MongoDB Sharding(二) - 搭建分片集群 - Powered by Discuz! Archiver

吉刚發表於 2021-1-16 14:08:00

MongoDB Sharding(二) -- 搭建分片集群

<blockquote>
在上一篇文章中，我们基本了解了分片的概念，本文将着手实践，进行分片集群的搭建
</blockquote>
 
首先我们再来了解一下分片集群的架构，分片集群由三部分构成：
<ul>
<li>mongos：查询路由，在客户端程序和分片之间提供接口。本次实验部署2个mongos实例</li>
<li>config：配置服务器存储集群的元数据，元数据反映分片集群的内所有数据和组件的状态和组织方式，元数据包含每个分片上的块列表以及定义块的范围。从3.4版本开始，已弃用镜像服务器用作配置服务器(SCCC)，config Server必须部署为副本集架构（CSRS）。本次实验配置一个3节点的副本集作为配置服务器</li>
<li>shard：每个shard包含集合的一部分数据，从3.6版本开始，每个shard必须部署为副本集（replica set）架构。本次实验部署3个分片存储数据。</li>
</ul>
<img src="https://img2020.cnblogs.com/blog/823295/202101/823295-20210116130529829-537460884.png" alt="" loading="lazy">
 
（一）主机信息
<img src="https://img2020.cnblogs.com/blog/823295/202101/823295-20210116131040587-603187584.png" alt="" loading="lazy">
 
（二）配置服务器副本集搭建
配置服务器三个实例的基础规划如下：
member0 192.168.10.80:27017 member1 192.168.10.80:27018 member2 192.168.10.80:27019
其参数规划如下：
<img src="https://img2020.cnblogs.com/blog/823295/202101/823295-20210116131453668-1363832161.png" alt="" loading="lazy">
 
 接下来，我们一步一步搭建config server的副本集。
STEP1：解压mongodb安装包到/mongo目录
<div class="cnblogs_code">
<pre># pwd
/mongo
# ls
bin LICENSE-Community.txt MPL-2 README THIRD-PARTY-NOTICES THIRD-PARTY-NOTICES.gotools</pre>
</div>
 STEP2：根据上面参数规划，创建数据存放相关路径
<div class="cnblogs_code">
<pre># 创建文件路径
mkdir -p /replset/repset1/data
mkdir -p /replset/repset1/log
mkdir -p /replset/repset2/data
mkdir -p /replset/repset2/log
mkdir -p /replset/repset3/data
mkdir -p /replset/repset3/log

# tree /replset/
/replset/
├── repset1
│   ├── data
│   ├── log
│   └── mongodb.conf
├── repset2
│   ├── data
│   └── log
└── repset3
├── data
└── log</pre>
</div>
 STEP3：为3个实例创建参数文件
实例1的参数文件  /replset/repset1/mongodb.conf  ：
<div class="cnblogs_code"><img src="https://images.cnblogs.com/OutliningIndicators/ContractedBlock.gif" id="code_img_closed_9361e884-a66e-4ed9-9e8a-895c7f023fe0" class="code_img_closed"><img src="https://images.cnblogs.com/OutliningIndicators/ExpandedBlockStart.gif" id="code_img_opened_9361e884-a66e-4ed9-9e8a-895c7f023fe0" class="code_img_opened" style="display: none">
<div id="cnblogs_code_open_9361e884-a66e-4ed9-9e8a-895c7f023fe0" class="cnblogs_code_hide">
<pre>systemLog:
destination: file
logAppend: true
path: /replset/repset1/log/mongodb.log

storage:
dbPath: /replset/repset1/data
journal:
enabled: true

processManagement:
fork: true# fork and run in background
pidFilePath: /replset/repset1/mongod.pid# location of pidfile
timeZoneInfo: /usr/share/zoneinfo

