前言:ORM中的两种创建方式
数据库优先:指的是先创建数据库,包括表和字段的建立,然后根据数据库生成ORM的代码,它是先创建数据库,再创建相关程序代码
代码优先:就是先写代码,然后根据代码去生成数据库结构。
代码优先创建数据库的本质:拿到类-->转换成table对象, 然后根据table对象生成sql语句--> 生成数据库表结构
SQLalchemy联表操作
一对多
# 表结构
class Group(Base): # 一对多的表,组中可能包含多个用户 __tablename__ = 'group' nid = Column(Integer, primary_key=True, autoincrement=True) caption = Column(String(32)) class User(Base): __tablename__ = 'user' uid = Column(Integer, primary_key=True, autoincrement=True) name = Column(String(32)) gid = Column(Integer, ForeignKey('group.nid'))
# 创建完表之后,建立连接 Session = sessionmaker(bind=engine) session = Session() 新增Group表数据 session.add_all([ Group(caption='DBA'), Group(caption='SA'), ] ) session.commit() session.add(Group(caption='QA')) 新建User表数据 session.add_all([ User(name='Bob', gid=1), User(name='Boss', gid=2), ] ) session.commit()
上边的代码中,定义了2个表,一个是“组”,一个是“用户表”。一对多表示:一个组中可能存在多个用户。那么,需求来了,我想要查找出用户表中每个用户对应的组。
常规的联表查询如下:
ret = session.query(User.name, Group.caption).join(Group).all() print(ret) # join默认是进行left join out: [('Bob', 'DBA'), ('Boss', 'SA')]
虽然这样也查询到结果了。但是感觉还是很复杂,那有没有更加简便的方式呢!有,SQLalchemy中为我们提供了relationship类,来帮助我们应对一对多以及多对多的这种复杂表结构。还是以上边的需求为例,用relationship的话,代码如下:
步骤一:需要在创建表的时候建立关系:在创建User表的时候,新增一种虚拟字段类型(数据库中并不真实存在) --> group = relationship("Group")
# 使用relationship,先必须在创建表的时候建立关系 class Group(Base): __tablename__ = 'group' nid = Column(Integer, primary_key=True, autoincrement=True) caption = Column(String(32)) class User(Base): __tablename__ = 'user' uid = Column(Integer, primary_key=True, autoincrement=True) name = Column(String(32)) gid = Column(Integer, ForeignKey('group.nid')) # 仅仅方便查询. group = relationship("Group")
步骤二:开始查询之“正向查询”
ret = session.query(User).all() for obj in ret: # obj 代指user表中的每一行数据,是User对象 # obj.group代指group对象,可以用obj.group.XXX来获取group表中的字段的值 print(obj.uid, obj.name, obj.gid, obj.group, obj.group.nid, obj.group.caption) out: # 1 Bob 1 <__main__.Group object at 0x0000000003B31438> 1 DBA # 2 Boss 2 <__main__.Group object at 0x0000000003B315F8> 2 SA
可以看出,使用了非常简单的查询就将数据全部取了出来。细心的同学可能已经发现,上边写到了正向查询,那么有没有“反向查询”呢?有!
