Route 53¶

only service provide : 100% availability
fact:: 53 is reference tradition DNS port

A. Key concept¶

DNS¶

registrar : dns record stores in zone file
Naming Server : resolve dns queries

TTL¶

client cache the result from DNS server, for that that period locally.
hence wont query again till TTl expire.
eg: 24 hr, 60 sec
Domain/subdomain Name (example.com) + value (11.22.33.44)

namespace:¶

root domain .
top level domain TLD
second level domain SLD
sub domain
FQDN
URL
- SLD/TLD === zone apex (eg: covid19survey.com)
- cannot create CNAME for covid19survey.com

DNS Records¶

A : domain -> IPv4
AAAA : domain -> Ipv6
CNAME : domain -> doanother Domain (except root/TLD)
eg: xxxx.us-east-1.alb.com --> my-elb.com
NS : domain -> Name Server
Alias (AWS specific)
benefit : free + native health-check
eg:
- domain --> aws resources
- domain --> root domain + TLD (works)
- domain --> another record in same hosted zone
- domain --> CAN NOT : ec2-instance-DN
- domain --> CAN NOT :www.google.com (not aws resource)
Configuring TTL not allowed, set bt R53
Alias === [ A or AAAA + flag:alias=True ]
advanced: [ CAA / DS / MX / NAPTR / PTR / SOA / TXT / SPF / SRV ] - skip
PTR: resolves IP address to FQDN

AWS :: Hosted Zone¶

cost : .50/m
Container for DNS Record/s
define traffic routing for its domain, its sub-domains
Type
public
- internet traffic, resolves to public IP
private ( domainName eg: internal.example.com)
- internal traffic within VPC, resolves to private IP
- internal DNS management, secured, scalable.
- if associated (1-2-M) it with non-default VPC/s, then need to enable (disabled by default):
- DNS hostnames / enable DnsHostnames
- DNS resolution / enable DnsSupport

C. R53::`Routing policy`¶

simple : domain -> [ip-1, domain-2, ...] multiple target,
randomly choose 1
irrespective of health
weighted : domain -> [ip-1::Weight10, domain-2::Weight70, ...]
choose by weight
Latency : domain -> [ip-1::region-1, domain-2:region-2, ...] multiple target,
AWS will look for latency and pick one with low value.
failover : domain -> [ip-1::Primary, domain-2::Seconday, ...] multiple target,
primary until healthy, else secondary.
Geolocation : domain -> [ip-1::Country-1, domain-2::Country-2, domain-3::default,...] multiple target,
from client get Country.
Geo-proximity : domain -> [ip-1::bais-0, domain-2::bais-50, ...] multiple target,
shift more traffic to domain-2
Route traffic to your resources based on the geographic location of users and resources
create with UI for better visuality.
baise: -99 to 99
IP based : domain -> [ip-1::Client-IP-range/CIDR-1, domain-2::Client-IP-range/CIDR-2, ...]
choose by client_IP
Multi value : domain -> [ip-1, domain-2, ...] multiple target, randomly choose 1 (healthy)
similar to simple.
randomly choose 1 but healthy only
3rd party: route by goDaddy policy.
link goDaddy with R53,
just by updating NS record in R53 hosted-zone(public), with goDaddy DNs.
can use R53 then.

hands on¶

https://us-east-1.console.aws.amazon.com/route53/trafficflow/home?region=us-east-2#/

create from UI, :)

- launch Ec2-i1, i2, and i3  (in az1, az2,az3), with simple webapp
- create ALB
    - `alb-1-dns` : hostname/domainname
    - choose az1, az2, az3
    - create TG (tg-1)
        - sg : allow all HTTPs traffic
        - AZ1, az2, az3
        - regsister Ec2-i1, i2, and i3
    - routing and filter: 
        - Listener-1::No-contion : outside traffic on `http:80`  --> forward to --> `tg-1` 

- create hosted Zone  [Zone : hz.com ]
  - NS hz.com -> *****, ****, ****, **** some entry comes automatically once HZ created : alrady present
  - A + alias:F, ec2-1.com ->   ec2-1 public IP  
  - A + alias:F, ec2-2.com ->   ec2-2 public IP  
  - A + alias:F, ec2-3.com ->   ec2-3 public IP  
  - CNAME        abl.com --> alb-1-dns
  - A + alias:T, abl-alias.com --> now can provide another domain --> alb-1-dns ; FREE
  - CNAME        hz.com -->   alb-1-dns : ERROR
  - A + alias:T, hz.com --> alb-1-dns : WORKS

