stop python program when ssh pipe is broken - python

I'm writing a python script with an infinite while loop that I am running over ssh. I would like the script to terminate when someone kills ssh. For example:
The script (script.py):
while True:
# do something
Will be run as:
ssh foo ./script.py
When I kill the ssh process, I would like the script on the other end to stop running.
I have tried looking for a closed stdout:
while not sys.stdout.closed:
# do something
but this didn't work.
How do I achieve this?
Edit:
The remote machine is a Mac which opens the program in a csh:
502 29352 ?? 0:00.01 tcsh -c python test.py
502 29354 ?? 0:00.04 python test.py
I'm opening the ssh process from a python script like so:
p = Popen(['ssh','foo','./script.py'],stdout=PIPE)
while True:
line = p.stdout.readline()
# etc
EDIT
Proposed Solutions:
Run the script with while os.getppid() != 1
This seems to work on Linux systems, but does not work when the remote machine is running OSX. The problem is that the command is launched in a csh (see above) and so the csh has its parent process id set to 1, but not the script.
Periodically log to stderr
This works, but the script is also run locally, and I don't want to print a heartbeat to stderr.
Run the script in a pseduo tty with ssh -tt.
This does work, but has some weird consequences. Consider the following:
remote_script:
#!/usr/bin/env python
import os
import time
import sys
while True:
print time.time()
sys.stdout.flush()
time.sleep(1)
local_script:
#!/usr/bin/env python
from subprocess import Popen, PIPE
import time
p = Popen(['ssh','-tt','user#foo','remote_script'],stdout=PIPE)
while True:
line = p.stdout.readline().strip()
if line:
print line
else:
break
time.sleep(10)
First of all, the output is really weird, it seems to keep adding tabs or something:
[user#local ~]$ local_script
1393608642.7
1393608643.71
1393608644.71
Connection to foo closed.
Second of all, the program does not quit the first time it receives a SIGINT, i.e. I have to hit Ctrl-C twice in order to kill the local_script.

Okay, I have a solution for you
When the ssh connection closes, the parent process id will change from the pid of the ssh-deamon (the fork that handles your connection) to 1.
Thus the following solves your problem.
#!/usr/local/bin/python
from time import sleep
import os
#os.getppid() returns parent pid
while (os.getppid() != 1):
sleep(1)
pass
Can you confirm this is working in your end too :)
edit
I saw you update.
This is not tested, but to get this idea working on OSX, you may be able to detect if the process of the csh changes. The code below only illustrates an idea and has not been tested. That said i think it would work, but it would not be the most elegant solution. If a cross platform solution using signals could be found, it would be preferred.
def do_stuff():
sleep(1)
if sys.platform == 'darwin':
tcsh_pid = os.getppid()
sshfork_pid = psutil.Process(tcsh_pid).ppid
while (sshfork_pid == psutil.Process(tcsh_pid).ppid)
do_stuff()
elif sys.platform == 'linux':
while (os.getppid() != 1):
sleep(1)
else:
raise Exception("platform not supported")
sys.exit(1)

Have you tried
ssh -tt foo ./script.py

When the terminal connection is lost, the application is supposed to receive SIGHUP signal, so all you have to do is to register a special handler using signal module.
import signal
def MyHandler(self, signum, stackFrame):
errorMessage = "I was stopped by %s" % signum
raise Exception(errorMessage)
# somewhere in the beginning of the __main__:
# registering the handler
signal.signal(signal.SIGHUP, MyHandler)
Note that most likely you'll have to handle some other signals. You can do it in absolutely the same way.

I'd suggest periodically logging to stderr.
This will cause an exception to occur when you no longer have a stderr to write to.

The running script is a child pid of the terminal session. If you close the SSH session properly it will terminate the process. But, another method of going about this is to connect your while loop to another factor and disconnect it from your SSH session.
You can have your script controlled by cron to execute regularly. You can have the while loop have a counter. You can have a sleep command in the loop to control execution. Pretty much anything other than having it connected to your SSH session is valid.
To do this you could use exec & to disconnect instances from your loop.

Related

How to restart a Python script?