反向查询需要在创建关系的时候,新增一个字段设置backref
group = relationship("Group", backref='uuu')
查询语句就变为了:
obj = session.query(Group).filter(Group.caption=='DBA').first() 带有筛选条件的语句,筛选DBA组所有成员 print(obj) # obj表示符合条件的Group对象 out: <__main__.Group object at 0x00000000032EB710>
print(obj.uuu) # obj.uuu 就是符合筛选条件的User对象 out: [<__main__.User object at 0x0000000003B15400>, <__main__.User object at 0x0000000003B15470>] for i in obj.uuu:# obj.uuu需要用for循环来查询结果 print(i.uid, i.name, i.gid) out: 1 Bob 1 2 Boss 2
小结:
relationship()函数:这个函数告诉ORM,通过使用user.Group,Group类应该和User类连接起来
relationship()使用外键明确这两张表的关系。决定User.group属性是多对一的,即多个用户可以在同一个组里。
relationship()的子函数backref()提供表达反向关系的细节:relationship()对象的集合被Group.uuu引用。多对一的反向关系总是一对多,即一个组可以包含多个用户
多对多
上边的内容介绍了一对多的情况,生产环境中,也经常遇到多对多的情况。比如这样一种情形:某公司有好多服务器,员工也很多,业务交错在这些服务器上,就会造成一种现象:某台机器,多个人的业务在上边部署,这多个人具有这台服务器的权限。而对于每个人来说,他的业务不仅仅需要一台服务器,有时会部署在多台服务器。这样的情形,就构成了多对对的情况,即:单台服务器被多人拥有,单人拥有多台服务器的权限。这时候需要额外的一张表,来专门存储主机与用户之间的对应关系。三张表结构如下:
class Host(Base): __tablename__ = 'host' nid = Column(Integer, primary_key=True, autoincrement=True) hostname = Column(String(32)) port = Column(String(32)) ip = Column(String(32)) class HostUser(Base): __tablename__ = 'host_user' nid = Column(Integer, primary_key=True, autoincrement=True) username = Column(String(32)) # 多对多 class HostToHostUser(Base): __tablename__ = 'host_to_host_user' nid = Column(Integer, primary_key=True, autoincrement=True) # 两个关联的id,以外键的形式存在 host_id = Column(Integer, ForeignKey('host.nid')) host_user_id = Column(Integer, ForeignKey('host_user.nid'))
# 创建完表之后,建立连接 Session = sessionmaker(bind=engine) session = Session() # 新增表数据 session.add_all([ Host(hostname='c1',port='22',ip='1.1.1.1'), Host(hostname='c2',port='22',ip='1.1.1.2'), Host(hostname='c3',port='22',ip='1.1.1.3'), Host(hostname='c4',port='22',ip='1.1.1.4'), Host(hostname='c5',port='22',ip='1.1.1.5'), ]) session.commit() session.add_all([ HostUser(username='root'), HostUser(username='db'), HostUser(username='nb'), HostUser(username='sb'), ]) session.commit() session.add_all([ HostToHostUser(host_id=1,host_user_id=1), HostToHostUser(host_id=1,host_user_id=2), HostToHostUser(host_id=1,host_user_id=3), HostToHostUser(host_id=2,host_user_id=2), HostToHostUser(host_id=2,host_user_id=4), HostToHostUser(host_id=2,host_user_id=3), ]) session.commit()
上边增加的数据,增加了五台服务器,分别是c1,c2,c3,c4,c5。对应ID:1,2,3,4,5。
增加了四个人,分别是root, db, nb, sb,对应ID:1,2,3,4
关系表中 用户与主机的对应关系:
那么,需求来了:获取拥有主机1权限的所有用户。
在没有用到relationship时,查询方式如下:
# 1. 获取主机为c1的对象 host_obj = session.query(Host).filter(Host.hostname == 'c1').first() # 2. 获取查询关系表中拥有主机c1的用户ID host_2_host_user = session.query(HostToHostUser.host_user_id).filter(HostToHostUser.host_id == host_obj.nid).all() print(host_2_host_user) # [(1,), (2,), (3,)] r = zip(*host_2_host_user) print(list(r)) #[(1, 2, 3)] 这就是用户ID列表 # 3. 根据用户ID列表,查询用户名 users = session.query(HostUser.username).filter(HostUser.nid.in_(list(r)[0])).all() print(users) # [('root',), ('db',), ('nb',)]
可以看出,用普通方法查询,非常的繁琐,实际上进行了3次sql查询。那么使用relationship的话,会有哪些改变?
