Mongodb 5节点异地两中心故障转移恢复测试案例

Mongodb5节点异地两中心故障转移恢复测试案例

架构方式：5节点，主中心（2数据1仲裁），备中心（1数据1仲裁）

1基本情况

操作系统:Red Hat Enterprise Linux Server release 6.3 (Santiago)

Mongodb版本：db version v3.6.3

Mongodb架构：

Ip，端口规划

"hosts" : [##数据节点

"*114:28001",#主中心

"*114:28002",#主中心

"*114:28004"#备份中心

],

"arbiters" : [##仲裁节点

"*114:28003",#主中心

"*114:28005"#备份中心

],

2 mongodb

配置文件，其他配置文件28001替换为28002~28007

注意相应的data，log等数据目录要存在，记住，所有的mongodb里面执行的命令，都要有返回ok为1才成功。

[root@mysqlt1 ~]# cat /data/mongodb/conf/28001.conf

port=28001

bind_ip=*

logpath=/data/mongodb/log/28001.log

dbpath=/data/mongodb/data/28001/

logappend=true

pidfilepath=/data/mongodb/28001.pid

fork=true

oplogSize=1024

replSet=MyMongo

[root@mysqlt1 conf]# ll /data/mongodb/conf/

total 32

-rw-r--r-- 1 root root 192 Oct 16 02:48 28001.conf

-rw-r--r-- 1 root root 225 Oct 16 07:21 28002.conf

-rw-r--r-- 1 root root 192 Oct 16 02:48 28003.conf

-rw-r--r-- 1 root root 192 Oct 11 03:37 28004.conf

-rw-r--r-- 1 root root 192 Oct 11 03:38 28005.conf

-rw-r--r-- 1 root root 192 Oct 16 07:18 28006.conf

-rw-r--r-- 1 root root 192 Oct 16 08:15 28007.conf

启动5个节点

[root@mysqlt1 data]# /usr/local/mongodb/bin/mongod -f /data/mongodb/conf/28001.conf

[root@mysqlt1 data]# /usr/local/mongodb/bin/mongod -f /data/mongodb/conf/28002.conf

[root@mysqlt1 data]# /usr/local/mongodb/bin/mongod -f /data/mongodb/conf/28003.conf

[root@mysqlt1 data]# /usr/local/mongodb/bin/mongod -f /data/mongodb/conf/28004.conf

[root@mysqlt1 data]# /usr/local/mongodb/bin/mongod -f /data/mongodb/conf/28005.conf

3 测试主中心全部宕机的情况

[root@mysqlt1 ~]# /usr/local/mongodb/bin/mongod  --shutdown -f /data/mongodb/conf/28001.conf

[root@mysqlt1 ~]# /usr/local/mongodb/bin/mongod  --shutdown -f /data/mongodb/conf/28002.conf

[root@mysqlt1 ~]# /usr/local/mongodb/bin/mongod  --shutdown -f /data/mongodb/conf/28003.conf

只剩备份中心的2个节点，这时候变为secondry，备份中心只能读取，不能写入

两种解决方案
1 把备份中心的secondry节点，作为单节点，去掉参数replSet重新启动，可以继续使用，但是由于单节点缺少oplog，后面主中心恢复，备份中心的数据不能恢复到整个副本集中，可以考虑备份方式（复杂，这里还有第二种方式）。

2 在备份中心，启动2个新的仲裁节点，强制加入副本集，使secondry节点变为primary节点，详细的操作方式

一：启动备份中心的2个新节点（28006,28007）

二：在备份中心的secondry节点，重新配置副本集，加入2个仲裁节点

MyMongo:SECONDARY> use admin

MyMongo:SECONDARY> config = {

"_id":"MyMongo",

members:[

{"_id":0,host:"*:28001"},

{"_id":1,host:"*:28002"},

{"_id":2,host:"*:28003", arbiterOnly: true},

{"_id":3,host:"*:28004"},

{"_id":4,host:"*:28005", arbiterOnly: true},

{"_id":5,host:"*:28006", arbiterOnly: true},

{"_id":6,host:"*:28007", arbiterOnly: true}]

}

MyMongo:SECONDARY> rs.reconfig(config,{force:true});

MyMongo:PRIMARY> rs.status() #查看副本集的状态，及各节点的状态

MyMongo:PRIMARY> db.isMaster()

在client端批量插入数据（简单的程序），这里可以配置集群方式连接，也可以指定主节点的方式进行插入，这里是直接指定主节点

#coding:utf-8
import time
from pymongo import MongoClient
#conn = MongoClient()
# keyword argument
conn = MongoClient('*', 28004)
# MongoDB URI
#conn = MongoClient('mongodb://localhost:27017/')
#from pymongo import ReplicaSetConnection
#conn = ReplicaSetConnection("*:28001,*:28002,*:28004", replicaSet="MyMongo", read_preference=2, safe=True)

for i in xrange(100):
    try:
        conn.test.tt3.insert({"name":"test" + str(i)})
        time.sleep(1)
        print conn.primary
        print conn.secondaries
    except:
        pass