In a program I am writing in python I need to completely restart the program if a variable becomes true, looking for a while I found this command:
while True:
if reboot == True:
os.execv(sys.argv[0], sys.argv)
When executed it returns the error [Errno 8] Exec format error. I searched for further documentation on os.execv, but didn't find anything relevant, so my question is if anyone knows what I did wrong or knows a better way to restart a script (by restarting I mean completely re-running the script, as if it were been opened for the first time, so with all unassigned variables and no thread running).
There are multiple ways to achieve the same thing. Start by modifying the program to exit whenever the flag turns True. Then there are various options, each one with its advantages and disadvantages.
Wrap it using a bash script.
The script should handle exits and restart your program. A really basic version could be:
#!/bin/bash
while :
do
python program.py
sleep 1
done
Start the program as a sub-process of another program.
Start by wrapping your program's code to a function. Then your __main__ could look like this:
def program():
### Here is the code of your program
...
while True:
from multiprocessing import Process
process = Process(target=program)
process.start()
process.join()
print("Restarting...")
This code is relatively basic, and it requires error handling to be implemented.
Use a process manager
There are a lot of tools available that can monitor the process, run multiple processes in parallel and automatically restart stopped processes. It's worth having a look at PM2 or similar.
IMHO the third option (process manager) looks like the safest approach. The other approaches will have edge cases and require implementation from your side to handle edge cases.
This has worked for me. Please add the shebang at the top of your code and os.execv() as shown below
#!/usr/bin/env python3
import os
import sys
if __name__ == '__main__':
while True:
reboot = input('Enter:')
if reboot == '1':
sys.stdout.flush()
os.execv(sys.executable, [sys.executable, __file__] + [sys.argv[0]])
else:
print('OLD')
I got the same "Exec Format Error", and I believe it is basically the same error you get when you simply type a python script name at the command prompt and expect it to execute. On linux it won't work because a path is required, and the execv method is basically encountering the same error.
You could add the pathname of your python compiler, and that error goes away, except that the name of your script then becomes a parameter and must be added to the argv list. To avoid that, make your script independently executable by adding "#!/usr/bin/python3" to the top of the script AND chmod 755.
This works for me:
#!/usr/bin/python3
# this script is called foo.py
import os
import sys
import time
if (len(sys.argv) >= 2):
Arg1 = int(sys.argv[1])
else:
sys.argv.append(None)
Arg1 = 1
print(f"Arg1: {Arg1}")
sys.argv[1] = str(Arg1 + 1)
time.sleep(3)
os.execv("./foo.py", sys.argv)
Output:
Arg1: 1
Arg1: 2
Arg1: 3
.
.
.