步骤1:创建关系表-->在创建表的时候,创建关系
class Host(Base): __tablename__ = 'host' nid = Column(Integer, primary_key=True, autoincrement=True) hostname = Column(String(32)) port = Column(String(32)) ip = Column(String(32)) class HostUser(Base): __tablename__ = 'host_user' nid = Column(Integer, primary_key=True, autoincrement=True) username = Column(String(32)) # 多对多 class HostToHostUser(Base): __tablename__ = 'host_to_host_user' nid = Column(Integer, primary_key=True, autoincrement=True) host_id = Column(Integer, ForeignKey('host.nid')) host_user_id = Column(Integer, ForeignKey('host_user.nid')) # 新写法中有用 host = relationship('Host', backref='h') host_user = relationship('HostUser', backref='u')
步骤2:反向查找Host,从Host表中获取到HostToHostUser表中的字段信息
host_obj = session.query(Host).filter(Host.hostname == 'c1').first()
print(host_obj) # <__main__.Host object at 0x0000000003C11128> print(host_obj.hostname) # 主机名:c1 print(host_obj.h) # host_obj.h表示HostToHostUser表中,符合筛选条件的HostToHostUser数据对象列表 [<__main__.HostToHostUser object at 0x0000000003B3F198>, <__main__.HostToHostUser object at 0x0000000003B3F898>, <__main__.HostToHostUser object at 0x0000000003B3F908>] # 循环获取用户信息 for item in host_obj.h: # print(item.host_user) # 一行用户的数据, HostUser 表的一行数据对象 print(item.host_user.username) # root db nb
可以看出,使用relationship,整个查询使用一个sql就搞定了。接下来介绍另外一种方式,和之前的差异也是在创建表的时候,具体看如下代码:
class HostToHostUser(Base): __tablename__ = 'host_to_host_user' nid = Column(Integer, primary_key=True, autoincrement=True) host_id = Column(Integer, ForeignKey('host.nid')) host_user_id = Column(Integer, ForeignKey('host_user.nid')) class Host(Base): __tablename__ = 'host' nid = Column(Integer, primary_key=True, autoincrement=True) hostname = Column(String(32)) port = Column(String(32)) ip = Column(String(32)) # 新的书写方式,不需要在多对多表中建立关系 host_user = relationship('HostUser', secondary=HostToHostUser.__table__, backref='h') class HostUser(Base): __tablename__ = 'host_user' nid = Column(Integer, primary_key=True, autoincrement=True) username = Column(String(32))
# 创建完表之后,建立连接 Session = sessionmaker(bind=engine) session = Session() # 步骤1: 获取主机名为c1的Host对象 host_obj = session.query(Host).filter(Host.hostname=='c1').first() # 步骤2: 通过Host对象,正向查找用户名 print(host_obj.host_user) # [<__main__.HostUser object at 0x0000000003BA1DA0>, <__main__.HostUser object at 0x0000000003BA1E10>, <__main__.HostUser object at 0x0000000003BA1E80>] for i in host_obj.host_user: print(i.username) out:root db nb
关于filter和filter_by
用法不同而已,filter 可以像写 sql 的 where 条件那样写 > < 等条件,但引用列名时,需要通过 类名.属性名 的方式,引用列值时使用'=='
filter_by 可以使用 python 的正常参数传递方法传递条件,指定列名时,不需要额外指定类名。,参数名对应名类中的属性名,但似乎不能使用 > < 等条件。
小结:
SQLalchemy联表操作的方式
方式1:使用普通方法:join,多次查询符合筛选条件的数据
方式2:relationshi关系 --》ORM提供的一种便捷查询方法。使用relationship可以方便的实现一对多,以及多对多的查询
一对多关系:
一般情况下:fk与关系放在一起
多对多关系:多一张表,用来存放fk,完成查询需求时有两种方法
方法1、将关系放到关系表中 --》既有正向也有反向
方法2、在某一张表中放关系 --》只有正向
paramiko-基础
paramiko的两种基本用法:基于用户名密码连接、基于公钥私钥连接。主要有两个大类:SSHClient(用于连接远程服务器并执行基本命令)、SFTPClient(用于连接远程服务器并执行上传下载)
SSHClient
基于用户名密码连接:
import paramiko # 创建SSH对象 ssh = paramiko.SSHClient() # 允许连接不在know_hosts文件中的主机 ssh.set_missing_host_key_policy(paramiko.AutoAddPolicy()) # 连接服务器 ssh.connect(hostname='172.25.50.13', port=22, username='work', password='123456') # 执行命令 stdin, stdout, stderr = ssh.exec_command('ls -l') # 获取命令结果 result = stdout.read() print(result.decode()) # 关闭连接 ssh.close() out: total 8 drwxr-xr-x 2 work work 4096 Mar 18 19:22 cn_market_lua drwxrwxr-x 3 work work 4096 Mar 18 19:09 www
基于公钥密钥连接:
