mongoDB高级应用

副本集介绍
副本集是什么

- MongoDB复制是将数据同步在多个服务器的过程。

-复制提供了数据的冗余备份,并在多个服务器上存储数据副本,提高了数据的可用性,

并可以保证数据的安全性。 -复制还允许您从硬件故障和服务中断中恢复数据

复制原理

副本集工作过程

mongodb的复制至少需要两个节点。其中一个是主节点,负责处理客户端请求,

其余的都是从节点,负责复制主节点上的数据。

mongodb各个节点常见的搭配方式为:

一主一从、一主多从。

主节点记录在其上的所有操作oplog,从节点定期轮询主节点获取这些操作,

然后对自己的数据副本执行这些操作,从而保证从节点的数据与主节点一致。

副本集实现方式(续1 )

Replica Sets复制集

MongoDB在1.6版本对开发了新功能replica set,这比之前的replication功能要强大一些 ,增加了故障

自动切换和自动修复成员节点,各个DB之间数据完全一致, 大大降低了维护成功。

使用replica set故障切换完全自动。

Replica Sets的结构类似一个集群,完全可以把它当成一个集群,因为它确实与集群实现

的作用是一样的:

如果其中一个节点出现故障,其他节点马上会将业务

接管过来而无须停机操作

副本集实现方式

Master- Slave主从复制

实现数据同步只需要在某一台服务器启动时加上"master"参数,以指明此服务器的角色

是primary ;另-台服务器加上"-slave‘和" -source参数,以指明此服务器的角色是slave。

.主从复制的优点如下:

从服务器可以执行查询工作,降低主服务器访问压力。-在从服务器执行备份,

避免备份期间锁定主服务器的数据。

当主服务器出现故障时,可以快速切换到从服务器,减少宕机时间。

====================================================================

具体要求:

准备3台mongodb服务器

验证副本集配置

准备三台虚拟机,配置mongodb副本集,ip分别为192.158.4.51, 192.168.4.52,

192.168.4.53其中一个是主节点,负责处理客户端请求,其余的都是从节点,

负责复制主节点上的数据,

实现存储数据副本,提高了数据的可用性,

运行服务

启动服务时,指定主机所在副本集名称 副本集成员间使用形同的副本集名称

[ ~]# mkdir -p /data/db

[ ~]# mongod --bind_ip 192.168.200.102 --port 27061 --logpath=/var/l
og/mongod.log --replSet rs1 &       //指定副本集名称

[ ~]# 
[ ~]# netstat -lnpt|grep mongod
tcp   0  0 192.168.200.102:27078   0.0.0.0:*         LISTEN      6097/mongod         
tcp   0  0 192.168.200.102:27061   0.0.0.0:*         LISTEN      6276/mongod  

[ ~]# jobs
[1]+  运行中               mongod --bind_ip 192.168.200.102 --port 27061 --logpat
h=/var/log/mongod.log --replSet rs1 &

步骤一:创建mongodb副本集

1)三台主机安装mongodb(以4.51为例)

[ ~]# cd mongodb/
[ mongodb]#  mkdir  /usr/local/mongodb
[ mongodb]# cd /usr/local/mongodb
[ mongodb]# cp -r \ 
 /root/mongodb/mongodb-linux-x86_64-rhel70-3.6.3/bin/  .
[ mongodb]# ls
bin
[ mongodb]# mkdir etc
[ mongodb]# mkdir log
[ mongodb]# mkdir -p data/db
[ mongodb]# vim etc/mongodb.conf
dbpath=/usr/local/mongodb/data/db/
logpath=/usr/local/mongodb/log/mongodb.log
logappend=true 
fork=true
bind_ip=192.168.4.51
port=27077
replSet=rs1        
//加入到副本集,rs1名字随便起,想知道谁和我在一个副本集里,三台机器的名字一样,都写rs1

2)设置PATH变量

[ mongodb]# vim /etc/profile
export PATH=/usr/local/mongodb/bin:$PATH
[ mongodb]# source  /etc/profile

3)由于启动和停止服务名字太长,可以起一个别名

给停止服务起一个别名
[ mongodb]# alias cmdb=‘mongod  --dbpath=/usr/local/mongodb/data/db/  --shutdown‘

给启动服务起一个别名
[ mongodb]# alias smdb=‘mongod -f /usr/local/mongodb/etc/mongodb.conf‘

