acolyer

References from my QCon London 2016 Keynote

• The Morning Paper blog

Thinking Tools

• Scalability! But at what COST?
• Tachyon: Reliable, Memory Speed Storage for Cluster Computing
• Musketeer - Part I: What's the best data processing system?
• Musketeer - Part II: all for one, and one for all in data processing systems

rohityadavcloud

[email protected]
cfkey=cf-key-here
zoneid=some-zoneid
dnsrec=dns-record-id
domain=somedomain.com

curl -X PUT "https://api.cloudflare.com/client/v4/zones/$zoneid/dns_records/$dnsrec" \
-H "X-Auth-Email: $cfuser" \
-H "X-Auth-Key: $cfkey" \
-H "Content-Type: application/json" \

GABeech

# This is an example of the Stack Exchange Tier 1 HAProxy config
# The only things that have been changed from what we are running are: 
# 1. User names have been removed
# 2. All Passwords have been remove
# 3. IPs have been changed to use the example/documentation ranges
# 4. Rate limit numbers have been changed to randome numbers, don't read into them

userlist stats-auth
        group admin users $admin_user
        user $admin_user insecure-password $some_password

cryptix

package main

import (
	"io"
	"log"
	"mime/multipart"
	"net/http"
	"os"
	"path/filepath"
	"runtime"

leverich

# Systemtap script to trace latency of concurrent memcached requests
# as they percolate through the Linux kernel.
# - Jacob Leverich <[email protected]> 2013
#
# Note: You MUST call this script with "-x <TID>" where TID is the
# *task id* of a memcached worker thread. This is usually PID+1 of
# memcached. So do something like:
#
#    % memcached -t 1 -d
#    % pgrep memcached

vishvananda

#!/bin/bash

if [ "$4" == "" ]; then
    echo "usage: $0 <local_ip> <remote_ip> <new_local_ip> <new_remote_ip>"
    echo "creates an ipsec tunnel between two machines"
    exit 1
fi

SRC="$1"; shift
DST="$1"; shift

audreyt

從 SocialCalc 到 EtherCalc

先前在《開源應用程式架構》 一書中，我介紹了 SocialCalc 這個在瀏覽器中運行的試算表編輯器，以取代伺服器為中心的 WikiCalc 架構。SocialCalc 在瀏覽器中執行所有的運算，只有在載入和儲存試算表時才會使用伺服器。

追求效能是 Socialtext 團隊在 2006 年時設計 SocialCalc 的主要目的。重點在於：在 JavaScript 環境下執行客戶端運算，儘管在當年的速度僅有伺服器端 Perl 運算的十分之一，但仍然勝過 AJAX 來回傳輸資料造成的網路延遲：

---

******

giannitedesco

#!/usr/bin/python

class TokenBucket:
	from time import sleep, time

	def __init__(self, rate, burst = None):
		'Rate in tokens per second, burst in number of tokens'

		self.rate = float(rate)
		if burst is None:

tumf

# ./vif eth1 10.10.1.1
cat <<EOF > /etc/sysconfig/network-scripts/ifcfg-$1
DEVICE=$1
BOOTPROTO=static
NETMASK=255.255.255.0
IPADDR=$2
ONBOOT=yes
EOF
cat /etc/sysconfig/network-scripts/ifcfg-$1

eqhmcow

#!/bin/bash

# As the "bufferbloat" folks have recently re-discovered and/or more widely
# publicized, congestion avoidance algorithms (such as those found in TCP) do
# a great job of allowing network endpoints to negotiate transfer rates that
# maximize a link's bandwidth usage without unduly penalizing any particular
# stream. This allows bulk transfer streams to use the maximum available
# bandwidth without affecting the latency of non-bulk (e.g. interactive)
# streams.