使用python来访问Hadoop HDFS存储实现文件的操作
如何连接hadoop集群？ fs = hadoop.HadoopDFS("username","password","ugi",64310)fs.disconnect()
如何获取当前工作目录?
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.getWorkingDirectory()
fs.disconnect()
如何更改当前工作目录？
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.setWorkingDirectory("/user/ns-lsp/logs")
fs.disconnect()
如果目录不存在setWorkingDirectory()返回-1,如果执行成功，返回0
如果目录不存在setWorkingDirectory()返回-1,如果执行成功，返回0
如何判断某个文件/目录是否存在？
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.pathExists("/user/ns-lsp/logs")
fs.disconnect()
文件/目录存在，返回0，如果不存在，返回-1
如何创建一个目录?
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.createDirectory("/user/ns-lsp/logs/cjj")
fs.disconnect()
如果目录已经存在，则返回-1，如果目录创建成功，返回0
如何获得当前默认块大小？
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.getDefaultBlockSize()
fs.disconnect()
如何获得当期目录下的文件/目录？
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.listDirectory("/user/ns-lsp/logs")
fs.disconnect()
如何移动一个文件/目录?
同一HDFS内移动文件：
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.move("/user/ns-lsp/logs/cjj","/user/ns-lsp/logs/cjj_new")
fs.disconnect()
不同HDFS之间移动文件：
target_fs = hadoop.HadoopDFS("username","password","ugi",64310)
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.move("/user/ns-lsp/logs/cjj","/user/ns-lsp/logs/cjj_new",target_fs)
fs.disconnect()
如何删除一个文件/目录?
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.delete("/user/ns-lsp/logs/cjj_new")
fs.disconnect()
如何重命名一个文件/目录?
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.rename("/user/ns-lsp/logs/cjj","/user/ns-lsp/logs/cjj1")
fs.disconnect()
如何修改一个文件/目录的权限?
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.chmod("/user/ns-lsp/logs/cjj",7)
fs.disconnect()
如何文件块所在的服务器名?
有时我们需要查找某些文件块所在的服务器名是什么，可以如下使用:
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.getHosts("/user/ns-lsp/logs/cjj/a",0,1)
fs.disconnect()
返回包含服务器名的列表.
$ python gethosts.py
['xxxx']
如何获取一个文件/目录的信息?
fs = hadoop.HadoopDFS("username","password","ugi",64310)
pathinfo = fs.getPathInfo("/user/ns-lsp/logs/cjj")
fs.disconnect()
getPathInfo()返回一个hdfsFileInfo类。
如何指定文件的备份数?
fs = hadoop.HadoopDFS("username","password","ugi",64310)
print fs.setReplication("/user/ns-lsp/logs/cjj/a",3)
fs.disconnect()
如何打开一个文件，并读取数据?
要操作文件，需要创建一个HadoopFile对象，并利用read()方法读取数据.
fs = hadoop.HadoopDFS("username","password","ugi",64310)
fh = hadoop.HadoopFile(fs,'/user/ns-lsp/logs/cjj/a')
print fh.read()
fh.close()
fs.disconnect()
页:
[1]