4)启动服务并连接

[ ~]# smdb
about to fork child process, waiting until server is ready for connections.
forked process: 5656
child process started successfully, parent exiting

[ ~]# mongo --host 192.168.4.51 --port 27077
MongoDB shell version v3.6.3
connecting to: mongodb://192.168.4.51:27077/
MongoDB server version: 3.6.3

5)配置集群信息,任意一台都可以,在这里在51上面操作

> rs1_config = {        //rs1_config随便起变量名,要记住
 _id:"rs1",         //必须为rs1这个,三台主机集群名,配置文件里面写的是这个
 members:[
 {_id:0,host:"192.168.4.51:27077"},        //_id值随意,host值固定
 {_id:1,host:"192.168.4.52:27078"},
 {_id:2,host:"192.168.4.53:27079"}
 ]
 };        //回车,出现下面情况为成功
{
    "_id" : "rs1",
    "members" : [
        {
            "_id" : 0,
            "host" : "192.168.4.51:27077"
        },
        {
            "_id" : 1,
            "host" : "192.168.4.52:27078"
        },
        {
            "_id" : 2,
            "host" : "192.168.4.53:27079"
        }
    ]
}
>

6)初始化Replica Sets环境

> rs.initiate(rs1_config)
{
    "ok" : 1,
    "operationTime" : Timestamp(1538187475, 1),
    "$clusterTime" : {
        "clusterTime" : Timestamp(1538187475, 1),
        "signature" : {
            "hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
            "keyId" : NumberLong(0)
        }
    }
}
rs1:SECONDARY> 
rs1:PRIMARY>            //提示PRIMARY,51为主

工作过程:

Bully算法(有最新操作为主、或者成为主的时间短)

选举   心跳   同步

rs1_config = {       
 _id:"rs1",         
 members:[
 {_id:0,host:"192.168.4.51:27077",priority:0},       
 {_id:1,host:"192.168.4.52:27078",priority:3},
 {_id:2,host:"192.168.4.53:27079",priority:2}
 ]
 };

7)在52和53上面查看

[ ~]# mongo --host 192.168.4.52 --port 27078
MongoDB shell version v3.6.3
connecting to: mongodb://192.168.4.52:27078/
MongoDB server version: 3.6.3
...
... 
rs1:SECONDARY>         //提示SECONDARY,52为从
rs1:SECONDARY> 
rs1:SECONDARY>
[ ~]# mongo --host 192.168.4.53 --port 27079
MongoDB shell version v3.6.3
connecting to: mongodb://192.168.4.53:27079/
MongoDB server version: 3.6.3
...
... 
rs1:SECONDARY>     //提示SECONDARY,53为从
rs1:SECONDARY>

注意:如果初始化错误,重启服务登陆之后重新设置变量,之后再重新初始化

8)查看状态信息

rs1:PRIMARY>  rs.status()
...
    "members" : [
        {
            "_id" : 0,
            "name" : "192.168.4.51:27077",
            "health" : 1,
            "state" : 1,
            "stateStr" : "PRIMARY",
            "uptime" : 2295,
            ...
...
        },
        {
            "_id" : 1,
            "name" : "192.168.4.52:27078",
            "health" : 1,
            "state" : 2,
            "stateStr" : "SECONDARY",
            "uptime" : 384,
            ...
          ...
        },
        {
            "_id" : 2,
            "name" : "192.168.4.53:27079",
            "health" : 1,
            "state" : 2,
            "stateStr" : "SECONDARY",
...
...

9)查看是否是master库

rs1:PRIMARY>  rs .isMaster( )
{
    "hosts" : [
        "192.168.4.51:27077",
        "192.168.4.52:27078",
        "192.168.4.53:27079"
    ],
    "setName" : "rs1",
    "setVersion" : 1,
    "ismaster" : true,        //主库
    "secondary" : false,
    "primary" : "192.168.4.51:27077",
    "me" : "192.168.4.51:27077",
...
...