Python script can't be terminated through Ctrl+C or Ctrl+Break

I have this simple python script called myMain.py to execute another python program automatically with incremental number, and I'm running it on CentOS 7:
#!/usr/bin/python
import os
import sys
import time
def main():
step_indicator = ""
arrow = ">"
step = 2
try:
for i in range(0,360, step):
step_percentage = float(i)/360.0 * 100
if i % 10 == 0:
step_indicator += "="
os.system("python myParsePDB.py -i BP1.pdb -c 1 -s %s" % step)
print("step_percentage%s%s%.2f" % (step_indicator,arrow,step_percentage)+"%")
except KeyboardInterrupt:
print("Stop me!")
sys.exit(0)
if __name__ == "__main__":
main()
For now I only know this script is single thread safe, but I can't terminate it with Ctrl+C keyboard interruption.
I have read some relative questions: such as Cannot kill Python script with Ctrl-C and Stopping python using ctrl+c I realized that Ctrl+Z does not kill the process, it only pauses the process and keep the process in background. Ctrl+Break does work for my case either, I think it only terminates my main thread but keeps the child process.
I also noticed that calling os.system() will spawn a child process from the current executing process. At the same time, I also have os file I/O functions and os.system("rm -rf legacy/*") will be invoked in myParsePDB.py which means this myParsePDB.py child process will spawn child process as well. Then, if I want to catch Ctrl+C in myMain.py, should I daemon only myMain.py or should I daemon each process when they spawn?
This is a general problem that could raise when dealing with signal handling. Python signal is not an exception, it's a wrapper of operating system signal. Therefore, signal processing in python depends on operating system, hardware and many conditions. However, how to deal with these problem is similar.
According to this tutorial, I'll quote the following paragraphs: signal – Receive notification of asynchronous system events
Signals are an operating system feature that provide a means of
notifying your program of an event, and having it handled
asynchronously. They can be generated by the system itself, or sent
from one process to another. Since signals interrupt the regular flow
of your program, it is possible that some operations (especially I/O)
may produce error if a signal is received in the middle.
Signals are identified by integers and are defined in the operating
system C headers. Python exposes the signals appropriate for the
platform as symbols in the signal module. For the examples below, I
will use SIGINT and SIGUSR1. Both are typically defined for all Unix
and Unix-like systems.
In my code:
os.system("python myParsePDB.py -i BP1.pdb -c 1 -s %s" % step) inside the for loop will be executed for a bit of time and will spend some time on I/O files. If the keyboard interrupt is passing too fast and do not catch asynchronously after writing files, the signal might be blocked in operating system, so my execution will still remain the try clause for loop. (Errors detected during execution are called exceptions and are not unconditionally fatal: Python Errors and Exceptions).
Therefore the simplest way to make them asynchonous is wait:
try:
for i in range(0,360, step):
os.system("python myParsePDB.py -i BP1.pdb -c 1 -s %s" % step)
time.sleep(0.2)
except KeyboardInterrupt:
print("Stop me!")
sys.exit(0)
It might hurt performance but it guaranteed that the signal can be caught after waiting the execution of os.system(). You might also want to use other sync/async functions to solve the problem if better performance is required.
For more unix signal reference, please also look at: Linux Signal Manpage

Self Restarting a Python Script

I have created a watchdog timer for my script (Python 3), which allows me to halt execution if anything goes wrong (not shown in code below). However, I would like to have the ability to restart the script automatically using only Python (no external scripts). The code needs to be cross platform compatible.
I have tried subprocess and execv (os.execv(sys.executable, ['python'] + sys.argv)), however I am seeing very weird functionality on Windows. I open the command line, and run the script ("python myscript.py"). The script stops but does not exit (verified through Task Manager), and it will not restart itself unless I press enter twice. I would like it to work automatically.
Any suggestions? Thanks for your help!
import threading
import time
import subprocess
import os
import sys
if __name__ == '__main__':
print("Starting thread list: " + str(threading.enumerate()))
for _ in range(3):
time.sleep(1)
print("Sleeping")
''' Attempt 1 with subprocess.Popen '''
# child = subprocess.Popen(['python',__file__], shell=True)
''' Attempt 2 with os.execv '''
args = sys.argv[:]
args.insert(0, sys.executable)
if sys.platform == 'win32':
args = ['"%s"' % arg for arg in args]
os.execv(sys.executable, args)
sys.exit()
Sounds like you are using threading in your original script, which explains why your can't break your original script when simply pressing Ctrl+C. In that case, you might want to add a KeyboardInterrupt exception to your script, like this:
from time import sleep
def interrupt_this()
try:
while True:
sleep(0.02)
except KeyboardInterrupt as ex:
# handle all exit procedures and data cleaning
print("[*] Handling all exit procedures...")
After this, you should be able to automatically restart your relevant procedure (even from within the script itself, without any external scripts). Anyway, it's a bit hard to know without seeing the relevant script, so maybe I can be of more help if you share some of it.

