I am relatively new to Python and very new to Simpy. I am trying to build a model of a hospital system that:
Has roughly 20 resources (units) where each resource has a capacity of 5 to 50 (beds)
Not all units can serve the same patients. They have specialties.
When a patient needs a bed, the hospital requests 1 bed from roughly 5 of the 20 units.
So, what I want to do is make a request against multiple Unit resources and only capture 1 bed from 1 available Unit. After many iterations, I think I have found a way of doing this but my approach feels overly complicated.
Below I am showing code using the conditional AnyOf of Simpy. The way AnyOf works is if more than 1 resource has the availability, then more than 1 resource will be captured. So, after the AnyOf request, I release any extra captured beds and cancel any requests still pending.
Is there an easier approach?
from dataclasses import dataclass
import simpy
from simpy.events import AnyOf, Event
import random
#dataclass
class Unit():
env: simpy.Environment()
identifier: str
capacity: int = 1
def __post_init__(self):
self.beds: simpy.Resource = simpy.Resource(env, self.capacity)
def make_request(env, units: list[Unit]):
# create a list of simpy request events for unit 1, 3 and 5
# purpose is to put in a conditional request
any_of_request: list[Event] = []
request_to_unit_dictionary = {}
random_pool_size = random.randint(1, 5)
for x in range(random_pool_size):
random_unit = random.randint(0, 9)
unit = units[random_unit]
res_request = unit.beds.request()
request_to_unit_dictionary[res_request] = unit
any_of_request.append(res_request)
get_one_bed_request = AnyOf(env, any_of_request)
captured_units = yield get_one_bed_request
# how do i determine what unit was captured by the request?
# it is possible that both resources are avaliable
# and both request get filled
captured_requests = list(captured_units.keys())
captured_request = captured_requests[0]
captured_unit: Unit = request_to_unit_dictionary[captured_request]
print(captured_unit)
# if I understand correctly, if I only want 1 request then
# release any "extra" captures
for r in captured_requests:
if r != captured_request:
r.resource.release(r)
# cancel any of the request not captured.
for r in any_of_request:
if r not in captured_requests:
r.cancel()
env = simpy.Environment()
# create 10 units, each with 1 capacity for this example
units: list[Unit] = []
for x in range(1, 10):
units.append(Unit(identifier=f'unit{x}', env=env, capacity=1))
env.process(make_request(env, units))
env.process(make_request(env, units))
env.process(make_request(env, units))
env.process(make_request(env, units))
env.run()
Thanks,
Dan
I have code like the one below :
def expensive(self,c,v):
.....
def inner_loop(self,c,collector):
self.db.query('SELECT ...',(c,))
for v in self.db.cursor.fetchall() :
collector.append( self.expensive(c,v) )
def method(self):
# create a Pool
#join the Pool ??
self.db.query('SELECT ...')
for c in self.db.cursor.fetchall() :
collector = []
#RUN the whole cycle in parallel in separate processes
self.inner_loop(c, collector)
#do stuff with the collector
#! close the pool ?
both the Outer and the Inner loop are thousands of steps ...
I think I understand how to run a Pool of couple of processes.
All the examples I found show that more or less.
But in my case I need to lunch a persistent Pool and then feed the data (c-value). Once a inner-loop process has finished I have to supply the next-available-c-value.
And keep the processes running and collect the results.
How do I do that ?
A clunky idea I have is :
def method(self):
ws = 4
with Pool(processes=ws) as pool :
cs = []
for i,c in enumerate(..) :
cs.append(c)
if i % ws == 0 :
res = [pool.apply(self.inner_loop, (c)) for i in range(ws)]
cs = []
collector.append(res)
will this keep the same pool running !! i.e. not lunch new process every time ?i
Do I need 'if i % ws == 0' part or I can use imap(), map_async() and the Pool obj will block the loop when available workers are exhausted and continue when some are freed ?
Yes, the way that multiprocessing.Pool works is:
Worker processes within a Pool typically live for the complete duration of the Pool’s work queue.