10)验证副本集,同步数据验证(51上面写数据)

rs1:PRIMARY> use gamedb2
switched to db gamedb2
rs1:PRIMARY> db.a.save({name:"yaya",age:75,em:""})
WriteResult({ "nInserted" : 1 })

52上面查看

[ ~]# mongo --host 192.168.4.52 --port 27078
rs1:SECONDARY> db.getMongo().setSlaveOk()       //允许从库查看数据
rs1:SECONDARY> show dbs        //有gamedb2库
admin    0.000GB
config   0.000GB
ddsdb    0.000GB
gamedb2  0.000GB
local    0.000GB
test     0.000GB

步骤三:切换主库验证

1)自动切换主库验证
关闭51

[ ~]# cmdb        //之前设置的别名
killing process with pid: 5656

查看52和53

[ ~]# mongo --host 192.168.4.52 --port 27078
MongoDB shell version v3.6.3
connecting to: mongodb://192.168.4.52:27078/
MongoDB server version: 3.6.3
...
... 
rs1:PRIMARY>         //52为主
rs1:PRIMARY>

[ ~]# mongo --host 192.168.4.53 --port 27079
MongoDB shell version v3.6.3
connecting to: mongodb://192.168.4.53:27079/
MongoDB server version: 3.6.3
...
... 
rs1:SECONDARY>        //53为从

启动51,启动后不会再变为主,会成为52的从

[ ~]# smdb
about to fork child process, waiting until server is ready for connections.
forked process: 6598
child process started successfully, parent exiting
rs1:SECONDARY> rs.isMaster()
{
    "hosts" : [
        "192.168.4.51:27077",
        "192.168.4.52:27078",
        "192.168.4.53:27079"
    ],
    "setName" : "rs1",
    "setVersion" : 1,
    "ismaster" : false,
    "secondary" : true,
    "primary" : "192.168.4.52:27078",
    "me" : "192.168.4.51:27077",
...

==================================
基于MongoDB环境完成下列练习:
插入文档
查询文档
更新文档
删除文档

步骤一:管理文档

1)把系统用户信息/etc/passwd存储到mdb库下的user集合里

rs1:PRIMARY> use mdb
switched to db mdb
rs1:PRIMARY> db.user.save({name:"yaya",password:"x",uid:9999,gid:9999,comment:"",homdir:"/home",shell:"/bin/bash"})
WriteResult({ "nInserted" : 1 })
rs1:PRIMARY> exit
bye
[ ~]#  mongoexport --host 192.168.4.52 --port 27078 -d mdb -c user -f name,password,uid,gid,comment,homdir,shell  --type=csv -o /tmp/user.csv
2018-09-29T11:04:14.967+0800    connected to: 192.168.4.52:27078
2018-09-29T11:04:14.968+0800    exported 1 record
[ ~]#  cp /etc/passwd /tmp
[ ~]# sed -i ‘s/:/,/g‘   /tmp/passwd
[ ~]#  sed -i ‘$r /tmp/passwd‘ /tmp/user.csv
[ ~]# mongoimport --host 192.168.4.52 --port 27078 -d mdb -c user --headerline  --type=csv  /tmp/user.csv
2018-09-29T11:06:08.355+0800    connected to: 192.168.4.52:27078
2018-09-29T11:06:08.363+0800    imported 41 documents

2)查看

[ ~]# mongo --host 192.168.4.52 --port 27078
rs1:PRIMARY> use mdb
switched to db mdb
rs1:PRIMARY> db.user.findOne()
{
    "_id" : ObjectId("5baeeb37ce3cc5539aa21f38"),
    "name" : "yaya",
    "password" : "x",
    "uid" : 9999,
    "gid" : 9999,
    "comment" : "",
    "homdir" : "/home",
    "shell" : "/bin/bash"
}

db.user.find({条件},{定义显示的字段}) #指定查询条件并指定显示的字段
0不显示 1显示

rs1:PRIMARY> db.user.find()
{ "_id" : ObjectId("5baeeb37ce3cc5539aa21f38"), "name" : "yaya", "password" : "x", "uid" : 9999, "gid" : 9999, "comment" : "", "homdir" : "/home", "shell" : "/bin/bash" }
...
Type "it" for more            //出现这个按it,默认出现20行

查看每行的name字段

rs1:PRIMARY>  db.user.find({},{name:1})        
{ "_id" : ObjectId("5baeeb37ce3cc5539aa21f38"), "name" : "yaya" }
...
...
{ "_id" : ObjectId("5baeec2001805180a1011843"), "name" : "rpc" }
Type "it" for more