主节点执行，100条

MyMongo:PRIMARY> db.tt3.find().count()

100

启动主中心的3个节点

[root@mysqlt1 conf]# /usr/local/mongodb/bin/mongod -f /data/mongodb/conf/28001.conf

[root@mysqlt1 conf]# /usr/local/mongodb/bin/mongod -f /data/mongodb/conf/28002.conf

[root@mysqlt1 conf]# /usr/local/mongodb/bin/mongod -f /data/mongodb/conf/28003.conf

[root@mysqlt1 ~]#  /usr/local/mongodb/bin/mongo *:28002/admin

MyMongo:SECONDARY> rs.slaveOk(true)

MyMongo:SECONDARY> use test;

switched to db test

MyMongo:SECONDARY> db.tt3.find().count() #数据同步成功

100

之前的5个节点，现在变成了7个节点，删除新加的2个仲裁节点

MyMongo:PRIMARY> rs.remove("*:28007");

MyMongo:PRIMARY> rs.remove("*:28006");

MyMongo:PRIMARY> db.isMaster() #变回之前的5个节点，1主，2secondry，2仲裁

{

"hosts" : [

"*:28001",

"*:28002",

"*:28004"

],

"arbiters" : [

"*:28003",

"*:28005"

MyMongo:PRIMARY> rs.status()

{

"set" : "MyMongo",

"date" : ISODate("2018-10-16T18:16:15.512Z"),

"myState" : 1,

"term" : NumberLong(7),

"heartbeatIntervalMillis" : NumberLong(2000),

"optimes" : {

"lastCommittedOpTime" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

},

"readConcernMajorityOpTime" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

},

"appliedOpTime" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

},

"durableOpTime" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

}

},

"members" : [

{

"_id" : 0,

"name" : "*:28001",

"health" : 1,

"state" : 2,

"stateStr" : "SECONDARY",

"uptime" : 237,

"optime" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

},

"optimeDurable" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

},

"optimeDate" : ISODate("2018-10-16T18:16:06Z"),

"optimeDurableDate" : ISODate("2018-10-16T18:16:06Z"),

"lastHeartbeat" : ISODate("2018-10-16T18:16:13.929Z"),

"lastHeartbeatRecv" : ISODate("2018-10-16T18:16:14.928Z"),

"pingMs" : NumberLong(0),

"syncingTo" : "*:28004",

"configVersion" : 102086

},

{

"_id" : 1,

"name" : "*:28002",

"health" : 1,

"state" : 2,

"stateStr" : "SECONDARY",

"uptime" : 269,

"optime" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

},

"optimeDurable" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

},

"optimeDate" : ISODate("2018-10-16T18:16:06Z"),

"optimeDurableDate" : ISODate("2018-10-16T18:16:06Z"),

"lastHeartbeat" : ISODate("2018-10-16T18:16:13.929Z"),

"lastHeartbeatRecv" : ISODate("2018-10-16T18:16:14.928Z"),

"pingMs" : NumberLong(0),

"syncingTo" : "*:28004",

"configVersion" : 102086

},

{

"_id" : 2,

"name" : "*:28003",

"health" : 1,

"state" : 7,

"stateStr" : "ARBITER",

"uptime" : 193,

"lastHeartbeat" : ISODate("2018-10-16T18:16:13.929Z"),

"lastHeartbeatRecv" : ISODate("2018-10-16T18:16:11.917Z"),

"pingMs" : NumberLong(0),

"configVersion" : 102086

},

{

"_id" : 3,

"name" : "*:28004",

"health" : 1,

"state" : 1,

"stateStr" : "PRIMARY",

"uptime" : 68054,

"optime" : {

"ts" : Timestamp(1539713766, 1),

"t" : NumberLong(7)

},

"optimeDate" : ISODate("2018-10-16T18:16:06Z"),

"electionTime" : Timestamp(1539712874, 1),

"electionDate" : ISODate("2018-10-16T18:01:14Z"),

"configVersion" : 102086,

"self" : true

},

{

"_id" : 4,

"name" : "*:28005",

"health" : 1,

"state" : 7,

"stateStr" : "ARBITER",

"uptime" : 66987,

"lastHeartbeat" : ISODate("2018-10-16T18:16:13.929Z"),

"lastHeartbeatRecv" : ISODate("2018-10-16T18:16:11.921Z"),

"pingMs" : NumberLong(0),

"configVersion" : 102086

}

],

"ok" : 1,

"operationTime" : Timestamp(1539713766, 1),

"$clusterTime" : {

"clusterTime" : Timestamp(1539713766, 1),

"signature" : {

"hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),

"keyId" : NumberLong(0)

}

}

}