00:00 - 00:03
have you ever wondered what really
00:01 - 00:05
happens behind the scenes when you click
00:03 - 00:07
a button on a website today we are
00:05 - 00:09
following a single request through the
00:07 - 00:11
entire architecture of a modern web
00:09 - 00:13
application so you can understand
00:11 - 00:16
exactly what happens from DNS resolution
00:13 - 00:17
to backend databases by the end of this
00:16 - 00:19
video you will have a deeper
00:17 - 00:23
appreciation for all the moving parts
00:19 - 00:23
that make your favorite apps run
00:25 - 00:30
smoothly when a user types example.com
00:27 - 00:33
into their browser the DNS acts as the
00:30 - 00:35
internet's phone book translating the
00:33 - 00:37
domain into an IP address this IP
00:35 - 00:39
address could point to a load balancer
00:37 - 00:41
or a AWS cloudfront distribution
00:39 - 00:44
depending on the architecture quick DNS
00:41 - 00:46
resolution is crucial because it helps
00:44 - 00:49
reduce initial load times ensuring users
00:46 - 00:51
can access your site quickly so if
00:49 - 00:53
you're using AWS Route 53 it helps you
00:51 - 00:56
direct the user to the closest available
00:53 - 00:59
server ensuring fast and efficient
00:56 - 01:02
routing next up is the content delivery
00:59 - 01:05
Network or CDN services like AWS Cloud
01:02 - 01:08
front cash static assets images videos
01:05 - 01:10
and CSS files at Edge locations closer
01:08 - 01:12
to your users this helps reduce latency
01:10 - 01:14
and speeds up load times especially for
01:12 - 01:16
users spread out across different
01:14 - 01:18
regions and in today's world where even
01:16 - 01:20
a 1second delay can lead to drop offs
01:18 - 01:22
speed is everything not only does
01:20 - 01:25
cloudfront improve speed but it also
01:22 - 01:27
helps reduce cost by offloading request
01:25 - 01:29
from your main server especially when
01:27 - 01:31
handling Global traffic before the
01:29 - 01:33
request reaches your backend Services it
01:31 - 01:36
passes through web application firewall
01:33 - 01:38
or vaav a vaav is your application's
01:36 - 01:40
first line of defense imagine someone
01:38 - 01:42
trying to hack into your system using
01:40 - 01:44
SQL injection or cross- siiz scripting
01:42 - 01:46
attacks VAV stops those attacks in their
01:44 - 01:48
tracks by filtering out malicious
01:46 - 01:50
traffic before it even hits your backend
01:48 - 01:52
Services after passing through VAV the
01:50 - 01:54
request reaches enginex which serves as
01:52 - 01:56
a reverse proxy it routes traffic
01:54 - 01:59
handles SSL termination and caches
01:56 - 02:02
Dynamic content now while engin X offers
01:59 - 02:04
fine grain control but services like
02:02 - 02:06
cloudfront or ALB can handle many of
02:04 - 02:08
these tasks making engine X an optional
02:06 - 02:10
component in certain architectures
02:08 - 02:13
however when you need more control over
02:10 - 02:16
routing caching or SSL termination
02:13 - 02:18
engine X becomes a valuable addition now
02:16 - 02:20
the load balancer ensures availability
02:18 - 02:23
by Distributing incoming traffic across
02:20 - 02:25
multiple backend servers with NLB or
02:23 - 02:27
network load balancer you get low
02:25 - 02:28
latency High throughput handling for
02:27 - 02:30
applications that need Extreme
02:28 - 02:33
Performance like gaming or Financial
02:30 - 02:34
Services on the other hand an ALB or
02:33 - 02:37
application load balancer gives you
02:34 - 02:39
Advanced routing capabilities ideal for
02:37 - 02:41
web apps and microservices where
02:39 - 02:43
requests need to be routed based on
02:41 - 02:45
specific paths or host names both
02:43 - 02:46
Network load balancer and application
02:45 - 02:49
load balancer operate at different
02:46 - 02:51
layers of OSI models previously I have
02:49 - 02:53
made a video on the OSI and TCP layer
02:51 - 02:55
breaking down and explaining each of
02:53 - 02:57
these layers with practical examples NLB
02:55 - 02:59
operates at the transport layer which is
02:57 - 03:01
the layer four of the OSI model this
02:59 - 03:04
means it forwards traffic based on the
03:01 - 03:05
IP address and port number without
03:04 - 03:08
inspecting the content of the request
03:05 - 03:11
for example HTTP headers or paths it is
03:08 - 03:14
designed to handle protocols like TCP or
03:11 - 03:16
transmission control protocol and UDP
03:14 - 03:17
which stands for user datagram protocol
03:16 - 03:19
which makes it suitable for high
03:17 - 03:21
performance low latency applications
03:19 - 03:23
handling very large amounts of traffic
03:21 - 03:26
with millions of requests per second
03:23 - 03:27
while maintaining very low latency this
03:26 - 03:29
makes it ideal for real-time
03:27 - 03:31
applications such as gaming or services
03:29 - 03:33
your speed and throughput are critical
03:31 - 03:35
ALB operates at the application layer
03:33 - 03:37
which is the layer seven of the OSI
03:35 - 03:40
model which means it can inspect the
03:37 - 03:43
content of HTTP or https request making
03:40 - 03:45
it suitable for web application rest
03:43 - 03:47
apis or microservices architecture it