不看_id字段

rs1:PRIMARY> db.user.find({},{_id:0})
{ "name" : "yaya", "password" : "x", "uid" : 9999, "gid" : 9999, "comment" : "", "homdir" : "/home", "shell" : "/bin/bash" }
...
...
{ "name" : "rpc", "password" : "x", "uid" : 32, "gid" : 32, "comment" : "Rpcbind Daemon", "homdir" : "/var/lib/rpcbind", "shell" : "/sbin/nologin" }
Type "it" for more

不看_id那一列,看name那一列

rs1:PRIMARY> db.user.find({},{_id:0,name:1})
{ "name" : "yaya" }
...
...
{ "name" : "rpc" }
Type "it" for more

查看以a开头的name字段

rs1:PRIMARY> db.user.find({name:/^a/},{_id:0})
{ "name" : "adm", "password" : "x", "uid" : 3, "gid" : 4, "comment" : "adm", "homdir" : "/var/adm", "shell" : "/sbin/nologin" }
{ "name" : "abrt", "password" : "x", "uid" : 173, "gid" : 173, "comment" : "", "homdir" : "/etc/abrt", "shell" : "/sbin/nologin" }
{ "name" : "avahi", "password" : "x", "uid" : 70, "gid" : 70, "comment" : "Avahi mDNS/DNS-SD Stack", "homdir" : "/var/run/avahi-daemon", "shell" : "/sbin/nologin" }

显示查询结果的前一行
limit 数字

rs1:PRIMARY>  db.user.find({name:/^a/},{_id:0}).limit (1)
{ "name" : "adm", "password" : "x", "uid" : 3, "gid" : 4, "comment" : "adm", "homdir" : "/var/adm", "shell" : "/sbin/nologin" }

显示name字段以a开头的第一行

rs1:PRIMARY> db.user.findOne({name:/^a/},{_id:0,name:1,shell:1,uid:1})
{ "name" : "adm", "uid" : 3, "shell" : "/sbin/nologin" }

跳过几行显示 (2行)
skip 数字

rs1:PRIMARY> db.user.find({name:/^a/},{_id:0,name:1,shell:1}).skip (2)
{ "name" : "avahi", "shell" : "/sbin/nologin" }

默认升序排序
sort 字段名

rs1:PRIMARY> db.user.find({name:/^a/},{_id:0,name:1,shell:1,uid:1}).sort({uid:1})
{ "name" : "adm", "uid" : 3, "shell" : "/sbin/nologin" }
{ "name" : "avahi", "uid" : 70, "shell" : "/sbin/nologin" }
{ "name" : "abrt", "uid" : 173, "shell" : "/sbin/nologin" }

降序排序

rs1:PRIMARY> db.user.find({name:/^a/},{_id:0,name:1,shell:1,uid:1}).sort({uid:-1})
{ "name" : "abrt", "uid" : 173, "shell" : "/sbin/nologin" }
{ "name" : "avahi", "uid" : 70, "shell" : "/sbin/nologin" }
{ "name" : "adm", "uid" : 3, "shell" : "/sbin/nologin" }

显示name字段以a开头和uid为3的所有行

rs1:PRIMARY> db.user.find({name:/^a/,uid:3},{_id:0,name:1,shell:1,uid:1})
{ "name" : "adm", "uid" : 3, "shell" : "/sbin/nologin" }

3)条件判断的表示方式

$in 在…里

rs1:PRIMARY> db.user.find({uid:{$in:[1,6,9]}})        //uid的为1或者6或者9的匹配

{ "_id" : ObjectId("5baeec2001805180a1011833"), "name" : "bin", "password" : "x", "uid" : 1, "gid" : 1, "comment" : "bin", "homdir" : "/bin", "shell" : "/sbin/nologin" }
{ "_id" : ObjectId("5baeec2001805180a1011838"), "name" : "shutdown", "password" : "x", "uid" : 6, "gid" : 0, "comment" : "shutdown", "homdir" : "/sbin", "shell" : "/sbin/shutdown" }

$nin 不在…里

rs1:PRIMARY> db.user.find({uid:{$nin:[1,6,9]}},{_id:0,name:1,uid:1})
{ "name" : "yaya", "uid" : 9999 }
...
...
{ "name" : "saslauth", "uid" : 996 }
Type "it" for more