# network interfaces
net:
port: 27017
bindIp: 0.0.0.0

# shard
sharding:
clusterRole: configsvr

# repliuca set
replication:
replSetName: conf</pre>
</div>
View Code</div>
 实例2的参数文件  /replset/repset2/mongodb.conf ：
<div class="cnblogs_code"><img src="https://images.cnblogs.com/OutliningIndicators/ContractedBlock.gif" id="code_img_closed_b858fc3b-90e1-436b-b2fe-d3e41c18ecaa" class="code_img_closed"><img src="https://images.cnblogs.com/OutliningIndicators/ExpandedBlockStart.gif" id="code_img_opened_b858fc3b-90e1-436b-b2fe-d3e41c18ecaa" class="code_img_opened" style="display: none">
<div id="cnblogs_code_open_b858fc3b-90e1-436b-b2fe-d3e41c18ecaa" class="cnblogs_code_hide">
<pre>systemLog:
destination: file
logAppend: true
path: /replset/repset2/log/mongodb.log

storage:
dbPath: /replset/repset2/data
 journal:
enabled: true

processManagement:
fork: true# fork and run in background
pidFilePath: /replset/repset2/mongod.pid# location of pidfile
timeZoneInfo: /usr/share/zoneinfo

# network interfaces
net:
port: 27018
bindIp: 0.0.0.0

# shard
sharding:
clusterRole: configsvr

# repliuca set
replication:
replSetName: conf
</pre>
</div>
View Code</div>
 实例3的参数文件  /replset/repset3/mongodb.conf ：
<div class="cnblogs_code"><img src="https://images.cnblogs.com/OutliningIndicators/ContractedBlock.gif" id="code_img_closed_88bde8b5-c91f-42bd-8e0f-afaed2752caa" class="code_img_closed"><img src="https://images.cnblogs.com/OutliningIndicators/ExpandedBlockStart.gif" id="code_img_opened_88bde8b5-c91f-42bd-8e0f-afaed2752caa" class="code_img_opened" style="display: none">
<div id="cnblogs_code_open_88bde8b5-c91f-42bd-8e0f-afaed2752caa" class="cnblogs_code_hide">
<pre>systemLog:
destination: file
logAppend: true
path: /replset/repset3/log/mongodb.log

storage:
dbPath: /replset/repset3/data
journal:
enabled: true

processManagement:
fork: true# fork and run in background
pidFilePath: /replset/repset3/mongod.pid# location of pidfile
timeZoneInfo: /usr/share/zoneinfo

# network interfaces
net:
port: 27019
bindIp: 0.0.0.0

# shard
sharding:
clusterRole: configsvr

# repliuca set
replication:
replSetName: conf</pre>
</div>
View Code</div>
 STEP4：启动三个mongod实例
<div class="cnblogs_code">
<pre>mongod -f /replset/repset1/mongodb.conf
mongod -f /replset/repset2/mongodb.conf
mongod -f /replset/repset3/mongodb.conf

# 查看是成功否启动
# netstat -nltp |grep mongod
tcp 0 0 0.0.0.0:27019 0.0.0.0:* LISTEN 28009/mongod
tcp 0 0 0.0.0.0:27017 0.0.0.0:* LISTEN 27928/mongod
tcp 0 0 0.0.0.0:27018 0.0.0.0:* LISTEN 27970/mongod</pre>
</div>
 STEP5：进入任意一个实例，初始化配置服务器的副本集
<div class="cnblogs_code">
<pre>rs.initiate(
{
_id: "conf",
configsvr: true,
members: [
 { _id : 0, host : "192.168.10.80:27017" },
 { _id : 1, host : "192.168.10.80:27018" },
 { _id : 2, host : "192.168.10.80:27019" }
]
}
)</pre>
</div>
  STEP6：[可选] 调整节点优先级，以便于确定主节点
<div class="cnblogs_code">
<pre>cfg = rs.conf()
cfg.members[0].priority = 3
cfg.members[1].priority = 2
cfg.members[2].priority = 1
rs.reconfig(cfg)</pre>
</div>
 对于members的定义：n是members数组中的数组位置，数组以0开始，千万不能将其理解为“members._id”的_id值。
查看节点优先级：
<div class="cnblogs_code">
<pre>conf:PRIMARY> rs.config()</pre>
</div>
 
