparse.ArgumentParser(description='TSG OAM Argument Parser')
                     
    parser.add_argument('-v', '--version', help = 'Print version information',
                        action='store_true', default = 0)

    return parser
    
if __name__ == '__main__':
    #global tsg_logger
    
    #if len(sys.argv) >= 2 and sys.argv[1] == "debug":
    #    tsg_logger = tsg_logger_init(logging.DEBUG)
    #else:
    #    tsg_logger = tsg_logger_init(logging.CRITICAL)
    
    comm_arg_parser = setup_cmd_args()
    arg_options = comm_arg_parser.parse_args()
    
    if arg_options.version != 0:
        print("%s version: %s" %(G_SOFT_REBOOT_NAME, G_SOFT_REBOOT_VERSION))
        sys.exit(0)
        
    tsg_software_reboot()