1、同步各个服务器时间

yum install ntp

ntpdate ntp.fudan.edu.cn

hdfs-site.xml配置

如果没有配置这一项，hadoop默认是0.0.0.0:50090，如果你的namenode和secondarynamenode配置在同一台服务器上，是没有问题的。如果分开部署没有指定该值，则hadoop会在namenode本机上找，就会出错

<property>
        <name>dfs.secondary.http.address</name>
        <value>snn0001:50090</value>
</property>

<property>
        <name>dfs.http.address</name>
        <value>nn0001:50070</value>
</property>

1、edits和fsimage

NameNode会把用户对FileSystem的操作保存在日志文件edits中

每次NameNode重启时，首先从镜像文件fsimage中读取HDFS数据，并把日志文件合并到fsimage中。

2、checkpoint

有两个参数控制SecondaryNameNode checkpoint

fs.checkpoint.period表示两次checkpoint的时间间隔，默认为3600s

fs.checkpoint.size规定edits文件最大值，超过该值即checkpoint，默认64M

可以通过NameNode的start-dfs.sh启动SecondaryNameNode

也可以通过./hadoop secondarynamenode -checkpoint或者./hadoop secondarynamenode -checkpoint force

3、恢复数据：

配置一台和NameNode一样的服务器

创建dfs.name.dir文件夹，注意：该文件夹不能包含合法的fsimage，否则会执行失败。因为NameNode会检查fs.checkpoint.dir目录下镜像的一致性，但是不会做任何改动。

注意：可以使用nfs备份dfs.name.dir和${hadoop.tmp.dir}/dfs/namesecondary

新建目录/hadoop/dfs/namenode和/hadoop/dfs/secondarynamenode

执行命令：./hadoop namenode -importCheckpoint，NameNode会读取checkpoint文件，保存到dfs.name.dir

出现以下错误：

12/01/24 00:02:56 WARN mortbay.log: /getimage: java.io.IOException: GetImage failed. java.net.ConnectException: Connection refused

这是因为没有在hdfs-site.xml配置dfs.secondary.http.address，上面已经配置了

下面的错误是由于数据块完整率没有达到hadoop规定的0.9990要求，所以namenode处于安全状态

把dfs.repliation值设置为2，重新格式化namenode，并上传数据，再次执行./hadoop namenode -importCheckpoint

在0.21.0中，可以通过Checkpoint Node和Backup Node做checkpoint

以下是官网对SecondaryNameNode的说明

The NameNode stores modifications to the file system as a log appended to a native file system file (edits). When a NameNode starts up, it reads HDFS state from an image file (fsimage) and then applies edits from the edits log file. It then writes new HDFS state to the fsimage and starts normal operation with an empty edits file. Since NameNode merges fsimage and edits files only during start up, the edits log file could get very large over time on a busy cluster. Another side effect of a larger edits file is that next restart of NameNode takes longer.

The secondary NameNode merges the fsimage and the edits log files periodically and keeps edits log size within a limit. It is usually run on a different machine than the primary NameNode since its memory requirements are on the same order as the primary NameNode. The secondary NameNode is started bybin/start-dfs.sh on the nodes specified in conf/masters file.

The start of the checkpoint process on the secondary NameNode is controlled by two configuration parameters.

• fs.checkpoint.period, set to 1 hour by default, specifies the maximum delay between two consecutive checkpoints, and
• fs.checkpoint.size, set to 64MB by default, defines the size of the edits log file that forces an urgent checkpoint even if the maximum checkpoint delay is not reached.

The secondary NameNode stores the latest checkpoint in a directory which is structured the same way as the primary NameNode's directory. So that the check pointed image is always ready to be read by the primary NameNode if necessary.

The latest checkpoint can be imported to the primary NameNode if all other copies of the image and the edits files are lost. In order to do that one should:

• Create an empty directory specified in the dfs.name.dir configuration variable;
• Specify the location of the checkpoint directory in the configuration variable fs.checkpoint.dir;
• and start the NameNode with -importCheckpoint option.

The NameNode will upload the checkpoint from the fs.checkpoint.dir directory and then save it to the NameNode directory(s) set in dfs.name.dir. The NameNode will fail if a legal image is contained indfs.name.dir. The NameNode verifies that the image in fs.checkpoint.dir is consistent, but does not modify it in any way.