Tutorial

Lets create simple aiohttp application, and see how aiomonitor can integrate with it.

Basic aiohttp server

import asyncio

import aiomonitor
from aiohttp import web

# Simple handler that returns response after 100s
async def simple(request):
    loop = request.app.loop

    print('Start sleeping')
    await asyncio.sleep(100)
    return web.Response(text="Simple answer")

loop = asyncio.get_event_loop()
# create application and register route create route
app = web.Application(loop=loop)
app.router.add_get('/simple', simple)

# init monitor just before run_app
with aiomonitor.start_monitor(loop):
    # run application with built in aoihttp run_app function
    web.run_app(app, port=8090, host='localhost')

Lets save this code in file simple_srv.py, so we can run it with command:

$ python simple_srv.py
======== Running on http://localhost:8090 ========
(Press CTRL+C to quit)

Connection over telnet

And now it is possible to connect to the running application from separate terminal, by execution nc command, immediately aiomonitor will respond with prompt:

$ nc localhost 50101
Asyncio Monitor: 1 tasks running
Type help for commands
monitor >>>

aiomonitor packaged with own telnet client, just in case you do not have nc or other related utility:

$ python -m aiomonitor.cli
Asyncio Monitor: 1 tasks running
Type help for commands
monitor >>>

Once connection established, one can type commands, for instance help:

monitor >>> help
Available Commands are:
             cancel taskid: Cancel an indicated task
             console: Switch to async Python REPL
             ps: Show task table
             quit: Leave the monitor
             signal signame: Send a Unix signal
             stacktrace: Print a stack trace from the event loop thread
             where taskid: Show stack frames for a task

Library will respond with list of supported commands:

  • ps – shows table of alive tasks with their id and state

  • where – prints stack frame for the task, taskid must be supplied

  • cancel – command cancels task, taskid must be supplied

  • signal – command sends unix signal to the app process

  • stacktrace – prints a stack trace from the event loop thread

  • console – switch to python REPL

  • quit – stops telnet session

Additional commands can be added by subclassing Monitor, see below <cust-commands>.

Python REPL

aiomonitor supports also async python console inside running event loop so you can explore state of your application:

monitor >>> console
Python 3.5.2 (default, Oct 11 2016, 05:05:28)
[GCC 4.2.1 Compatible Apple LLVM 8.0.0 (clang-800.0.38)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
---
This console is running in an asyncio event loop.
It allows you to wait for coroutines using the 'await' syntax.
Try: await asyncio.sleep(1, result=3)
---
>>>

Now you may execute regular function as well as coroutines by adding await keyword:

>>> import aiohttp
>>> session = aiohttp.ClientSession()
>>> resp = await session.get('http://python.org')
>>> resp.status
200
>>> data = await resp.read()
>>> len(data)
47373
>>>

To leave console type exit():

>>> exit()
monitor >>>

Expose Local Variables in Python REPL

Local variables can be exposed in Python REPL by passing additional locals dictionary with mapping variable name in console to the value.

locals = {"foo": "bar"}
with aiomonitor.start_monitor(loop):
    web.run_app(app, port=8090, host='localhost')

As result variable foo available in console:

monitor >>> console
>>> foo
>>> bar
>>> exit()
monitor >>>

Adding custom commands

By employing a custom Monitor subclass, we can add our own commands to the telnet REPL. These are simply methods with names starting with do_. These methods can use the in- and outgoing REPL sockets self._sin and self._sout for IO, like self._sout.write(string) to print to the REPL.

Any parameters to the method will receive their value as a string, if they are meant to be used as e.g. numbers, manual casting is needed.

class MyMon(Monitor):
    @alt_names('moc own')
    def do_my_own_command(self, some_argument):
        """This is a short description

        The first line of the doc will be shown in the help overview, this rest
        will only show up in the "help my_own_command" output.
        This command will have aliases "moc" and "own", just like "help" has "h"
        and "?".
        """
        results = self._do_stuff(self._locals['my_app_instance'])
        self._sout.write('The results are: {}\n'.format(results))

This custom command will be able to do anything you could do in the python REPL, so you can add custom shortcuts here, that would be tedious to do manually in the console.

By using the “locals” argument to start_monitor you can give any of your commands access to anything they might need to do their jobs.

Modify the basic behaviour of the command loop

Like the standard library’s cmd_ module, you can customise how the behaviour of the Monitor in various ways, see api_reference.