So simply submitting all your work to the pool via imap should be sufficient:
with Pool(processes=4) as pool:
initial_results = db.fetchall("SELECT c FROM outer")
results = [pool.imap(self.inner_loop, (c,)) for c in initial_results]
That said, if you really are doing this to fetch things from the DB, it may make more sense to move more processing down into that layer (bring the computation to the data rather than bringing the data to the computation).
I am using flask, and have a route that sends emails to people. I am using threading to send them faster. When i run it on my local machine it takes about 12 seconds to send 300 emails. But when I run it on lambda thorough API Gateway it times out.
here's my code:
import threading
def async_mail(app, msg):
with app.app_context():
mail.send(msg)
def mass_mail_sender(order, user, header):
html = render_template('emails/pickup_mail.html', bruger_order=order.ordre, produkt=order.produkt)
msg = Message(recipients=[user],
sender=('Sender', 'infor#example.com'),
html=html,
subject=header)
thread = threading.Thread(target=async_mail, args=[create_app(), msg])
thread.start()
return thread
#admin.route('/lager/<url_id>/opdater', methods=['POST'])
def update_stock(url_id):
start = time.time()
if current_user.navn != 'Admin':
abort(403)
if request.method == 'POST':
produkt = Produkt.query.filter_by(url_id=url_id)
nyt_antal = int(request.form['bestilt_hjem'])
produkt.balance = nyt_antal
produkt.bestilt_hjem = nyt_antal
db.session.commit()
orders = OrdreBog.query.filter(OrdreBog.produkt.has(func.lower(Produkt.url_id == url_id))) \
.filter(OrdreBog.produkt_status == 'Ikke klar').all()
threads = []
for order in orders:
if order.antal <= nyt_antal:
nyt_antal -= order.antal
new_thread = mass_mail_sender(order, order.ordre.bruger.email, f'Din bog {order.produkt.titel} er klar til afhentning')
threads.append(new_thread)
order.produkt_status = 'Klar til afhentning'
db.session.commit()
for thread in threads:
try:
thread.join()
except Exception:
pass
end = time.time()
print(end - start)
return 'Emails sendt'
return ''
AWS lambda functions designed to run functions within these constraints:
Memory– The amount of memory available to the function during execution. Choose an amount between 128 MB and 3,008 MB in 64-MB increments.
Lambda allocates CPU power linearly in proportion to the amount of memory configured. At 1,792 MB, a function has the equivalent of one full vCPU (one vCPU-second of credits per second).
Timeout – The amount of time that Lambda allows a function to run before stopping it. The default is 3 seconds. The maximum allowed value is 900 seconds.
To put this in your mail sending multi threaded python code. The function will terminate automatically either when your function execution completes successfully or it reaches configured timeout.
I understand you want single python function to send n number of emails "concurrently". To achieve this with lambda try the "Concurrency" setting and trigger your lambda function through a local script, S3 hosted html/js triggered by cloud watch or EC2 instance.
Concurrency – Reserve concurrency for a function to set the maximum number of simultaneous executions for a function. Provision concurrency to ensure that a function can scale without fluctuations in latency.
https://docs.aws.amazon.com/lambda/latest/dg/configuration-console.html
Important: All above settings will affect your lambda function execution cost significantly. So plan and compare before applying.
If you need any more help, let me know.
Thank you.
I am a newbie. My current project is when the current end decides to start the modbus service, I will create a process for the modbus service. Then the value is obtained in the parent process, through the ZeroMQ PUB/SUB to pass the value, I now want to update the value of the modbus register in the modbus service process.
I tried the method mentioned by pymodbus provided by updating_server.py, and twisted.internet.task.LoopingCall() to update the value of the register, but this will make it impossible for me to connect to my server with the client. I don't know why?
Use LoopingCall() to establish the server, the log when the client connects.
Then I tried to put both the uploading and startTCPserver in the async loop, but the update was only entered for the first time after the startup, and then it was not entered.
Currently, I'm using the LoopingCall() to handle updates, but I don't think this is a good way.
