<span>mongodb启动关闭</span>

在mongodb的启动时,在数据目录下,会生成一个mongod.lock文件。如果在正常退出时,会清除这个mongod.lock文件,若要是异常退出,在下次启动的时候,会禁止启动,从而保留一份干净的一份副本数据。有人可能会想到删除这个文件,

建议请不要这么做。如果这么做,我们也不知道数据文件是否会损坏,如果mongod.lock文件阻止mongod的启动,请对数据文件进行修复,而不是简单的删除该文件。而这里的mongod.lock文件存放的是:启动mongod的进程号.

 

正常启动:

/home/work/mongodb/4.0.17/bin/mongod --config /home/work/mongodb/mongo_27000/etc/mongodb.conf

#############################

直接运行单实例:
/home/work/mongodb/4.0.17/bin/mongod --replSet myapp --dbpath  /home/work/mongodb/mongo_27000/data --port 27017  --logpath  /home/work/mongodb/mongo_27000/log/mongodb.log  --fork
 

 


异常关闭后启动:

比如kill -9关闭实例

正常启动出现以下报错:

waiting until server is ready for connections.

 


解决:

1, 删除data目录下的lock文件:

[work@ data]$ tree -L 2
.
├── admin
│ ├── collection
│ └── index
├── apple
│ ├── collection
│ └── index
├── config
│ ├── collection
│ └── index
├── diagnostic.data
│ ├── metrics.2020-09-03T06-58-33Z-00000
│ ├── metrics.2020-09-04T07-03-33Z-00000
│ ├── metrics.2020-09-05T07-03-33Z-00000
│ ├── metrics.2020-09-06T06-58-33Z-00000
│ ├── metrics.2020-09-07T06-53-33Z-00000
│ ├── metrics.2020-09-08T06-38-33Z-00000
│ ├── metrics.2020-09-09T06-18-33Z-00000
│ ├── metrics.2020-09-10T06-03-33Z-00000
│ ├── metrics.2020-09-11T05-48-33Z-00000
│ ├── metrics.2020-09-12T05-28-33Z-00000
│ ├── metrics.2020-09-13T05-08-33Z-00000
│ ├── metrics.2020-09-14T04-48-33Z-00000
│ ├── metrics.2020-09-15T04-28-33Z-00000
│ ├── metrics.2020-09-16T04-13-33Z-00000
│ ├── metrics.2020-09-17T03-28-33Z-00000
│ ├── metrics.2020-09-18T02-43-33Z-00000
│ ├── metrics.2020-09-19T02-08-33Z-00000
│ ├── metrics.2020-09-20T01-38-33Z-00000
│ ├── metrics.2020-09-21T01-03-33Z-00000
│ ├── metrics.2020-09-22T00-18-33Z-00000
│ ├── metrics.2020-09-22T03-13-43Z-00000
│ ├── metrics.2020-09-22T03-18-43Z-00000
│ └── metrics.interim
├── google
│ ├── collection
│ └── index
├── journal
│ ├── WiredTigerLog.0000000503
│ ├── WiredTigerPreplog.0000000001
│ └── WiredTigerPreplog.0000000002
├── local
│ ├── collection
│ └── index
├── _mdb_catalog.wt
├── mongod.lock
├── sizeStorer.wt
├── storage.bson
├── test
│ ├── collection
│ └── index
├── WiredTiger
├── WiredTigerLAS.wt
├── WiredTiger.lock
├── WiredTiger.turtle
└── WiredTiger.wt

 


2,修复:

./mongod --repair --config /home/work/mongodb/mongo_27000/etc/mongodb.conf

 


3,开启:

/home/work/mongodb/4.0.17/bin/mongod --config /home/work/mongodb/mongo_27000/etc/mongodb.conf

 


 

   关闭:

1. 登陆 正常关闭

use admin;
db.shutdownServer();
use admin
db.shutdownServer();

 

#########################################

glc-test:PRIMARY> use admin
switched to db admin
glc-test:PRIMARY> db.shutdownServer();
server should be down...
2020-09-22T14:02:54.171+0800 I NETWORK [js] trying reconnect to XXX:27000 failed
2020-09-22T14:02:55.282+0800 I NETWORK [js] reconnect XXX:28042 failed failed 
2020-09-22T14:02:55.285+0800 I NETWORK [js] trying reconnect to XXX:27000 failed
2020-09-22T14:02:55.285+0800 I NETWORK [js] reconnect XXX:28042 failed failed 
> ^C
2020-09-22T14:04:18.256+0800 I NETWORK [js] trying reconnect to XXX:27000 failed
2020-09-22T14:04:18.258+0800 I NETWORK [js] reconnect XXX:27000 ok
exit status 0

 


2. mongod 命令 正常关闭,最佳方式

[work@xxx ] /home/work/mongodb/4.0.17/bin/mongod --shutdown --dbpath /home/work/mongodb/mongo_27000/data


2020-09-22T14:09:29.221+0800 I CONTROL [main] Automatically disabling TLS 1.0, to force-enable TLS 1.0 specify --sslDisabledProtocols 'none'
killing process with pid: 371416

 

3. kill 命令 正常关闭

[work@xxx]$ ps aux|grep mongod
work 394902 6.6 0.0 1432792 57620 ? SLl 14:12 0:02 /home/work/mongodb/4.0.17/bin/mongod --config /home/work/mongodb/mongo_27000/etcmongodb.conf
work 397490 0.0 0.0 112644 964 pts/0 S+ 14:12 0:00 grep --color=auto mongod
[work@xxx mongodb]$ kill 394902

 


4. kill -9 命令 异常关闭,最坏方式

[work@xxx]$ ps aux|grep mongod
work 394902 6.6 0.0 1432792 57620 ? SLl 14:12 0:02 /home/work/mongodb/4.0.17/bin/mongod --config /home/work/mongodb/mongo_27000/etcmongodb.conf
work 397490 0.0 0.0 112644 964 pts/0 S+ 14:12 0:00 grep --color=auto mongod
[work@xxx mongodb]$ kill -9 394902
# 建议不要使用 ”kill -9 pid“,因为如果 MongoDB 运行在没开启日志(--journal)的情况下,
可能会造成数据损失。

  无论如何都要避免直接 kill -9 MongoDB 进程,如需正常安全的退出,向进程发送 SIGINT 或 SIGTERM 信号,如:

$ sudo kill -2 pid // SIGINT// 或$ sudo kill pid // SIGTERM

 

全部评论

相关推荐

死在JAVA的王小美:哈哈哈哈哈哈哈哈哈哈哈哈哈哈哈哈哈哈,我也是,让我免了一轮,但是硬气拒绝了
点赞 评论 收藏
分享
评论
点赞
收藏
分享
牛客网
牛客企业服务