$30 off During Our Annual Pro Sale. View Details »
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Beyond HTTP, Breaking Free of The Web
Search
Mark Wolfe
February 09, 2015
Technology
0
330
Beyond HTTP, Breaking Free of The Web
Talk I gave at APIdays in Sydney 2015 and Microservices Meetup
Mark Wolfe
February 09, 2015
Tweet
Share
More Decks by Mark Wolfe
See All by Mark Wolfe
IP Networking in AWS
wolfeidau
0
73
Getting Into Vue.js
wolfeidau
0
190
AWS Automation With Lambda
wolfeidau
1
92
Ansible and Cloudformation
wolfeidau
1
230
Docker Security
wolfeidau
0
200
Building a Proxy in Go
wolfeidau
2
380
Using JWT to Authenticate Microservices
wolfeidau
1
250
Managing Secrets IN AWS
wolfeidau
0
150
ESPlant Workshop Firmware Overview and Demos
wolfeidau
0
140
Other Decks in Technology
See All in Technology
regrowth_tokyo_2025_securityagent
hiashisan
0
160
21st ACRi Webinar - Univ of Tokyo Presentation Slide (Ayumi Ohno)
nao_sumikawa
0
120
Playwrightのソースコードに見る、自動テストを自動で書く技術
yusukeiwaki
13
4.7k
プロダクトマネージャーが押さえておくべき、ソフトウェア資産とAIエージェント投資効果 / pmconf2025
i35_267
2
580
ログ管理の新たな可能性?CloudWatchの新機能をご紹介
ikumi_ono
0
460
【CEDEC+KYUSHU2025】学生・若手必見!テクニカルアーティスト 大全 ~仕事・スキル・キャリアパス、TAの「わからない」を徹底解剖~
cygames
PRO
0
140
Oracle Database@Google Cloud:サービス概要のご紹介
oracle4engineer
PRO
0
690
グレートファイアウォールを自宅に建てよう
ctes091x
0
140
Noを伝える技術2025: 爆速合意形成のためのNICOフレームワーク速習 #pmconf2025
aki_iinuma
2
2k
RAG/Agent開発のアップデートまとめ
taka0709
0
130
Gemini でコードレビュー知見を見える化
zozotech
PRO
1
180
品質のための共通認識
kakehashi
PRO
1
190
Featured
See All Featured
CoffeeScript is Beautiful & I Never Want to Write Plain JavaScript Again
sstephenson
162
15k
Code Review Best Practice
trishagee
74
19k
GraphQLの誤解/rethinking-graphql
sonatard
73
11k
Practical Tips for Bootstrapping Information Extraction Pipelines
honnibal
25
1.6k
What’s in a name? Adding method to the madness
productmarketing
PRO
24
3.8k
Rebuilding a faster, lazier Slack
samanthasiow
84
9.3k
Balancing Empowerment & Direction
lara
5
790
Documentation Writing (for coders)
carmenintech
76
5.2k
Helping Users Find Their Own Way: Creating Modern Search Experiences
danielanewman
31
3k
実際に使うSQLの書き方 徹底解説 / pgcon21j-tutorial
soudai
PRO
196
69k
Reflections from 52 weeks, 52 projects
jeffersonlam
355
21k
A designer walks into a library…
pauljervisheath
210
24k
Transcript
Beyond HTTP, Breaking Free of The Web Mark Wolfe
–From http://www.w3.org/DesignIssues/Architecture.html “The most fundamental specification of Web architecture, while
one of the simpler, is that of the Universal Resource Identifier, or URI.”
–From https://tools.ietf.org/html/rfc2616 “The Hypertext Transfer Protocol (HTTP) is an application
protocol for distributed, collaborative, hypermedia information systems.”
– From http://en.wikipedia.org/wiki/Application_programming_interface “In computer programming, an application programming interface
(API) is a set of routines, protocols, and tools for building software applications.”
APIs We Use != http DNS Used to resolve host
names Also used for service discovery MDNS SPDY
Wireshark Network protocol analyser Our window into the wonderful world
of protocols and APIs Available for Windows, OSX and Linux https://www.wireshark.org/ The wire doesn’t lie
DNS? Protocol uses UDP or TCP for transport Consists of
Questions and Answers Includes some information on authority Uses "intelligent retransmission” Small messages
DNS Continued.. Binary protocol Fixed header Payload uses a compact
self describing binary format TLV (type length value)
DNS Question
DNS Answer
DNS Exchange
HTTP Get Request
In Review Compact messages More likely to transit “wireless” networks
Extensible self describing binary payload Established encoding so others can introspect without a definition! Fast
MDNS Uses UDP Over Multicast Designed for resolving names in
adhoc networks Uses the existing DNS message structure, name syntax, and resource record types
Something Simpler? So lets look at something simpler than DNS
What if we just put text in a UDP packet What would that look like?
Problem I want to send 10s of 1000s of metrics
to a server for aggregation with minimal performance impact I want to keep it simple I want the server to be a staging area for data Perform Aggregation Flush to timeseries at regular intervals
statsd Developed by etsy Uses UDP as a transport Data
is encoded using text Metrics are either a Gauge, Counter, Timer or Meter No reply necessary!
statsd protocol Metrics are generally of the form <metric name>:<value>|<type>
outside_temp|12.2|g get_home_request_time|123|ms get_home_request|1|c
In Review Very simple implementation Doesn’t block Fast Loss doesn’t
really matter in the scheme of things Maths happens in the server
None
The End @wolfeidau on twitter https://github.com/wolfeidau
[email protected]
http://wolfe.id.au
Links Good article on DNS message format https://www.wireshark.org/ Read this
book High Performance Browser Networking Etsy Statsd