This is the code I initialized the PUB and all the tags that can read the tag.
from loop import cycle
import asyncio
from multiprocessing import Process
from persistence import models as pmodels
from persistence import service as pservice
from persistence import basic as pbasic
import zmq
from zmq.asyncio import Context
from common import logging
from server.modbustcp import i3ot_tcp as sertcp
import common.config as cfg
import communication.admin as ca
import json
import os
import signal
from datetime import datetime
from server.opcuaserver import i3ot_opc as seropc
async def main():
future = []
task = []
global readers, readers_old, task_flag
logger.debug("connecting to database and create table.")
pmodels.connect_create()
logger.debug("init read all address to create loop task.")
cycle.init_readers(readers)
ctx = Context()
publisher = ctx.socket(zmq.PUB)
logger.debug("init publish [%s].", addrs)
publisher.bind(addrs)
readers_old = readers.copy()
for reader in readers:
task.append(asyncio.ensure_future(
cycle.run_readers(readers[reader], publisher)))
if not len(task):
task_flag = True
logger.debug("task length [%s - %s].", len(task), task)
opcua_server = LocalServer(seropc.opc_server, "opcua")
future = [
start_get_all_address(),
start_api(),
create_address_loop(publisher, task),
modbus_server(),
opcua_server.run()
]
logger.debug("run loop...")
await asyncio.gather(*future)
asyncio.run(main(), debug=False)
This is to get the device tag value and publish it.
async def run_readers(reader, publisher):
while True:
await reader.run(publisher)
class DataReader:
def __init__(self, freq, clients):
self._addresses = []
self._frequency = freq
self._stop_signal = False
self._clients = clients
self.signature = sign_data_reader(self._addresses)
async def run(self, publisher):
while not self._stop_signal:
for addr in self._addresses:
await addr.read()
data = {
"type": "value",
"data": addr._final_value
}
publisher.send_pyobj(data)
if addr._status:
if addr.alarm_log:
return_alarm_log = pbasic.get_log_by_time(addr.alarm_log['date'])
if return_alarm_log:
data = {
"type": "alarm",
"data": return_alarm_log
}
publisher.send_pyobj(data)
self.data_send(addr)
logger.debug("run send data")
await asyncio.sleep(int(self._frequency))
def stop(self):
self._stop_signal = True
modbus server imports
from common import logging
from pymodbus.server.asynchronous import StartTcpServer
from pymodbus.device import ModbusDeviceIdentification
from pymodbus.datastore import ModbusSequentialDataBlock
from pymodbus.datastore import ModbusSlaveContext, ModbusServerContext
from persistence import service as pservice
from persistence import basic as pbasic
import zmq
import common.config as cfg
import struct
import os
import signal
from datetime import datetime
from twisted.internet.task import LoopingCall
def updating_writer(a):
logger.info("in updates of modbus tcp server.")
context = a[0]
# while True:
if check_pid(os.getppid()) is False:
os.kill(os.getpid(), signal.SIGKILL)
url = ("ipc://{}" .format(cfg.get('ipc', 'pubsub')))
logger.debug("connecting to [%s].", url)
ctx = zmq.Context()
subscriber = ctx.socket(zmq.SUB)
subscriber.connect(url)
subscriber.setsockopt(zmq.SUBSCRIBE, b"")
slave_id = 0x00
msg = subscriber.recv_pyobj()
logger.debug("updates.")
if msg['data']['data_type'] in modbus_server_type and msg['type'] == 'value':
addr = pservice.get_mbaddress_to_write_value(msg['data']['id'])
if addr:
logger.debug(
"local address and length [%s - %s].",
addr['local_address'], addr['length'])
values = get_value_by_type(msg['data']['data_type'], msg['data']['final'])
logger.debug("modbus server updates values [%s].", values)
register = get_register(addr['type'])
logger.debug(
"register [%d] local address [%d] and value [%s].",
register, addr['local_address'], values)
context[slave_id].setValues(register, addr['local_address'], values)
# time.sleep(1)
def tcp_server(pid):
logger.info("Get server configure and device's tags.")
st = datetime.now()
data = get_servie_and_all_tags()
if data:
logger.debug("register address space.")
register_address_space(data)
else:
logger.debug("no data to create address space.")
