Increasing the OpLog Size in MongoDB(don't need full resync)

OpLog用于MongoDB数据库复制场合,特征:
1. 设置容量上限,
2. 循环使用
3. local数据库专用

对于MongoDB来说设置好适当的OpLog上限是非常有必要的，因为加大OpLog Size需要重启Master和Slave数据库，并且Slave数据库需要重新同步(full sync).
如果不做FULL SYNC，看看会出现什么样的情况:
MASTER:
[root@db6 data1]# /opt/mongodb-linux-x86_64-1.6.4/bin/mongo 127.0.0.1:5282/test
MongoDB shell version: 1.6.4
connecting to: 127.0.0.1:5282/test
> db.tbl_test.insert({"id":2,"name":"digoal"})      
> exit
bye

SlAVE
[root@db6 data1]# /opt/mongodb-linux-x86_64-1.6.4/bin/mongo 127.0.0.1:5283/test
MongoDB shell version: 1.6.4
connecting to: 127.0.0.1:5283/test
> db.tbl_test.find()
{ "_id" : ObjectId("4cfcbbffb405701a1a121820"), "id" : 1, "name" : "digoal.zhou" }
已经无法完成同步:
分析原因:
MASTER:
> db.printReplicationInfo()      
configured oplog size:   990MB
log length start to end: 91secs (0.03hrs)
oplog first event time:  Mon Dec 06 2010 18:38:14 GMT+0800 (CST)
oplog last event time:   Mon Dec 06 2010 18:39:45 GMT+0800 (CST)
now:                     Mon Dec 06 2010 18:39:52 GMT+0800 (CST)
> exit

SLAVE:
> db.printSlaveReplicationInfo()
source:   db6:5282
         syncedTo: Mon Dec 06 2010 18:34:41 GMT+0800 (CST)
                 = 286secs ago (0.08hrs)
> exit
bye
时间已经错过了,取不到所要的数据.
解决办法:
(首先确保MASTER SLAVE在关闭时数据是一致的，否则数据还是会有问题)
在删除OpLog后，启动MASTER前调整MASTER服务器的时间到db.printSlaveReplicationInfo()输出的时间之前(Mon Dec 06 2010 18:34:41 GMT+0800 (CST)),这样的话启动MASTER得到的db.printReplicationInfo()时间会比SLAVE需要的时间早,数据会继续同步.确认同步后再把两台服务器的时间同步一下.以正确的时间为准.（OPLOG里面是存储时间戳的,oplog.$main具体的字段如下）
ts
Timestamp for the operation. The timestamp type is an internal type used to track
when operations are performed. It is composed of a 4-byte timestamp and a 4-byte
incrementing counter.
op
Type of operation performed as a 1-byte code (e.g., “i” for an insert).
ns
Namespace (collection name) where the operation was performed.
o
Document further specifying the operation to perform. For an insert, this would
be the document to insert.


加大OpLog Size的几个原因分析:
1. Slave长时间Down机或与Master断开连接，超过OpLog轮循时间。
 如，在SLAVE节点执行
> db.printSlaveReplicationInfo();
    source:   localhost:27017
    syncedTo: Tue Mar 30 2010 16:44:01 GMT-0400 (EDT)
    = 12secs ago (0hrs)
同步截至12秒前。如果OpLog中最早的记录 比 Tue Mar 30 2010 16:44:01 GMT-0400 (EDT)  还要新的话,完蛋,需要做full sync。
2. SLAVE数据库做FULL SYNC的时间过长，超过了OpLog可以容纳的操作时间。
如: 在SLAVE数据库使用如下命令开始从master数据库full sync :  
> use admin  
> db.runCommand({resync: 1})  
或 调用--autoresync参数重启SLAVE: mongod --slave --autoresync  

例 : 增加OpLog SIZE

1. 查看Master OpLog Size

> db.printReplicationInfo();

configured oplog size:   1048.576MB

log length start to end: 7200secs (2hrs)

oplog first event time:  Wed Mar 03 2010 16:20:39 GMT-0500 (EST)

oplog last event time:   Wed Mar 03 2010 18:20:39 GMT-0500 (EST)

now:                     Wed Mar 03 2010 18:40:34 GMT-0500 (EST)


2. 关闭master mongod

$ # Stop mongod - killall mongod or kill -2 or ctrl-c) - then:

$ rm /data/db/local.*

$ mongod --oplog=8038 --master

加到8G


3. 重启slave

mongod --slave --autoresync


例 :  Manually Allocating OpLog Files

MongoDB允许在数据库开启前先手工分配OpLog 文件.可能是出于文件在DISK上存储连续性的考虑吧.(猜测)

1. 生成文件 (20GB)

cd /tmp/local
for i in {0..9}
do

echo $i

head -c 2146435072 /dev/zero > local.$i

done

Note that the datafiles aren't exactly 2GB due MongoDB's max int size.

2. 关闭master mongodb

3. 移动文件 (假设$MONGODB = /data/db )

$ mv /data/db/local.* /safe/place

$ mv /tmp/local/* /data/db/

4. 重启master

$ mongod --master --oplogSize=20000

5. Finally, resync the slave. This can be done by shutting down the slave, deleting all its datafiles, and restarting it.