sshpipe 0.5.2

Loads and dumps

loads(), load() methods are used to read SSH configuration from string stream or file respectively into SSHConfig object. dumps(), dump() methods are used to write SSHConfig object to string stream or file respectively.

SSHConfig Class

SSHConfig class holds SSH configuration as read by load() or loads(). It can then be stored back into SSH configuration file with sshconfig’s dump() and dumps(). *SSHConfig provides get() method to retrieve SSH settings.

Future functionality:

1. validation of configuration.
2. manipulation of configuration (e.g., add key, change flags, etc.)

Example

This example shows a SSHPipe creation from one host to another with. The service in the remote host will accept string message sent via the pipe and would store them into a file.

sshremotehandlerusage.py, below, initiates

 1 importos 2 importmultiprocessingasmp 3 fromsshpipeimportSSHPipe 4  5 defrun(): 6 agent_dir='/var/acrisel/sand/acris/sshpipe/sshpipe/sshpipe_examples' 7 agentpy=os.path.join(agent_dir,"sshremotehandler.py") 8 host='ubuntud01_eventor'# SSH config host name of remote server. 9 10 sshagent=SSHPipe(host,agentpy)11 sshagent.start()12 13 ifnotsshagent.is_alive():14 print("Agent not alive",sshagent.response())15 exit(1)16 17 sshagent.send("This is life.\n")18 sshagent.send("This is also life.\n")19 sshagent.send("This is yet another life.\n")20 sshagent.send("That is all, life.\n")21 sshagent.send("TERM")22 23 ifnotsshagent.is_alive():24 print(sshagent.response())25 exit()26 27 response=sshagent.close()28 ifresponse:29 exitcode,stdout,stderr=response30 print('Response: ',response)31 32 if__name__=='__main__':33 mp.set_start_method('spawn')34 run()

The remote agent sshremotehandler.py is would run through SHHPipe and would loop awaiting input on its stdin stream.

 1 importlogging 2 fromsshpipeimportSSHPipeHandler 3  4 module_logger=logging.getLogger(__file__) 5  6 classMySSHPipeHandler(SSHPipeHandler): 7  8 def__init__(self,*args,**kwargs): 9 super(MySSHPipeHandler,self).__init__(*args,**kwargs)10 self.file=None11 12 defatstart(self,received):13 file="{}{}".format(__file__,".remote.log")14 self.module_logger.debug("Opening file: {}.".format(file))15 self.file=open(file,'w')16 17 defatexit(self,received):18 ifself.fileisnotNone:19 self.file.close()20 super(MySSHPipeHandler,self).atexit(received)21 22 defhandle(self,received):23 self.file.write(str(received))24 25 if__name__=='__main__':26 client=MySSHPipeHandler()27 client.service_loop()

The handler overrides the four methods of SSHPipeHandler. __init__() defines an instance member file, atstart() opens file to which records would be written, atexit() closes the file, and handle() writes received record to file.

Example Explanation

假设我们在某个服务器ubuntud20上运行sshremotehandlerusage.py程序

Classes and Methods

SSHPipe(host,command,name=None,user=None,term_message='TERM',config=None,encoding='utf8',callback=None,logger=None)SSHPipeestablishesconnectiontoremote*host*andruns*command*.*host*canbeipaddress,hostname,orSSHhostname.*name*associatesandidtothepipe.If*user*isprovided,itwillusefortheSSHconnectivity.term_message,is

example

在主进程中

importtimeimportrandomimportloggingfromacrisimportMpLoggerimportosimportmultiprocessingasmpdefsubproc(limit=1,logger_info=None):logger=MpLogger.get_logger(logger_info,name="acrilog.subproc",)foriinrange(limit):sleep_time=3/random.randint(1,10)time.sleep(sleep_time)logger.info("proc [%s]: %s/%s - sleep %4.4ssec"%(os.getpid(),i,limit,sleep_time))level_formats={logging.DEBUG:"[ %(asctime)s ][ %(levelname)s ][ %(message)s ][ %(module)s.%(funcName)s(%(lineno)d) ]",'default':"[ %(asctime)s ][ %(levelname)s ][ %(message)s ]",}mplogger=MpLogger(logging_level=logging.DEBUG,level_formats=level_formats,datefmt='%Y-%m-%d,%H:%M:%S.%f')logger=mplogger.start(name='main_process')logger.debug("starting sub processes")procs=list()forlimitin[1,1]:proc=mp.Process(target=subproc,args=(limit,mplogger.logger_info(),))procs.append(proc)proc.start()forprocinprocs:ifproc:proc.join()logger.debug("sub processes completed")mplogger.stop()

Example output

[2016-12-19,11:39:44.953189][DEBUG][startingsubprocesses][mplogger.<module>(45)][2016-12-19,11:39:45.258794][INFO][proc[932]:0/1-sleep0.3sec][2016-12-19,11:39:45.707914][INFO][proc[931]:0/1-sleep0.75sec][2016-12-19,11:39:45.710487][DEBUG][subprocessescompleted][mplogger.<module>(56)]

name

name标识记录器的基名称。注意，此参数在mplogger init方法及其start方法中都可用。

当设置了consolidate时，mplogger init的name参数用于合并记录器。如果调用start（）方法时未提供主进程的私有记录器，则它也用于主进程的私有记录器。

proecess_key

process_key定义一个或多个日志记录字段，该字段将成为日志文件名的一部分。如果使用它，记录器将为每个记录的进程密钥提供一个文件。如果设置了consolidate，这将是对合并日志的补充。

默认情况下，mplogger使用name作为进程键。如果提供了其他内容，例如processname，则它将作为后缀连接到name。

file_prefix and file_suffix

允许通过设置file_prefix和file_suffix中的一个（或两个）来区分不同运行的日志集。通常，使用pid和粒度datetime作为前缀或后缀会创建一组唯一的日志。

file_mode

file_mode让程序定义如何打开日志。默认情况下，日志以追加模式打开。母鸡，历史被收集起来，并在一夜之间按大小归档。

consolidate

consolidate设置后，将从所有处理日志创建合并日志。 如果设置了consolidated，并且在没有name的情况下调用start（），则将合并到主进程中。

kwargs

kwargs是将传递给filehandler的命名参数。这包括：

文件模式='a'，用于旋转文件处理程序

maxbytes=0，用于旋转文件处理程序

backupcount=0，用于rotatingfilehandler和timedrotingfilehandler

编码='ascii'，用于旋转filehandler和timedrotingfilehandler

delay=false，对于timedrotatingfilehandler

when='h'，用于timedrotatingfilehandler

间隔=1，timedrotatingfilehandler

utc=false，timedrotatingfilehandler

attime=none，对于timedrotatingfilehandler