（三）分片副本集搭建
分片1副本集成员： member0 192.168.10.81:27017 member1 192.168.10.81:27018 member2 192.168.10.81:27019
分片2副本集成员： member0 192.168.10.82:27017 member1 192.168.10.82:27018 member2 192.168.10.82:27019
分片3副本集成员： member0 192.168.10.83:27017 member1 192.168.10.83:27018 member2 192.168.10.83:27019
 
其参数规划如下：
<img src="https://img2020.cnblogs.com/blog/823295/202101/823295-20210116132649817-573510342.png" alt="" loading="lazy">
 这里一共有3个分片，每个分片都是3个节点的副本集，副本集的搭建过程与上面config server副本集搭建过程相似，这里不再重复赘述，唯一不同的是副本集的初始化。shard副本集的初始化与配置副本集初始化过程相比，少了 configsvr: true 的参数配置。
三个shard副本集的初始化：
<div class="cnblogs_code">
<pre># shard001 rs.initiate(
{
_id: "shard001",
members: [
 { _id : 0, host : "192.168.10.81:27017" },
 { _id : 1, host : "192.168.10.81:27018" },
 { _id : 2, host : "192.168.10.81:27019" }
]
}
)
</pre>
<pre># shard002 rs.initiate(
{
_id: "shard002",
members: [
 { _id : 0, host : "192.168.10.82:27017" },
 { _id : 1, host : "192.168.10.82:27018" },
 { _id : 2, host : "192.168.10.82:27019" }
]
}
)

# shard003
rs.initiate(
{
_id: "shard003",
members: [
 { _id : 0, host : "192.168.10.83:27017" },
 { _id : 1, host : "192.168.10.83:27018" },
 { _id : 2, host : "192.168.10.83:27019" }
]
}
)</pre>
</div>
 
（四）配置并启动mongos
本次试验在192.168.10.100服务器上启动2个mongos进程，分别使用端口27000和28000。
STEP1：配置mongos实例的参数
端口27000参数配置，特别注意，需要先创建涉及到的路径：
<div class="cnblogs_code">
<pre>systemLog:
destination: file
logAppend: true
path: /mongo/log/mongos-27000.log

processManagement:
fork: true# fork and run in background
pidFilePath: /mongo/mongod-27000.pid# location of pidfile
timeZoneInfo: /usr/share/zoneinfo

# network interfaces
net:
port: 27000
bindIp: 0.0.0.0

sharding:
configDB: conf/192.168.10.80:27017,192.168.10.80:27018,192.168.10.80:27019</pre>
</div>
 端口28000参数配置，特别注意，需要先创建涉及到的路径：
<div class="cnblogs_code">
<pre>systemLog:
destination: file
logAppend: true
path: /mongo/log/mongos-28000.log

processManagement:
fork: true# fork and run in background
pidFilePath: /mongo/mongod-28000.pid# location of pidfile
timeZoneInfo: /usr/share/zoneinfo

# network interfaces
net:
port: 28000
bindIp: 0.0.0.0

sharding:
configDB: conf/192.168.10.80:27017,192.168.10.80:27018,192.168.10.80:27019</pre>
</div>
 STEP2：启动mongos实例
<div class="cnblogs_code">
<pre># 启动mongos实例
# mongos -f /mongo/mongos-27000.conf
# mongos -f /mongo/mongos-28000.conf

# 查看实例信息
# netstat -nltp|grep mongos
tcp 0 0 0.0.0.0:27000 0.0.0.0:* LISTEN 2209/mongos
tcp 0 0 0.0.0.0:28000 0.0.0.0:* LISTEN 2241/mongos</pre>
</div>
 