Executing 2 blocking processes and need the control back in the same ssh session

I'm trying to automate the start up procedure for a Linux Agent machine. I need to run two executable files located in two different directories and get the control back to the prompt so as to proceed with some other task like do a grep to check if the two processes are still running. Here are two different ways using Python in which I tried to do this:
Code snippet 1:(By forking a child process)
import os
import pdb
def child():
cwd = os.getcwd()
os.chdir("THoT")
os.chdir("TH_Node")
print "Executing TH_Node.exe........."
command = "mono TH_Node.exe"
os.system(command)
os._exit(0)
def parent():
i = 0
while i < 1:
i = i + 1
newpid = os.fork()
if newpid == 0:
child()
else:
cwd = os.getcwd()
os.chdir("THoT")
os.chdir("TH_Protocol")
print "Executing TH_Protocol.exe........."
command1 = "mono TH_Protocol.exe"
os.system(command1)
parent()
Code snippet 2:(Using multiprocessing)
import multiprocessing
import time
import sys
import os
import pdb
def TH_Protocol():
os.chdir("THoT")
os.chdir("TH_Protocol")
command = "mono TH_Protocol.exe"
os.system(command)
def TH_Node():
os.chdir("THoT")
os.chdir("TH_Node")
command1 = "mono TH_Node.exe"
os.system(command1)
if __name__ == '__main__':
d = multiprocessing.Process(name='TH_Protocol', target=TH_Protocol)
d.TH_Protocol = True
n = multiprocessing.Process(name='TH_Node', target=TH_Node)
n.TH_Protocol = False
d.start()
n.start()
d.join(1)
n.join()
The problem is although I get both the processes TH_Protocol.exe and TH_Node.exe to run, I need to ssh to another session to run a grep command to check if the two processes are running. I need to get the control back in the same session as the session in which I run my python script. I tried to use the subprocess.Popen as well, but I face the same problem. Is there any way I can solve this issue?
If you just want to run this script in the background, and get control of your ssh session back while it's running… that has nothing to do with Python, or ssh, it's basic shell job control.
For example, assuming your shell on the remote machine is sh/bash/similar, instead of this:
remote_machine$ python script.py
… do this:
remote_machine$ python script.py &
[1] 55341
remote_machine$
Now you've got the prompt back. You can interact with the main interpreter process as %1 or PID 55341. After it finally finishes, the next prompt you get will show something like this:
[1]+ Done python
You can't directly interact with the two child processes this way. You can always grep for them if you want, or search for child processes of PID 55341… but you might find your life easier if you had the child processes do something like print('TH_Protocol on {}'.format(os.getpid())) as soon as they start up, so you don't have to do that.

How to kill headless X server started via Python?

I want to get screenshots of a webpage in Python. For this I am using http://github.com/AdamN/python-webkit2png/ .
newArgs = ["xvfb-run", "--server-args=-screen 0, 640x480x24", sys.argv[0]]
for i in range(1, len(sys.argv)):
if sys.argv[i] not in ["-x", "--xvfb"]:
newArgs.append(sys.argv[i])
logging.debug("Executing %s" % " ".join(newArgs))
os.execvp(newArgs[0], newArgs)
Basically calls xvfb-run with the correct args. But man xvfb says:
Note that the demo X clients used in the above examples will not exit on their own, so they will have to be killed before xvfb-run will exit.
So that means that this script will <????> if this whole thing is in a loop, (To get multiple screenshots) unless the X server is killed. How can I do that?
The documentation for os.execvp states:
These functions all execute a new
program, replacing the current
process; they do not return. [..]
So after calling os.execvp no other statement in the program will be executed. You may want to use subprocess.Popen instead:
The subprocess module allows you to
spawn new processes, connect to their
input/output/error pipes, and obtain
their return codes. This module
intends to replace several other,
older modules and functions, such as:
Using subprocess.Popen, the code to run xlogo in the virtual framebuffer X server becomes:
import subprocess
xvfb_args = ['xvfb-run', '--server-args=-screen 0, 640x480x24', 'xlogo']
process = subprocess.Popen(xvfb_args)
Now the problem is that xvfb-run launches Xvfb in a background process. Calling process.kill() will not kill Xvfb (at least not on my machine...). I have been fiddling around with this a bit, and so far the only thing that works for me is:
import os
import signal
import subprocess
SERVER_NUM = 99 # 99 is the default used by xvfb-run; you can leave this out.
xvfb_args = ['xvfb-run', '--server-num=%d' % SERVER_NUM,
'--server-args=-screen 0, 640x480x24', 'xlogo']
subprocess.Popen(xvfb_args)
# ... do whatever you want to do here...
pid = int(open('/tmp/.X%s-lock' % SERVER_NUM).read().strip())
os.kill(pid, signal.SIGINT)
So this code reads the process ID of Xvfb from /tmp/.X99-lock and sends the process an interrupt. It works, but does yield an error message every now and then (I suppose you can ignore it, though). Hopefully somebody else can provide a more elegant solution. Cheers.

Categories