JacksonTian

Node.js异步处理CPU密集型任务

Node.js擅长数据密集型实时（data-intensive real-time）交互的应用场景。然而数据密集型实时应用程序并不是只有I/O密集型任务，当碰到CPU密集型任务时，比如要对数据加解密（node.bcrypt.js），数据压缩和解压（node-tar），或者要根据用户的身份对图片做些个性化处理，在这些场景下，主线程致力于做复杂的CPU计算，I/O请求队列中的任务就被阻塞。

Node.js主线程的event loop在处理所有的任务/事件时，都是沿着事件队列顺序执行的，所以在其中任何一个任务/事件本身没有完成之前，其它的回调、监听器、超时、nextTick()的函数都得不到运行的机会，因为被阻塞的event loop根本没机会处理它们，此时程序最好的情况是变慢，最糟的情况是停滞不动，像死掉一样。

一个可行的解决方案是新开进程，通过IPC通信，将CPU密集型任务交给子进程，子进程计算完毕后，再通过ipc消息通知主进程，并将结果返回给主进程。

和创建线程相比，开辟新进程的系统资源占用率大，进程间通信效率也不高。如果能不开新进程而是新开线程，将CPU耗时任务交给一个工作线程去做，然后主线程立即返回，处理其他的I/O请求，等到工作线程计算完毕后，通知主线程并将结果返回给主线程。那么在同时面对I/O密集型和CPU密集型服务的场景下，Node的主线程也会变得轻松，并能时刻保持高响应度。

ahankinson

from rest_framework import mixins
from rest_framework.generics import SingleObjectAPIView
from yourapp.mixins import PartialUpdateModelMixin


class RetrievePartialUpdateDestroyAPIView(PartialUpdateModelMixin,
                                    mixins.RetrieveModelMixin,
                                    mixins.UpdateModelMixin,
                                    mixins.DestroyModelMixin,
                                    SingleObjectAPIView):

acu192

from django.conf import settings


def SessionCookieSameSiteWorkaround(get_response):
    def middleware(request):
        # See:
        #   https://blog.chromium.org/2019/10/developers-get-ready-for-new.html
        #   https://github.com/django/django/pull/11894
        #   https://web.dev/samesite-cookie-recipes/#handling-incompatible-clients
        #   https://www.chromium.org/updates/same-site/incompatible-clients

fkztw

#!/usr/bin/env expect

# Login into PTT via SSH.
# Combine crontab with this script so you don't forget to login PTT everyday.

# Usage: /bin/bash -c ${FILE_PATH_OF_THIS_SCRIPT}

# Example in crontab:
# ptt auto login, every day at 12:00 PM                                            
# 0 12 * * * /bin/bash -c ${ABSOLUTE_FILE_PATH_OF_THIS_SCRIPT}

dhbradshaw

To change a field name in django 1.7+

1.  Edit the field name in the model (but remember the old field name: you need it for step 3!)
2.  Create an empty migration
$ python manage.py makemigrations --empty myApp
3.  Edit the empty migration (it's in the migrations folder in your app folder, and will be the most recent migration) by adding

migrations.RenameField('MyModel', 'old_field_name', 'new_field_name'),

to the operations list.

msabramo

@contextlib.contextmanager
def multiple_targets(mock_patches):
    """
    `mock_patches` is a list (or iterable) of mock.patch objects

    Example usage:

        with mock.patch.multiple_targets([
                mock.patch('os.path.exists', side_effect=mock_path_exists),
                mock.patch('subprocess.Popen'),

cespare

package main

import (
	"log"
	"myserver"
	"net/http"
)

const addr = "localhost:12345"

itaysk

###
# There's a newer version available here:
# https://github.com/itaysk/kube-imagepuller
# All future updates will be made there.
# Please also post you questions as issues on that repo instead of commenting here
###


apiVersion: apps/v1beta2
kind: DaemonSet

innovia

#!/bin/bash
set -e
set -o pipefail

# Add user to k8s using service account, no RBAC (must create RBAC after this script)
if [[ -z "$1" ]] || [[ -z "$2" ]]; then
 echo "usage: $0 <service_account_name> <namespace>"
 exit 1
fi

mmellison

import asyncio
from concurrent import futures
import functools
import inspect
import threading

from grpc import _server

def _loop_mgr(loop: asyncio.AbstractEventLoop):