（五）添加分片到集群配置服务器
STEP1：使用mongo连接到mongos
<div class="cnblogs_code">
<pre>mongo --host 192.168.10.100 --port 27000
# 或者
mongo --host 192.168.10.100 --port 28000</pre>
</div>
 STEP2：添加分片到集群
<div class="cnblogs_code">
<pre>sh.addShard( "shard001/192.168.10.81:27017,192.168.10.81:27018,192.168.10.81:27019")
sh.addShard( "shard002/192.168.10.82:27017,192.168.10.82:27018,192.168.10.82:27019")
sh.addShard( "shard003/192.168.10.83:27017,192.168.10.83:27018,192.168.10.83:27019") </pre>
</div>
 STEP3：查看分片信息
<div class="cnblogs_code">
<pre>mongos> sh.status()
--- Sharding Status ---
sharding version: {
 "_id" : 1,
 "minCompatibleVersion" : 5,
 "currentVersion" : 6,
 "clusterId" : ObjectId("5ffc0709b040c53d59c15c66")
}
shards:
 {"_id" : "shard001","host" : "shard001/192.168.10.81:27017,192.168.10.81:27018,192.168.10.81:27019","state" : 1 }
 {"_id" : "shard002","host" : "shard002/192.168.10.82:27017,192.168.10.82:27018,192.168.10.82:27019","state" : 1 }
 {"_id" : "shard003","host" : "shard003/192.168.10.83:27017,192.168.10.83:27018,192.168.10.83:27019","state" : 1 }
active mongoses:
 "4.2.10" : 2
autosplit:
 Currently enabled: yes
balancer:
 Currently enabled:yes
 Currently running:no
 Failed balancer rounds in last 5 attempts:0
 Migration Results for the last 24 hours:
 No recent migrations
databases:
 {"_id" : "config","primary" : "config","partitioned" : true }

mongos> </pre>
</div>
 
 
（六）启用分片
（6.1）对数据库启用分片 分片是以集合为单位进行的，在对一个集合进行分片之前，需要先对其数据库启用分片，对数据库启用分片并不会重新分发数据，只是说明该数据库上的集合可以进行分片操作。
<div class="cnblogs_code">
<pre>sh.enableSharding("lijiamandb");</pre>
</div>
 
（6.2）对集合启用分片
如果集合已经存在数据，必须手动创建在分片键上创建索引，然后再对集合进行分片，如果集合为空，MongoDB会在分片的时候自动在分片键上创建索引。 mongodb提供了2种策略来对集合进行分片：
<ul>
<li>哈希(hash)分片，对单列使用hash索引作为分片键</li>

</ul>
<div class="cnblogs_code">
<pre>sh.shardCollection("<database>.<collection>",{shard key field : "hashed"}) </pre>
</div>
<ul>
<li>范围(range)分片，可以使用多个字段作为分片键，并将数据划分为由分片键确定的连续范围</li>
</ul>
<div class="cnblogs_code">
<pre>sh.shardCollection("<database>.<collection>",{<shard key field>:1,...} )</pre>
</div>
 
例子:对集合user进行hash分片
<div class="cnblogs_code">
<pre>// 连接到mongos，进入lijiamandb数据库，对新集合users插入10万条数据
use lijiamandb

for (i=1;i<100000;i++){
db.user.insert({
"id" : i,
"name" : "name"+i,
"age" : Math.floor(Math.random()*120),
"created" : new Date()
});
}

// 使用mongostat可以看到，所有数据都写入到了主节点(shard2)，每个数据库的主节点可能不同，可以使用sh.status()查看。
# mongostat --port 27000 5 --discover
 host insert query update delete getmore command dirty used flushes mapped vsize res faults qrw arw net_in net_out conn set repl time
localhost:27000 352 *0 *0 *0 0 704|0 0 0B356M 32.0M 0 0|0 0|0 224k 140k 10 RTR Jan 15 10:52:32.046