length = register_number()
store = ModbusSlaveContext(
di=ModbusSequentialDataBlock(0, [0] * length),
co=ModbusSequentialDataBlock(0, [0] * length),
hr=ModbusSequentialDataBlock(0, [0] * length),
ir=ModbusSequentialDataBlock(0, [0] * length)
)
context = ModbusServerContext(slaves=store, single=True)
identity = ModbusDeviceIdentification()
identity.VendorName = 'pymodbus'
identity.ProductCode = 'PM'
identity.VendorUrl = 'http://github.com/bashwork/pymodbus/'
identity.ProductName = 'pymodbus Server'
identity.ModelName = 'pymodbus Server'
identity.MajorMinorRevision = '2.2.0'
# ----------------------------------------------------------------------- #
# set loop call and run server
# ----------------------------------------------------------------------- #
try:
logger.debug("thread start.")
loop = LoopingCall(updating_writer, (context, ))
loop.start(1, now=False)
# process = Process(target=updating_writer, args=(context, os.getpid(),))
# process.start()
address = (data['tcp_ip'], int(data['tcp_port']))
nt = datetime.now() - st
logger.info("modbus tcp server begin has used [%s] s.", nt.seconds)
pservice.write_server_status_by_type('modbus', 'running')
StartTcpServer(context, identity=identity, address=address)
except Exception as e:
logger.debug("modbus server start error [%s].", e)
pservice.write_server_status_by_type('modbus', 'closed')
This is the code I created for the modbus process.
def process_stop(p_to_stop):
global ptcp_flag
pid = p_to_stop.pid
os.kill(pid, signal.SIGKILL)
logger.debug("process has closed.")
ptcp_flag = False
def ptcp_create():
global ptcp_flag
pid = os.getpid()
logger.debug("sentry pid [%s].", pid)
ptcp = Process(target=sertcp.tcp_server, args=(pid,))
ptcp_flag = True
return ptcp
async def modbus_server():
logger.debug("get mosbuc server's status.")
global ptcp_flag
name = 'modbus'
while True:
ser = pservice.get_server_status_by_name(name)
if ser['enabled']:
if ser['tcp_status'] == 'closed' or ser['tcp_status'] == 'running':
tags = pbasic.get_tag_by_name(name)
if len(tags):
if ptcp_flag is False:
logger.debug("[%s] status [%s].", ser['tcp_name'], ptcp_flag)
ptcp = ptcp_create()
ptcp.start()
else:
logger.debug("modbus server is running ...")
else:
logger.debug("no address to create [%s] server.", ser['tcp_name'])
pservice.write_server_status_by_type(name, "closed")
else:
logger.debug("[%s] server is running ...", name)
else:
if ptcp_flag:
process_stop(ptcp)
logger.debug("[%s] has been closed.", ser['tcp_name'])
pservice.write_server_status_by_type(name, "closed")
logger.debug("[%s] server not allowed to running.", name)
await asyncio.sleep(5)
This is the command that Docker runs.
/usr/bin/docker run --privileged --network host --name scout-sentry -v /etc/scout.cfg:/etc/scout.cfg -v /var/run:/var/run -v /sys:/sys -v /dev/mem:/dev/mem -v /var/lib/scout:/data --rm shulian/scout-sentry
This is the Docker configuration file /etc/scout.cfg.
[scout]
mode=product
[logging]
level=DEBUG
[db]
path=/data
[ipc]
cs=/var/run/scout-cs.sock
pubsub=/var/run/pubsub.sock
I want to be able to trigger the modbus value update function when there is a message coming from ZeroMQ, and it will be updated correctly.
Let's start from inside out.
Q : ...this will make it impossible for me to connect to my server with the client. I don't know why?
ZeroMQ is a smart broker-less messaging / signaling middleware or better a platform for smart-messaging. In case one feels not so much familiar with the art of Zen-of-Zero as present in ZeroMQ Architecture, one may like to start with ZeroMQ Principles in less than Five Seconds before diving into further details.
The Basis :
The Scalable Formal Communication Archetype, borrowed from ZeroMQ PUB/SUB, does not come at zero-cost.
This means that each infrastructure setup ( both on PUB-side and on SUB-side ) takes some, rather remarkable time and no one can be sure of when the AccessNode cnfiguration results in RTO-state. So the SUB-side (as proposed above) ought be either a permanent entity, or the user shall not expect to make it RTO in zero-time, after a twisted.internet.task.LoopingCall() gets reinstated.
