hyperdev.fr

What does a web application architecture include?

And what are the code patterns?

In this note, I will describe and introduce concepts related to web developpement from the big picture to code patterns.

This article links to RFCs and man pages, it's a reminder for me to read them one day.

Big pictures

At the very beginning you type an Uniform Resource Locator aka. URL RFC 1738 in the location bar. Something like http://hyperdev.fr/.

The browser will request the Internet Protocol address aka. IP RFC 791 of the URL using a Domain Name Service RFC 1035 prolly configured in /etc/resolv.conf.

You query the DNS service using the dig command:

amirouche@ubudec:~$ dig hyperdev.fr

; <<>> DiG 9.10.3-P4-Ubuntu <<>> hyperdev.fr
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 2891
;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 65494
;; QUESTION SECTION:
;hyperdev.fr.           IN  A

;; ANSWER SECTION:
hyperdev.fr.        2952    IN  A   92.222.88.65

;; Query time: 0 msec
;; SERVER: 127.0.0.53#53(127.0.0.53)
;; WHEN: Mon Nov 06 20:24:48 CET 2017
;; MSG SIZE  rcvd: 56

You decypher this without knowing DNS or whatever. There is only one IP in the output.

At this point, the browser with the IP and the URL will construct a HyperText Transfer Protocol aka. HTTP/1.1 RFC 2616. This HTTP packet is a request for the page you want.

You can simulate what the browser does at this point with the nc command. For instance:

amirouche@ubudec:~$ nc 92.222.88.65 80

Then paste the following in the console:

GET / HTTP/1.1
Host: hyperdev.fr

Mind the last blank line.

The output will start with something like that:

HTTP/1.1 200 OK
Server: nginx/1.10.3
Date: Mon, 06 Nov 2017 19:27:02 GMT
Content-Type: text/html
Content-Length: 3875
Last-Modified: Sat, 19 Aug 2017 13:31:40 GMT
Connection: keep-alive
ETag: "59983dbc-f23"
Accept-Ranges: bytes

<html><head><meta charset="utf-8" /><title>hyperdev.fr</title><link rel="stylesheet" href="/static/css/normalize.css" /><link rel="stylesheet" href="/static/css/app.css" /></head><body><div>
...

And even more HTML code. This is the server response.

At his point the browser will parse the response and prolly request more files from the server and trigger at the same time the rendering of the page [TODO: link how a browser render a page].

But a question remains, who answered the HTTP request? Well the answer is: it depends. It depends whether there is cache or some other kind of proxy between the browser and the server.

Also, nowdays HTTP requests happens most of the time over a secure layer sometime called SSL but it's more likely using Transport Layer Security protocol aka. TLS RFC 2246.

Thanks to Let's Encrypt everybody can use TLS for free.

To keep things simple we will imagine that there is no proxy and you directly access the application server.