 host insert query update delete getmore command dirtyused flushes mapped vsize res faults qrw arw net_in net_out conn set repl time
192.168.10.81:27017 *0 *0 *0 *0 0 2|00.3%0.8% 0 1.90G133M n/a 0|0 1|0 417b 9.67k 23 shard001SEC Jan 15 10:52:32.061
192.168.10.81:27018 *0 *0 *0 *0 0 3|00.3%0.8% 1 1.93G132M n/a 0|0 1|01.39k 11.0k 28 shard001PRI Jan 15 10:52:32.067
192.168.10.81:27019 *0 *0 *0 *0 0 2|00.3%0.8% 0 1.95G148M n/a 0|0 1|0 942b 10.2k 26 shard001SEC Jan 15 10:52:32.070
192.168.10.82:27017 352 *0 *0 *0 4071192|02.5% 11.7% 1 1.99G180M n/a 0|0 1|01.52m 1.15m 29 shard002PRI Jan 15 10:52:32.075
192.168.10.82:27018 *352 *0 *0 *0 409 441|04.5%8.9% 0 1.96G163M n/a 0|0 1|0 566k 650k 25 shard002SEC Jan 15 10:52:32.085
192.168.10.82:27019 *352 *0 *0 *0 0 2|04.4%9.7% 0 1.92G168M n/a 0|0 1|0 406b 9.51k 24 shard002SEC Jan 15 10:52:32.093
192.168.10.83:27017 *0 *0 *0 *0 0 1|00.2%0.6% 1 1.89G130M n/a 0|0 1|0 342b 9.17k 22 shard003SEC Jan 15 10:52:32.099
192.168.10.83:27018 *0 *0 *0 *0 0 2|00.2%0.6% 0 1.95G139M n/a 0|0 1|0 877b 9.92k 28 shard003PRI Jan 15 10:52:32.107
192.168.10.83:27019 *0 *0 *0 *0 0 1|00.2%0.6% 0 1.90G133M n/a 0|0 1|0 342b 9.17k 21 shard003SEC Jan 15 10:52:32.113
localhost:27000 365 *0 *0 *0 0 731|0 0 0B356M 32.0M 0 0|0 0|0 233k 145k 10 RTR Jan 15 10:52:37.047

// 使用分片键id创建hash分片，因为id上没有hash索引，会报错
sh.shardCollection("lijiamandb.user",{"id":"hashed"})
/* 1 */
{
"ok" : 0.0,
"errmsg" : "Please create an index that starts with the proposed shard key before sharding the collection",
"code" : 72,
"codeName" : "InvalidOptions",
"operationTime" : Timestamp(1610679762, 4),
"$clusterTime" : {
 "clusterTime" : Timestamp(1610679762, 4),
 "signature" : {
 "hash" : { "$binary" : "AAAAAAAAAAAAAAAAAAAAAAAAAAA=", "$type" : "00" },
 "keyId" : NumberLong(0)
 }
}
}

// 需要手动创建hash索引
db.user.ensureIndex()

// 查看索引
/* 1 */
[
{
 "v" : 2,
 "key" : {
 "_id" : 1
 },
 "name" : "_id_",
 "ns" : "lijiamandb.user"
},
{
 "v" : 2,
 "key" : {
 "id" : "hashed"
 },
 "name" : "id_hashed",
 "ns" : "lijiamandb.user"
}
]

# 最后再重新分片即可
sh.shardCollection("lijiamandb".user,{"id":"hashed"})</pre>
</div>
 
到这里，我们分片集群环境已经搭建完成，接下来我们将会学习分片键的选择机制。
 
【完】
 
 
<table style="background-color: rgba(44, 180, 55, 1); height: 133px; width: 778px" border="0" align="left">
<tbody>
<tr>
<td>
相关文档合集：
1. MongoDB Sharding(一) -- 分片的概念 2. MongoDB Sharding(二) -- 搭建分片集群 3. MongoDB Sharding(三) -- zone 4. MongoDB Sharding(四) -- 分片集群的维护管理

</td>

</tr>

</tbody>

</table> 
来源：https://www.cnblogs.com/lijiaman/p/14285489.html

頁: [1]

琼殿技术论坛's Archiver

MongoDB Sharding(二) -- 搭建分片集群