Preferred way: instantiate your (semi-)persistent zmq.Context(), get it configured so as to serve the <aContextInstance>.socket( zmq.PUB ) as needed, a minimum safeguarding setup being the <aSocketInstance>.setsockopt( zmq.LINGER, 0 ) and all transport / queuing / security-handling details, that the exosystem exposes to your code ( whitelisting and secure sizing and resources protection being the most probable candidates - but details are related to your application domain and the risks that you are willing to face being prepared to handle them ).
ZeroMQ strongly discourages from sharing ( zero-sharing ) <aContextInstance>.socket()-instances, yet the zmq.Context()-instance can be shared / re-used (ref. ZeroMQ Principles... ) / passed to more than one threads ( if needed ).
All <aSocketInstance>{.bind()|.connect()}- methods are expensive, so try to setup the infrastructure AccessPoint(s) and their due error-handling way before one tries to use the their-mediated communication services.
Each <aSocketInstance>.setsockopt( zmq.SUBSCRIBE, ... ) is expensive in that it may take ( depending on (local/remote) version ) a form of a non-local, distributed-behaviour - local side "sets" the subscription, yet the remote side has to "be informed" about such state-change and "implements" the operations in line with the actual (propagated) state. While in earlier versions, all messages were dispatched from the PUB-side and all the SUB-side(s) were flooded with such data and were left for "filtering" which will be moved into a local-side internal-Queue, the newer versions "implement" the Topic-Filter on the PUB-side, which further increases the latency of setting the new modus-operandi in action.
Next comes the modus-operandi: how <aSocketInstance>.recv() gets results:
In their default API-state, .recv()-methods are blocking, potentially infinitely blocking, if no messages arrive.
Solution: avoid blocking-forms of calling ZeroMQ <aSocket>.recv()-methods by always using the zmq.NOBLOCK-modes thereof or rather test a presence or absence of any expected-message(s) with <aSocket>.poll( zmq.POLLIN, <timeout> )-methods available, with zero or controlled-timeouts. This makes you the master, who decides about the flow of code-execution. Not doing so, you knowingly let your code depend on external sequence ( or absence ) of events and your architecture is prone to awful problems with handling infinite blocking-states ( or potential unsalvageable many-agents' distributed behaviour live-locks or dead-locks )
Avoid uncontrolled cross-breeding of event-loops - like passing ZeroMQ-driven-loops into an external "callback"-alike handler or async-decorated code-blocks, where the stack of (non-)blocking logics may wreck havoc the original idea just by throwing the system into an unresolvable state, where events miss expected sequence of events and live-locks are unsalvagable or just the first pass happen to go through.
Stacking asyncio-code with twisted-LoopingCall()-s and async/await-decorated code + ZeroMQ blocking .recv()-s is either a Piece-of-Filligrane-Precise-Art-of-Truly-a-Zen-Master, or a sure ticket to Hell - with all respect to the Art-of-Truly-Zen-Masters :o)
So, yes, complex thinking is needed -- welcome to the realms of distributed-computing!
Is there a way to specify a "maximum" = inf timeout for a worker?
I have some long-running tasks and if something fails due to timeouts I handle it internally within the worker.
Can a specify this through the cli?
timeout argument specifies the maximum runtime of the task before it's considered 'lost'. Can be used with #job, Queue, enqueue & enqueue_call.
from rq.decorators import job
#job('low', connection=my_redis_conn, timeout=600)
def long_running_task(x, y):
# Code
python-rq.org/docs
Setting Queue(default_timeout=-1) will do the trick. Here is a reference to their source code:
def create_job(self, func, args=None, kwargs=None, timeout=None,
result_ttl=None, ttl=None, failure_ttl=None,
description=None, depends_on=None, job_id=None,
meta=None, status=JobStatus.QUEUED, retry=None):
"""Creates a job based on parameters given."""
timeout = parse_timeout(timeout)
if timeout is None:
timeout = self._default_timeout
elif timeout == 0:
raise ValueError('0 timeout is not allowed. Use -1 for infinite timeout')