import paramiko # 创建key文件 private_key = paramiko.RSAKey.from_private_key_file('/home/auto/.ssh/id_rsa') # 创建SSH对象 ssh = paramiko.SSHClient() # 允许连接不在know_hosts文件中的主机 ssh.set_missing_host_key_policy(paramiko.AutoAddPolicy()) # 连接服务器 ssh.connect(hostname='172.25.50.13', port=22, username='work', key=private_key) # 执行命令 stdin, stdout, stderr = ssh.exec_command('df -h') # 获取命令结果 result = stdout.read() # 关闭连接 ssh.close() out: Filesystem Size Used Avail Use% Mounted on /dev/vda1 20G 4.2G 15G 23% / tmpfs 1.9G 0 1.9G 0% /dev/shm /dev/vdb1 99G 499M 93G 1% /data0
SFTPClient
基于用户名密码连接:
import paramiko # 创建transport transport = paramiko.Transport(('172.25.50.13',22)) transport.connect(username='work',password='123456') # 创建sftpclient,并基于transport连接,把他俩进行绑定 sftp = paramiko.SFTPClient.from_transport(transport) # 将location.py 上传至服务器 /tmp/test.py sftp.put('/tmp/location.py', '/tmp/test.py') # 将remove_path 下载到本地 local_path sftp.get('remove_path', 'local_path') # 关闭session transport.close()
基于公钥密钥连接:
import paramiko # 创建key文件 private_key = paramiko.RSAKey.from_private_key_file('/home/auto/.ssh/id_rsa') transport = paramiko.Transport(('172.25.50.13', 22)) transport.connect(username='work', pkey=private_key ) sftp = paramiko.SFTPClient.from_transport(transport) # 将location.py 上传至服务器 /tmp/test.py sftp.put('/tmp/location.py', '/tmp/test.py') # 将remove_path 下载到本地 local_path sftp.get('remove_path', 'local_path') transport.close()
Paramiko进阶
从基础环节,我们学到了基本的使用方法。现在来看下源代码中这块的执行
第一步:创建SSHclient对象,这部没啥说的,关注一个初始化字段self._transport=None
ssh = paramiko.SSHClient()
第二步:连接到SSH主机,完成认证
ssh.connect(hostname='172.25.50.13', port=22, username='work', password='123456')
这步到底做了什么呢?看下源代码
def connect( self, hostname, port=SSH_PORT, username=None, password=None, pkey=None, key_filename=None, timeout=None, allow_agent=True, look_for_keys=True, compress=False, sock=None, gss_auth=False, gss_kex=False, gss_deleg_creds=True, gss_host=None, banner_timeout=None ): """ Connect to an SSH server and authenticate to it. The server's host key is checked against the system host keys (see `load_system_host_keys`) and any local host keys (`load_host_keys`). If the server's hostname is not found in either set of host keys, the missing host key policy is used (see `set_missing_host_key_policy`). The default policy is to reject the key and raise an `.SSHException`. Authentication is attempted in the following order of priority: - The ``pkey`` or ``key_filename`` passed in (if any) - Any key we can find through an SSH agent - Any "id_rsa", "id_dsa" or "id_ecdsa" key discoverable in ``~/.ssh/`` - Plain username/password auth, if a password was given If a private key requires a password to unlock it, and a password is passed in, that password will be used to attempt to unlock the key. :param str hostname: the server to connect to :param int port: the server port to connect to :param str username: the username to authenticate as (defaults to the current local username) :param str password: a password to use for authentication or for unlocking a private key :param .PKey pkey: an optional private key to use for authentication :param str key_filename: the filename, or list of filenames, of optional private key(s) to try for authentication :param float timeout: an optional timeout (in seconds) for the TCP connect :param bool allow_agent: set to False to disable connecting to the SSH agent :param bool look_for_keys: set to False to disable searching for discoverable private key files in ``~/.ssh/`` :param bool compress: set to True to turn on compression :param socket sock: an open socket or socket-like object (such as a `.Channel`) to use for communication to the target host :param bool gss_auth: ``True`` if you want to use GSS-API authentication :param bool gss_kex: Perform GSS-API Key Exchange and user authentication :param bool gss_deleg_creds: Delegate GSS-API client credentials or not :param str gss_host: The targets name in