  // Rourting policies
  // SIMPLE
  -  record-id-1, A + alias:F, webapp1.com ->   ec2-1 public IP  SIMPLE 
  -  record-id-2, A + alias:F, webapp1.com ->   ec2-2 public IP  SIMPLE 
  -  record-id-3, A + alias:F, webapp1.com ->   ec2-3 public IP  SIMPLE
  // WEIGHTED
  -  record-id-11, A + alias:F, webapp2.com ->   ec2-1 public IP  WEIGHTED 10
  -  record-id-21, A + alias:F, webapp2.com ->   ec2-2 public IP  WEIGHTED 20
  -  record-id-33, A + alias:F, webapp2.com ->   ec2-3 public IP  WEIGHTED 70
  // LATENCY
  -  record-id-111, A + alias:F, webapp3.com ->   ec2-1 public IP  LATENCY us-east-1
  -  record-id-222, A + alias:F, webapp3.com ->   ec2-2 public IP  LATENCY us-east-1
  -  record-id-333, A + alias:F, webapp3.com ->   ec2-3 public IP  LATENCY us-west-2
  // failover
  -  record-id-1111, A + alias:F, webapp4.com ->   ec2-1 public IP  failover "primary",  R53:healthCheck-1
  -  record-id-2222, A + alias:F, webapp4.com ->   ec2-2 public IP  failover "seconday", R53:healthCheck-2 (optional)
  // Geolocation
  -  record-id-1111, A + alias:F, webapp5.com ->   ec2-1 public IP  Geolocation "ASIA",  
  -  record-id-2222, A + alias:F, webapp5.com ->   ec2-2 public IP  Geolocation "EUROPE"
  // Geo-proximity
  -  record-id-1111, A + alias:F, webapp6.com ->   ec2-1 public IP  Geo-proximity Bias-10  
  -  record-id-2222, A + alias:F, webapp6.com ->   ec2-2 public IP  Geo-proximity Bias-50
  // Multi-value
  -  record-id-1111, A + alias:F, webapp7.com ->   ec2-1 public IP Multi-value  R53:healthCheck-1 
  -  record-id-2222, A + alias:F, webapp7.com ->   ec2-2 public IP Multi-value  R53:healthCheck-1

- acess webapp:
  - diectly from ec2-i public IPS
  - add (A + alias:F) record and access
  - via ELB 
    - alb-1-dns
    - abl-alias.com
  - hz.com
  - Rourting policies demo : webapp[1,2,3,...7].com

D. R53 :: `health Check`¶

create health heck and use in DNS entries.

a. health checker : `Endpoint`¶

update sg to allow: runs outside private VPC
provide health endpoints.
set interval : 10 sec or 30 sec
eg: R53 alais record
point to ALB
health check point to ALB (Evaluated Health Check = True)

b. health checker : `Caculated`¶

choose : state of other health checks
AND OR NOT, etc. mix amd match.

c. health checker : `CW Alarm state`¶

create CW alarm, first.
select it. while creating R53 health check

E. Scenario¶

1. resolve Outside AWS dns ¶

A retail company has connected its on-premises data center to the AWS Cloud via AWS Direct Connect.
The company wants to be able to resolve DNS queries
for any resources in the on-premises from the AWS VPC
for any resources in the AWS VPC from the on-premises
solution:
Create an outbound endpoint on Route 53 Resolver
- and then Amazon Route 53 Resolver
- can conditionally forward DNS queries to on-premises resolvers, via this endpoint.
Create an inbound endpoint on Route 53 Resolver
- and then on-premises DNS resolvers
- can conditionally forward DNS queries to Route 53 Resolver, via this endpoint.

Route 53¶

A. Key concept¶

DNS¶

TTL¶

namespace:¶

DNS Records¶

AWS :: Hosted Zone¶

C. R53::Routing policy¶

hands on¶

D. R53 :: health Check¶

a. health checker : Endpoint¶

b. health checker : Caculated¶

c. health checker : CW Alarm state¶

E. Scenario¶

1. resolve Outside AWS dns ¶

C. R53::`Routing policy`¶

D. R53 :: `health Check`¶

a. health checker : `Endpoint`¶

b. health checker : `Caculated`¶

c. health checker : `CW Alarm state`¶