03:45 - 03:49
can Route traffic based on attributes
03:47 - 03:52
like request paths for example routing
03:49 - 03:54
based on specific API paths like API
03:52 - 03:56
users or API products or it can also
03:54 - 03:59
drout traffic based on attributes such
03:56 - 04:02
as host name for example a request to
03:59 - 04:05
example ./ API can be routed to a
04:02 - 04:06
backend API service or a request to
04:06 - 04:11
static can be routed to a different
04:08 - 04:13
service that handles static content the
04:11 - 04:15
API Gateway acts like a traffic cop
04:13 - 04:18
validating request ensuring the user is
04:15 - 04:20
authorized and enforcing rate limits
04:18 - 04:22
rate limiting and throttling prevent any
04:20 - 04:24
single client from overwhelming your
04:22 - 04:26
system keeping everything running
04:24 - 04:29
smoothly after this the request is
04:26 - 04:31
routed to the right backend service
04:29 - 04:33
whether that's an easy to instance a AWS
04:31 - 04:35
Lambda function or a microservice
04:33 - 04:38
running in kubernetes once the API gate
04:35 - 04:41
fit does its job it's up to the backend
04:38 - 04:43
services to handle the heavy lifting now
04:41 - 04:45
in microservices architecture each
04:43 - 04:47
service is dedicated to a specific
04:45 - 04:50
function allowing teams to work
04:47 - 04:51
independently scale Services more easily
04:50 - 04:54
and make updates without impacting the
04:51 - 04:56
entire system for instance one service
04:54 - 04:58
might handle user authentication while
04:56 - 05:01
another process payments with each
04:58 - 05:03
service able to scale or up separately
05:01 - 05:06
all while interacting with databases
05:03 - 05:08
like post SQL or Dynamo TV speaking of
05:06 - 05:10
databases they are the heart of your
05:08 - 05:12
application where your persistent data
05:10 - 05:14
is stored for distributed applications
05:12 - 05:16
that need to scale dynamically a nosql
05:14 - 05:18
database like Dynamo DV works best due
05:16 - 05:21
to its flexibility and ability to handle
05:18 - 05:23
unstructured data meanwhile for
05:21 - 05:25
application with structured relational
05:23 - 05:27
data and SQL database like post SQL is
05:25 - 05:30
often the best choice for asynchronous
05:27 - 05:32
task request might be passed to HQ like
05:30 - 05:34
Kafka or sqs which handles STS that
05:32 - 05:37
don't need immediate processing such as
05:34 - 05:38
email notification or background jobs in
05:37 - 05:40
addition to handling request your
05:38 - 05:43
application likely uses additional
05:40 - 05:47
layers like authentication using or 2.0
05:43 - 05:50
or open ID connect or JWT basically to
05:47 - 05:52
verify users caching can be used to
05:50 - 05:54
speed up responses and logging for
05:52 - 05:57
tracking system Health Services like
05:54 - 05:58
redis or mcast can catch frequently
05:57 - 06:00
requested data reducing the load on
05:58 - 06:02
backend services
06:00 - 06:04
monitoring tools like promas or AWS
06:02 - 06:06
Cloud watch ensure your system is
06:04 - 06:08
running smoothly by tracking metrics and
06:06 - 06:09
identifying issues before the escalate
06:08 - 06:11
now that your backend services and
06:09 - 06:14
additional components have done their
06:11 - 06:16
job it's time for the response to make
06:14 - 06:18
its way back to the user this journey is
06:16 - 06:20
just as important as the request itself
06:18 - 06:22
and it follows a similar path back
06:20 - 06:24
through the system so the process
06:22 - 06:26
response is sent back to the API Gateway
06:24 - 06:28
which may apply additional formatting or
06:26 - 06:30
Transformations before routing it to the
06:28 - 06:32
client the load balancer efficiently
06:30 - 06:35
routes the response back to the user's
06:32 - 06:37
original request balancing any outgoing
06:35 - 06:39
traffic if necessary and if any static
06:37 - 06:41
content was part of the response such as
06:39 - 06:43
images or cach data it can be served
06:41 - 06:46
directly from a CDN Edge location to
06:43 - 06:48
minimize latency ensuring a quicker
06:46 - 06:50
response finally the response reaches
06:48 - 06:52
the users browser or mobile app
06:50 - 06:54
displaying the requested data completing
06:52 - 06:56
the full request response cycle in a
06:54 - 06:58
well-designed system this entire Journey
06:56 - 07:00
from the user's request to the final
06:58 - 07:03
response happens milliseconds ensuring
07:00 - 07:04
fast reliable performance from DNS
07:03 - 07:06
resolution to the database and
07:04 - 07:08
everything in between every layer plays
07:06 - 07:10
a crucial role in delivering a smooth
07:08 - 07:12
fast experience to users understanding
07:10 - 07:13
both the request and response life cycle
07:12 - 07:16
gives you a solid foundation for
07:13 - 07:17
Designing resilient scalable systems let
07:16 - 07:19
me know in the comments if you'd like me
07:17 - 07:21
to dive deeper into any of these topics
07:19 - 07:23
and don't forget to like subscribe and
07:21 - 07:25
hit the notification Bell for more deep
07:23 - 07:28
dives into Cloud architecture and system
video to article