the kerberos database. default: hostname :param float banner_timeout: an optional timeout (in seconds) to wait for the SSH banner to be presented. :raises BadHostKeyException: if the server's host key could not be verified :raises AuthenticationException: if authentication failed :raises SSHException: if there was any other error connecting or establishing an SSH session :raises socket.error: if a socket error occurred while connecting .. versionchanged:: 1.15 Added the ``banner_timeout``, ``gss_auth``, ``gss_kex``, ``gss_deleg_creds`` and ``gss_host`` arguments. """ if not sock: errors = {} # Try multiple possible address families (e.g. IPv4 vs IPv6) to_try = list(self._families_and_addresses(hostname, port)) for af, addr in to_try: try: sock = socket.socket(af, socket.SOCK_STREAM) if timeout is not None: try: sock.settimeout(timeout) except: pass retry_on_signal(lambda: sock.connect(addr)) # Break out of the loop on success break except socket.error as e: # Raise anything that isn't a straight up connection error # (such as a resolution error) if e.errno not in (ECONNREFUSED, EHOSTUNREACH): raise # Capture anything else so we know how the run looks once # iteration is complete. Retain info about which attempt # this was. errors[addr] = e # Make sure we explode usefully if no address family attempts # succeeded. We've no way of knowing which error is the "right" # one, so we construct a hybrid exception containing all the real # ones, of a subclass that client code should still be watching for # (socket.error) if len(errors) == len(to_try): raise NoValidConnectionsError(errors) t = self._transport = Transport(sock, gss_kex=gss_kex, gss_deleg_creds=gss_deleg_creds) t.use_compression(compress=compress) if gss_kex and gss_host is None: t.set_gss_host(hostname) elif gss_kex and gss_host is not None: t.set_gss_host(gss_host) else: pass if self._log_channel is not None: t.set_log_channel(self._log_channel) if banner_timeout is not None: t.banner_timeout = banner_timeout t.start_client() ResourceManager.register(self, t) server_key = t.get_remote_server_key() keytype = server_key.get_name() if port == SSH_PORT: server_hostkey_name = hostname else: server_hostkey_name = "[%s]:%d" % (hostname, port) # If GSS-API Key Exchange is performed we are not required to check the # host key, because the host is authenticated via GSS-API / SSPI as # well as our client. if not self._transport.use_gss_kex: our_server_key = self._system_host_keys.get(server_hostkey_name, {}).get(keytype, None) if our_server_key is None: our_server_key = self._host_keys.get(server_hostkey_name, {}).get(keytype, None) if our_server_key is None: # will raise exception if the key is rejected; let that fall out self._policy.missing_host_key(self, server_hostkey_name, server_key) # if the callback returns, assume the key is ok our_server_key = server_key if server_key != our_server_key: raise BadHostKeyException(hostname, server_key, our_server_key) if username is None: username = getpass.getuser() if key_filename is None: key_filenames = [] elif isinstance(key_filename, string_types): key_filenames = [key_filename] else: key_filenames = key_filename if gss_host is None: gss_host = hostname self._auth(username, password, pkey, key_filenames, allow_agent, look_for_keys, gss_auth, gss_kex, gss_deleg_creds, gss_host)
最关键的一步:t = self._transport = Transport(sock, gss_kex=gss_kex, gss_deleg_creds=gss_deleg_creds)
这步源码中说明如下:
Create a new SSH session over an existing socket, or socket-like object. This only creates the `.Transport` object; it doesn't begin the SSH session yet. Use `connect` or `start_client` to begin a client session, or `start_server` to begin a server session.