$or 条件满足任意其中一个即可

rs1:PRIMARY> db.user.find({$or:[{name:"root"},{uid:1}]},{_id:0,name:1,uid:1})
{ "name" : "root", "uid" : 0 }
{ "name" : "bin", "uid" : 1 }

4)正则匹配,以a开头的name字段

rs1:PRIMARY> db.user.find({name:/^a/},{_id:0,name:1,uid:1})
{ "name" : "adm", "uid" : 3 }
{ "name" : "abrt", "uid" : 173 }
{ "name" : "avahi", "uid" : 70 }

objectId对象类型 使用12字节的存储空间每个字节2位十六进制数字共计24位字符串.

0 1 2 3 4 5 6 7 8 9 10 11
时间戳 机器 PID 计数器

5)数值比较

$lt(小于) $lte(小于等于) $gt(大于) $gte(大于等于) $ne(不等于)

rs1:PRIMARY> db.user.find({uid:{$gte:10,$lte:40}},{_id:0,name:1,uid:1})
{ "name" : "operator", "uid" : 11 }
{ "name" : "games", "uid" : 12 }
{ "name" : "ftp", "uid" : 14 }
{ "name" : "rpc", "uid" : 32 }
{ "name" : "rpcuser", "uid" : 29 }
{ "name" : "ntp", "uid" : 38 }

匹配null:可以匹配没有的字段,也可以检查这个字段有没有

rs1:PRIMARY> db.user.save({name:null,uid:null})
WriteResult({ "nInserted" : 1 })

rs1:PRIMARY> db.user.find({name:null})
{ "_id" : ObjectId("5baef385f9f3bf625ea1dbd6"), "name" : null, "uid" : null }

rs1:PRIMARY>  db.user.find({shell:null})    //表示此条文档没有shell字段
{ "_id" : ObjectId("5baef385f9f3bf625ea1dbd6"), "name" : null, "uid" : null }

6)save和insert的区别

相同点:当集合不存在时创建集合,并插入记录
不同点:

save() _id字段值已经存在时,修改文档字段值

insert() _id字段值已经存在时,放弃修改文档字段值

rs1:PRIMARY> db.t1.save({name:"bob",age:19})

rs1:PRIMARY> db.t1.insert({name:"bob",age:19})

rs1:PRIMARY> db.t1.save({_id:7,name:"bob",age:19})
WriteResult({ "nMatched" : 0, "nUpserted" : 1, "nModified" : 0, "_id" : 7 })

rs1:PRIMARY>  db.t1.find()
...
...
{ "_id" : 7, "name" : "bob", "age" : 19 }

rs1:PRIMARY> db.t1.save({_id:7,name:"tom",age:19})        //把上一条的记录直接修改

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

rs1:PRIMARY> db.t1.find()
...
...
{ "_id" : 7, "name" : "tom", "age" : 19 }

rs1:PRIMARY> db.t1.insert({_id:8,name:"tom",age:19})        //可以存上
WriteResult({ "nInserted" : 1 })

rs1:PRIMARY> db.t1.insert({_id:8,name:"tom",age:19})        //存不上
WriteResult({
    "nInserted" : 0,
    "writeError" : {
        "code" : 11000,
        "errmsg" : "E11000 duplicate key error collection: mdb.t1 index: _id_ dup key: { : 8.0 }"
    }
})

7)插入多行文档

rs1:PRIMARY> db.t1.insertMany([{name:"xiaojiu"},{name:"laoshi"}])
{
    "acknowledged" : true,
    "insertedIds" : [
        ObjectId("5baef526f9f3bf625ea1dbd9"),
        ObjectId("5baef526f9f3bf625ea1dbda")
    ]
}

rs1:PRIMARY> db.t1.find()
...
...
{ "_id" : ObjectId("5baef526f9f3bf625ea1dbd9"), "name" : "xiaojiu" }
{ "_id" : ObjectId("5baef526f9f3bf625ea1dbda"), "name" : "laoshi" }

8)update修改

>db.user.find.update({条件},{修改的字段})

rs1:PRIMARY> db.user.update({name:"root"},{password:"XXX"})  

//如果这一列不写完整,这一行除了password这一行,这一列的其他值都没有了

相当于删除(要写完整)

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

rs1:PRIMARY> db.t1.find({name:"root"})
rs1:PRIMARY> db.user.find({name:"root"})   //没有东西,除了password:"XXX"

