mongodb 集群重构和释放磁盘空间实例详解

gcgl000hugang

2019-04-02

关注关注

MongoDB集群重构，释放磁盘空间

由于mongodb删除了一部分数据后，不会回收相应的磁盘空间，所以这里通过重建数据目录的方式释放磁盘空间。

一实验环境

配置了一个副本集，该副本集由以下三个节点组成：

10.192.203.201:27017 PRIMARY

10.192.203.202:27017 SECONDARY

10.192.203.202:10001 ARBITER

二实验步骤

2.1 模拟环境

use dba; 
for(var i=0;i<1000000;i++)db.c.insert({uid:i,uname:'osqlfan'+i}); 
db.c.find().count();#1000000 
 db.stats(); 
{ 
  "db" : "dba", 
  "collections" : 5, 
  "objects" : 1000111, 
  "avgObjSize" : 111.9994880568257, 
  "dataSize" : 112011920, 
  "storageSize" : 174796800, 
  "numExtents" : 17, 
  "indexes" : 3, 
  "indexSize" : 32475072, 
  "fileSize" : 469762048, 
  "nsSizeMB" : 16, 
  "extentFreeList" : { 
    "num" : 0, 
    "totalSize" : 0 
  }, 
  "dataFileVersion" : { 
    "major" : 4, 
    "minor" : 22 
  }, 
  "ok" : 1 
}

磁盘空间增加了400M数据：

-rw-------. 1 root root 134217728 Nov 7 13:38 dba.1

-rw-------. 1 root root 268435456 Nov 7 13:38 dba.2

[root@slave2 ~]# du -sh /data/mongo/data

4.7G /data/mongo/data

#删除dba.c表数据: 
MyReplset:PRIMARY> db.c.drop(); 
true 
MyReplset:PRIMARY> db.c.find().count(); 
0 
MyReplset:PRIMARY> db.stats(); 
{ 
"db" : "dba", 
"collections" : 4, 
"objects" : 108, 
"avgObjSize" : 108.44444444444444, 
"dataSize" : 11712, 
"storageSize" : 61440, 
"numExtents" : 5, 
"indexes" : 2, 
"indexSize" : 16352, 
"fileSize" : 469762048, 
"nsSizeMB" : 16, 
"extentFreeList" : { 
"num" : 18, 
"totalSize" : 212492288 
}, 
"dataFileVersion" : { 
"major" : 4, 
"minor" : 22 
}, 
"ok" : 1 
}

看到dataSize，indexSize，storageSize都变小了，但是fileSize没有变化，而且mongo data目录仍然占用4.7G。

2.2 确保先在从库10.192.203.202:27017上进行重构

#查看主从关系

MyReplset:PRIMARY>rs.status(); 
{ 
"set" : "MyReplset", 
"date" :ISODate("2016-11-07T07:10:50.717Z"), 
"myState" : 1, 
"members" : [ 
      { 
          "_id" : 0, 
          "name" :"10.192.203.201:27017", 
          "health" : 1, 
          "state" : 1, 
          "stateStr" :"PRIMARY", 
          "uptime" : 964, 
          "optime" :Timestamp(1478239977, 594), 
          "optimeDate" :ISODate("2016-11-04T06:12:57Z"), 
          "electionTime" :Timestamp(1478502021, 1), 
          "electionDate" :ISODate("2016-11-07T07:00:21Z"), 
          "configVersion" :2, 
          "self" : true 
      }, 
      { 
          "_id" : 1, 
          "name" :"10.192.203.202:27017", 
          "health" : 1, 
          "state" : 2, 
          "stateStr" :"SECONDARY", 
          "uptime" : 628, 
          "optime" :Timestamp(1478239977, 594), 
          "optimeDate" :ISODate("2016-11-04T06:12:57Z"), 
          "lastHeartbeat" :ISODate("2016-11-07T07:10:49.257Z"), 
          "lastHeartbeatRecv": ISODate("2016-11-07T07:10:50.143Z"), 
          "pingMs" : 2, 
          "configVersion" :2 
      }, 
      { 
          "_id" : 2, 
          "name" :"10.192.203.202:10001", 
          "health" : 1, 
          "state": 7, 
          "stateStr" :"ARBITER", 
          "uptime" : 618, 
          "lastHeartbeat" :ISODate("2016-11-07T07:10:49.416Z"), 
          "lastHeartbeatRecv": ISODate("2016-11-07T07:10:49.847Z"), 
          "pingMs" : 2, 
          "configVersion" :2 
      } 
], 
"ok" : 1 
}

2.2.1 关闭数据库

MyReplset:SECONDARY> use admin; 
switched to db admin 
MyReplset:SECONDARY> db.shutdownServer(); 
2016-11-07T15:14:42.548+0800 I NETWORK DBClientCursor::init call() failed 
server should be down... 
2016-11-07T15:14:42.571+0800 I NETWORK trying reconnect to 127.0.0.1:27017(127.0.0.1) failed 
2016-11-07T15:14:42.575+0800 W NETWORK Failed to connect to 127.0.0.1:27017, reason:errno:111 Connection refused 
2016-11-07T15:14:42.575+0800 I NETWORK reconnect 127.0.0.1:27017 (127.0.0.1) failedfailed couldn't connect to server 127.0.0.1:27017 (127.0.0.1), connectionattempt failed 
2016-11-07T15:14:42.634+0800 I NETWORK trying reconnect to 127.0.0.1:27017(127.0.0.1) failed 
2016-11-07T15:14:42.637+0800 W NETWORK Failed to connect to 127.0.0.1:27017, reason:errno:111 Connection refused 
2016-11-07T15:14:42.638+0800I NETWORK reconnect 127.0.0.1:27017(127.0.0.1) failed failed couldn't connect to server 127.0.0.1:27017(127.0.0.1), connection attempt failed

2.2.2 备份，删除，重建数据目录

备份10.192.203.202:27017的数据目录，这里省略

备份完成后，删除，重建该目录。

rm-rf /data/mongo/data

mkdir/data/mongo/data

2.2.3 启动数据库

启动10.192.203.202:27017进程:

/usr/local/mongodb/bin/mongod--config /usr/local/mongodb/mongod.cnf --replSet MyReplset -rest

2.2.4 检查

检查数据库是否正常，之前的数据库是否都存在。

检查下磁盘空间是否缩小。

经检查，空间缩小为4.3G，收缩了400MB。