那么可以看出,实际上是这句只是创建了一个session,或者是socket-like的对象。且这个session并没有启动,只有在调用connect或者start_client方法时,session才会生效。
那么,我们是否也可以自己封装一个tranport来定制一些功能呢?
还是以基础中的案例,现在就定制一个基于用户名密码的transport
步骤1:创建一个transport
# 创建transport transport = paramiko.Transport(('172.25.50.13', 22))
步骤2:调用connect方法启动session
# 调用connect方法启动session transport.connect(username='work', password='123456')
步骤3:覆盖原SSHClient对象对于self._transport的定义
ssh = paramiko.SSHClient() ssh._transport = transport
步骤4:执行命令,执行完毕后关闭session
stdin, stdout, stderr = ssh.exec_command('df -h') print(stdout.read().decode()) transport.close()
import paramiko # 创建transport transport = paramiko.Transport(('172.25.50.13', 22)) # 调用connect方法启动session transport.connect(username='work', password='123456') ssh = paramiko.SSHClient() ssh._transport = transport stdin, stdout, stderr = ssh.exec_command('df -h') print(stdout.read().decode()) transport.close()
import paramiko private_key = paramiko.RSAKey.from_private_key_file('/home/auto/.ssh/id_rsa') transport = paramiko.Transport(('172.25.50.13', 22)) transport.connect(username='work', pkey=private_key) ssh = paramiko.SSHClient() ssh._transport = transport stdin, stdout, stderr = ssh.exec_command('df') print(stdout.read().decode()) transport.close()
那么,现在需求又来了。我不仅要实现单纯的执行命令,还要在执行命令之后,上传一个文件,上传文件之后依然能执行命令。
像这样一个复杂的需求,就需要我们自己来封装一个SSHclient,本质上也是对self._transport进行重新封装
# 自己封装一个类似SSHClient的类 import paramiko class SSHConnection(object): def __init__(self, host='172.25.50.13', port=22, username='work',pwd='123456'): self.host = host self.port = port self.username = username self.pwd = pwd self.__k = None def run(self): self.connect() pass self.close() def connect(self): # 创建transport transport = paramiko.Transport((self.host,self.port)) # 启动session transport.connect(username=self.username,password=self.pwd) self.__transport = transport def close(self): self.__transport.close() def cmd(self, command): ssh = paramiko.SSHClient() ssh._transport = self.__transport # 执行命令 stdin, stdout, stderr = ssh.exec_command(command) # 获取命令结果 result = stdout.read() return result def upload(self,local_path, target_path): # 连接,上传 sftp = paramiko.SFTPClient.from_transport(self.__transport) # 将location.py 上传至服务器 /tmp/test.py sftp.put(local_path, target_path) ssh = SSHConnection() ssh.connect() # 执行命令 r1 = ssh.cmd('df') print(r1.decode()) # 上传文件 ssh.upload('s2.py', "/home/alex/s7.py") ssh.close()
堡垒机
上图是一个基本的堡垒机模型,大概就是:
- 管理员为用户在服务器上创建账号(将公钥放置服务器,或者使用用户名密码)
- 用户登陆堡垒机,输入堡垒机用户名密码,显示当前用户管理的服务器列表
- 用户选择服务器,并自动登陆
- 执行操作并同时记录用户操作
使用paramiko就可以实现上述功能,这里先省略数据库方面:
版本一: 1)用户在终端输入内容,并将内容发送至远程服务器,
2)远程服务器执行命令,并将结果返回
3)用户终端显示内容
import paramiko import sys import os import socket import select import getpass from paramiko.py3compat import u # py27中注释掉这行 tran = paramiko.Transport(('172.25.50.13', 22,)) tran.start_client() tran.auth_password('work', '123456') # 打开一个通道 chan = tran.open_session() # 获取一个终端 chan.get_pty() # 激活器 chan.invoke_shell() while True: # 监视用户输入和服务器返回数据 # sys.stdin 处理用户输入 # chan 是之前创建的通道,用于接收服务器返回信息 readable, writeable, error = select.select([chan, sys.stdin, ],[],[],1) if chan in readable: try: x = u(chan.recv(1024)) # py3中 代码 # x = chan.recv(1024) # py2中代码 if len(x) == 0: print('\r\n*** EOF\r\n') break sys.stdout.write(x) sys.stdout.flush() except socket.timeout: pass if sys.stdin in readable: inp = sys.stdin.readline() chan.sendall(inp) chan.close() tran.close()
