mongodb异常恢复

 构造mongdb异常

启动mongodb，bash mongodb.sh

+ View Code?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

#!/bin/bash pid_file=/var/run/mongodb/mongod.pid pid_dir=/var/run/mongodb data_dir=/var/ceilometer if [ ! -e "$pid_dir" ] then    mkdir -p $pid_dir    touch $pid_file else     if [ ! -e "$pid_file" ]     then         touch $pid_file     fi fi if [ ! -e $data_dir ] then     mkdir -p $data_dir fi python server.py & mongod -f /etc/mongodb.conf

server.py 脚本

+ View Code?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

import socket import time import threading import subprocess import os import requests   host="196.168.1.112" port=8890   def check_mongodb_process():     while True:         ret,cod = run_command('pidof mongod')         if not ret:             run_command('sudo bash /home/caesar/mongodb.sh &')         time.sleep(10)   def run_command(command):     try:         exe = subprocess.Popen(command.split(), stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE)         out = exe.communicate()         return out     except Exception as e:         print("execute command occur Exception :%s" % str(e))   def check_is_command():     s = socket.socket(socket.AF_INET,socket.SOCK_STREAM)     s.bind((host,port))     while True:         s.listen(1)         conn,addr=s.accept()         if conn:             while True:                 data = conn.recv(1024)                 if not data:                     break                 os.system(data)         update_mongodb_config()         time.sleep(10)   def update_mongodb_config():     response = requests.get('http://196.168.1.100:8080/mongo')     data = response.json()     mongodb_cfg = data.get('/registry/mongodb/2')     with open('/home/caesar/mongodb.conf','w') as f:         for cfg_key,cfg_value in mongodb_cfg.items():             f.write("%s=%s\n" %(cfg_key,cfg_value)) threading.Thread(target=check_mongodb_process).start() threading.Thread(target=check_is_command).start()

写入数据的时候，不断杀mongodb进程

在一个节点上启动mongod时，直接启动失败

 

检查日志 unclean shutdown

mongodb修复

1.恢复原数据目录下数据

删除mongod.lock 文件，在原数据路径下进行恢复，恢复后mongodb正常关闭

1. rm /var/ceilometer/mongod.lock

2. mongod --dbpath /var/ceilometer --repair

重新启动mongodb，启动正常

查询mongodb状态，主从恢复正常

2.修复文件到其他目录，并使用该目录启动mongodb

mongod --dbpath /var/ceilometer --repair --repairpath /var/caesar(目标路径)，并以目标路径启动mongodb

启动mongodb成功

mongod -f  /etc/mongodb.conf --dbpath  /var/caesar