9)$set 条件匹配时,修改指定字段的值(局部修改)

rs1:PRIMARY> db.user.update({name:"adm"},{$set:{password:"AAA"}})
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

rs1:PRIMARY> db.user.find({name:"adm"})        //还存在
{ "_id" : ObjectId("5baeec2001805180a1011835"), "name" : "adm", "password" : "AAA", "uid" : 3, "gid" : 4, "comment" : "adm", "homdir" : "/var/adm", "shell" : "/sbin/nologin" }

rs1:PRIMARY> db.user.update({name:/^r/},{$set:{password:"FFF"}})        
//默认修改匹配条件的第一行

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })    

rs1:PRIMARY> db.user.update({name:/^a/},{$set:{password:"FFF"}},false,true)

//改匹配到的所有

WriteResult({ "nMatched" : 3, "nUpserted" : 0, "nModified" : 3 })

10)$unset 删除与条件匹配文档的字段

rs1:PRIMARY> db.user.update({name:"sync"},{$unset:{password:1}})     

//删除password字段
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

11)数组

rs1:PRIMARY> db.user.insert({name:"bob",like:["a","b","c","d","e","f",]})
WriteResult({ "nInserted" : 1 })

$pop

删除数组末尾一个元素,1删除最后一个,-1删除第一个

rs1:PRIMARY> db.user.update({name:"bob"},{$pop:{like:1}})    

//删除匹配的第一条的最后一个

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

rs1:PRIMARY> db.user.update({name:"bob"},{$pop:{like:-1}})    

//删除匹配的第一条的第一个

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

$push 向数组中添加新元素

rs1:PRIMARY> db.user.update({name:"bob"},{$push:{like:"Z"}})        //默认添加到最后

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

rs1:PRIMARY> db.user.update({name:"bob"},{$push:{like:"W"}})
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

rs1:PRIMARY> db.user.find({name:"bob"})
{ "_id" : ObjectId("5baef7b2034891a205de2959"), "name" : "bob", "like" : [ "b", "c", "d", "e", "Z", "W" ] }

$addToSet 避免重复添加

rs1:PRIMARY> db.user.update({name:"bob"},{$addToSet:{like:"W"}})

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 0 })

rs1:PRIMARY> db.user.find({name:"bob"})
{ "_id" : ObjectId("5baef7b2034891a205de2959"), "name" : "bob", "like" : [ "b", "c", "d", "e", "Z", "W" ] }

$pull 删除数组里的指定元素,若有两个bob可以用_id值定义把name:"bob"换成id值

rs1:PRIMARY> db.user.update({name:"bob"},{$pull:{like:"c"}})
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

rs1:PRIMARY> db.user.find({name:"bob"})
{ "_id" : ObjectId("5baef7b2034891a205de2959"), "name" : "bob", "like" : [ "b", "d", "e", "Z", "W" ] }

rs1:PRIMARY> db.user.update({"_id":ObjectId("5afc1a717eff45e9cfc57ed3")},{$push:{like:"S"}})
WriteResult({ "nMatched" : 0, "nUpserted" : 0, "nModified" : 0 })

rs1:PRIMARY>

$inc 条件匹配时,字段值自加或自减

rs1:PRIMARY> db.user.update({uid:{$lte:10}},{$inc:{uid:2}})         

//设置字段值自加2,默认改第一行

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })   

rs1:PRIMARY> db.user.update({uid:{$lte:10}},{$inc:{uid:2}},false,true)    

//设置字段值自加2,false,true改全部

WriteResult({ "nMatched" : 8, "nUpserted" : 0, "nModified" : 8 })   

rs1:PRIMARY> db.user.update({uid:{$lte:10}},{$inc:{uid:-1}})    

//负数时是自减1,默认改第一行

WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })

12)删除文档
remove()与drop()的区别

remove()删除文档时不删除索引

drop()删除集合的时候同时删除索引

rs1:PRIMARY> db.t1.remove({})
WriteResult({ "nRemoved" : 6 })

rs1:PRIMARY> db.user.remove({name:"/^a/"})        //删除以a开头的记录
WriteResult({ "nRemoved" : 0 })

rs1:PRIMARY> db.t1.drop()        //删除集合t1
true

相关推荐