版本二:在版本一中,可以发现和真是的shell环境是有区别的,不支持补全,非常难用。那么版本二就可以完成tab补全等功能,和真是shell一样。
核心要点:用户每次输入1个字符,就立即发送到服务端。然后在client端接收服务端的返回
# 肆意妄为 2 版本 改变终端模式,每输入一个字符,立即发送 import paramiko import sys import os import socket import select import getpass import termios # windows中没有这个模块 import tty from paramiko.py3compat import u # py2 中需要注释掉这行 tran = paramiko.Transport(('172.25.50.13', 22,)) tran.start_client() tran.auth_password('work', '123456') # 打开一个通道 chan = tran.open_session() # 获取一个终端 chan.get_pty() # 激活器 chan.invoke_shell() # 获取原tty属性 oldtty = termios.tcgetattr(sys.stdin) try: # 为tty设置新属性 # 默认当前tty设备属性: # 输入一行回车,执行 # CTRL+C 进程退出,遇到特殊字符,特殊处理。 # 这是为原始模式,不认识所有特殊符号 # 放置特殊字符应用在当前终端,如此设置,将所有的用户输入均发送到远程服务器 tty.setraw(sys.stdin.fileno()) chan.settimeout(0.0) while True: # 监视 用户输入 和 远程服务器返回数据(socket) # 阻塞,直到句柄可读 r, w, e = select.select([chan, sys.stdin], [], [], 1) if chan in r: try: # x = u(chan.recv(1024)) # py2中需要修改 x = chan.recv(1024) if len(x) == 0: print('\r\n*** EOF\r\n') break sys.stdout.write(x) sys.stdout.flush() except socket.timeout: pass if sys.stdin in r: x = sys.stdin.read(1) if len(x) == 0: break chan.send(x) finally: # 重新设置终端属性 termios.tcsetattr(sys.stdin, termios.TCSADRAIN, oldtty) chan.close() tran.close()
版本二种,实现了一个shell的完善环境。但是实际生产中,还有一个很重要的环节“用户验证”。版本1和版本2中,并不具备这个功能,现在来看版本3
版本三:带有用户验证功能的堡垒机
# 不带日志版 import paramiko import sys import os import socket import getpass from paramiko.py3compat import u # windows does not have termios... try: import termios import tty has_termios = True except ImportError: has_termios = False def interactive_shell(chan): if has_termios: posix_shell(chan) else: windows_shell(chan) def posix_shell(chan): import select oldtty = termios.tcgetattr(sys.stdin) try: tty.setraw(sys.stdin.fileno()) tty.setcbreak(sys.stdin.fileno()) chan.settimeout(0.0) flag = False temp_list = [] while True: r, w, e = select.select([chan, sys.stdin], [], []) if chan in r: try: x = u(chan.recv(1024)) if len(x) == 0: sys.stdout.write('\r\n*** EOF\r\n') break sys.stdout.write(x) sys.stdout.flush() except socket.timeout: pass if sys.stdin in r: x = sys.stdin.read(1) import json if len(x) == 0: break chan.send(x) finally: termios.tcsetattr(sys.stdin, termios.TCSADRAIN, oldtty) def windows_shell(chan): import threading sys.stdout.write("Line-buffered terminal emulation. Press F6 or ^Z to send EOF.\r\n\r\n") def writeall(sock): while True: data = sock.recv(256) if not data: sys.stdout.write('\r\n*** EOF ***\r\n\r\n') sys.stdout.flush() break sys.stdout.write(data) sys.stdout.flush() writer = threading.Thread(target=writeall, args=(chan,)) writer.start() try: while True: d = sys.stdin.read(1) if not d: break chan.send(d) except EOFError: # user hit ^Z or F6 pass def run(): # 获取当前登录用户 username = raw_input('Username ') hostname = raw_input('Hostname: ') pwd = raw_input('password: ') tran = paramiko.Transport((hostname, 22,)) tran.start_client() tran.auth_password(username, pwd) # 打开一个通道 chan = tran.open_session() # 获取一个终端 chan.get_pty() # 激活器 chan.invoke_shell() interactive_shell(chan) chan.close() tran.close() if __name__ == '__main__': run()
版本三在不考虑数据库的情况下,其实已经相对完善了。但是堡垒机还得有一个必要的功能,“记录用户操作日志”
终极版本:带有用户日志的堡垒机
# 记录用户日志 import paramiko import sys import os import socket import getpass from paramiko.py3compat import u # windows does not have termios... try: import termios import tty has_termios = True except ImportError: has_termios = False def interactive_shell(chan): if has_termios: posix_shell(chan) else: windows_shell(chan) def posix_shell(chan): import select oldtty = termios.tcgetattr(sys.stdin) try: tty.setraw(sys.stdin.fileno()) tty.setcbreak(sys.stdin.fileno()) chan.settimeout(0.0) log = open('handle.log', 'a+', encoding='utf-8') flag = False temp_list = [] while True: r, w, e = select.select([chan, sys.stdin], [], []) if chan in r: try: x = u(chan.recv(1024)) if len(x) == 0: sys.stdout.write('\r\n*** EOF\r\n') break if flag: if x.startswith('\r\n'): pass else: temp_list.append(x) flag = False sys.stdout.write(x) sys.stdout.flush() except socket.timeout: pass if sys.stdin in r: x = sys.stdin.read(1) import json if len(x) == 0: break if x == '\t': flag = True else: temp_list.append(x) if x == '\r': log.write(''.join(temp_list)) log.flush() temp_list.clear() chan.send(x) finally: termios.tcsetattr(sys.stdin, termios.TCSADRAIN, oldtty) def windows_shell(chan): import threading sys.stdout.write("Line-buffered terminal emulation. Press F6 or ^Z to send EOF.\r\n\r\n") def writeall(sock): while True: data = sock.recv(256) if not data: sys.stdout.write('\r\n*** EOF ***\r\n\r\n') sys.stdout.flush() break sys.stdout.write(data) sys.stdout.flush() writer = threading.Thread(target=writeall, args=(chan,)) writer.start() try: while True: d = sys.stdin.read(1) if not d: break chan.send(d) except EOFError: # user hit ^Z or F6 pass def run(): default_username = getpass.getuser() username = input('Username [%s]: ' % default_username) if len(username) == 0: username = default_username hostname = input('Hostname: ') if len(hostname) == 0: print('*** Hostname required.') sys.exit(1) tran = paramiko.Transport((hostname, 22,)) tran.start_client() default_auth = "p" auth = input('Auth by (p)assword or (r)sa key[%s] ' % default_auth) if len(auth) == 0: auth = default_auth if auth == 'r': default_path = os.path.join(os.environ['HOME'], '.ssh', 'id_rsa') path = input('RSA key [%s]: ' % default_path) if len(path) == 0: path = default_path try: key = paramiko.RSAKey.from_private_key_file(path) except paramiko.PasswordRequiredException: password = getpass.getpass('RSA key password: ') key = paramiko.RSAKey.from_private_key_file(path, password) tran.auth_publickey(username, key) else: pw = getpass.getpass('Password for %s@%s: ' % (username, hostname)) tran.auth_password(username, pw) # 打开一个通道 chan = tran.open_session() # 获取一个终端 chan.get_pty() # 激活器 chan.invoke_shell() interactive_shell(chan) chan.close() tran.close() if __name__ == '__main__': run()