AWS Builder Center

Builder Center

EC2 Auto Scaling in Tanglish: Festival Sale Example

EC2 Auto Scaling in Tanglish: Festival Sale Example

Auto Scaling na enna? Festival sale-la traffic athigam aanaa servers thaanaa add aagum, kammi aanaa remove aagum. AWS-la epdi nu Tanglish-la paapom.

Tharsan

Introduction

Namma online cloth shop-la festival sale varudhu. Normal naal-la 100 per website paarpaanga. Aana sale naal-la 10,000 per ore neram-la varuvaanga.
Ippo rendu problem:
  • Servers kammi-a irundhaa, website slow aagum, illa crash aagum. Customers poiduvaanga.
  • Servers niraiya vechirundhaa, sale mudinja piragu kooda paisa pogum.
Idhukku solution dhaan EC2 Auto Scaling. Indha article-la real example oda Tanglish-la paapom.

Auto Scaling na enna?

Auto Scaling na traffic-ku thagundha maadhiri servers-a thaanaave add pannum, remove pannum.
  • Traffic athigam aanaa: pudhu servers add aagum (scale out)
  • Traffic kammi aanaa: extra servers remove aagum (scale in)
Idha restaurant-la nenaichu paarunga. Normal naal-la 2 waiters podhum. Festival naal-la customers niraiya varuvaanga, so extra waiters-a call pannuvaanga. Kootam kammi aanadhum extra waiters-a veetuku anuppiduvaanga. Idhu dhaan Auto Scaling.

Rendu vagai scaling

VagaiArthamExample
Horizontal (Scale out/in)Servers-oda ennikkaiyai maathuradhu2 servers-a 6 servers-aa aakkuradhu
Vertical (Scale up/down)Oru server-oda size-a maathuradhut3.small-a t3.large-aa maathuradhu
EC2 Auto Scaling horizontal scaling pannum.

Mukkiyamaana 3 numbers

Auto Scaling Group (ASG) create pannum bodhu 3 numbers set pannuveenga:
SettingArthamNamma shop example
MinimumKammi-pakshama irukka vendiya servers2
DesiredIppo irukka vendiya servers2
MaximumEvlo vara allow pannuvom (upper limit)6
Maximum romba mukkiyam. Edhuvo bug-naala servers-a aayirakkanakka-la create aagama, bill-a kaapaathum.

Real-time example: Festival sale timeline

NeramTrafficEnna nadakkumServers
Kaalai 9 maniNormalMinimum-la run aagum2
Madhiyam 12 maniSale arambam, athigamCPU high aagudhu, pudhu servers add aagum4
Maalai 6 maniPeakInnum add aagum, maximum varaikkum6
Iravu 11 maniTraffic kuraiyudhuExtra servers remove aagum3
Mudhal naal kaalaiNormalThirumbi minimum-ku2
Website fast-a irukkum, sale mudinja piragu extra servers-ku paisa pogaadhu.

Architecture

1
2
3
4
5
6
Customers --> Application Load Balancer --> Auto Scaling Group
|-- EC2 server 1
|-- EC2 server 2
|-- ... (thevai-na add aagum)

CloudWatch (CPU paakkum) --> Scaling policy --> ASG-ku "add / remove" nu solluum
  • Load balancer customers-a ellaa servers-kum pirichi anuppum
  • CloudWatch CPU madhiri metrics-a paakkum
  • Scaling policy "ippo add pannu / remove pannu" nu decide pannum

Moondru scaling method

1. Target Tracking (beginners-ku best)

Oru target sollunga, example: "average CPU 50%-la irukkanum". CPU 50%-ku mela pona servers add aagum, kammiyaa pona remove aagum. Thermostat madhiri, neenga temperature sollunga, AC thaanaave velai seiyum.

2. Scheduled Scaling

Festival epo nu mundhaiye theriyum na, advance-a schedule pannalaam. Example: "Sale arambam aagura 11:30 mani-ku minimum-a 4-aa maathu". Ippadi pannina traffic varum munnaadhiye servers ready-a irukkum.

3. Step Scaling

CloudWatch alarm adhigam aagura alavukku thagundha madhiri add pannum. Example: CPU 60% na 1 server add, 80% na 3 servers add. Idhu konjam advanced.
Namma shop-ku best combination: Target Tracking + Scheduled Scaling.

Hands-on: Simple Auto Scaling setup

Mudhalla ennena venum
  • Oru Launch template (AMI, instance type, security group, app start aagura script)
  • Oru Application Load Balancer + target group
  • Rendu subnets (rendu Availability Zones-la)
Steps
  1. EC2 console-la Launch template create pannunga. Instance type (example: t3.micro), AMI, security group, user data script kudunga.
  2. Auto Scaling Groups → Create Auto Scaling group click pannunga.
  3. Launch template-a select pannunga.
  4. Rendu Availability Zones-la subnets select pannunga (reliability-ku).
  5. Load balancer-a attach pannunga, target group select pannunga.
  6. ELB health checks enable pannunga, so unhealthy server-a replace pannum.
  7. Capacity set pannunga: Min 2, Desired 2, Max 6.
  8. Scaling policy-la Target tracking select panni, Average CPU utilization = 50 nu vainga.
  9. Create pannunga.
Menu names maaritte irukkum, so current AWS docs-a paarthu follow pannunga.

Test epdi pannuradhu?

  1. ASG-la 2 instances run aagudha nu check pannunga.
  2. Oru test tool vechu load anuppunga (example: Apache Bench, hey).
  3. CloudWatch-la CPU eruvadhai paarunga.
  4. ASG-oda Activity tab-la pudhu instances launch aagudha nu paarunga.
  5. Load nirutthina piragu konja neram kazhichi instances kammi aagudha nu paarunga.
Scale in konjam late-a nadakkum. Adhu normal, thirumba thirumba up/down aagaama irukka AWS purposely wait pannum.

Auto Scaling velai seiya, app-la ethu irukkanum?

1. Stateless-a irukkanum
Server memory-la user session vechirundhaa, server remove aanaa user logout aagidalaam. Sessions-a DynamoDB / ElastiCache madhiri shared place-la vainga.
2. Server start aagura neram kammi
Pudhu server seekiram ready aaganum. Idhukku nalla AMI, user data script vechukkonga. Illa na traffic varum bodhu servers ready aagaama problem aagum.
3. Data server-kulla store panna vendaam
Orders, images ellaam DynamoDB / RDS / S3-la vainga. Server remove aanaa data poidum.

Cost patthi

  • Auto Scaling Group-kae separate charge illai nu naan puriyudhu. Run aagura EC2 instances, load balancer ku dhaan paisa. Exact details-a AWS pricing page-la check pannikonga.
  • Maximum limit vainga, bill-a control pannum.
  • Budget and billing alert set pannunga.
  • Sale mudinja piragu minimum-ku thirumbi varudha nu check pannunga.

Common mistakes

  • Maximum set pannama vitturadhu
  • Server start aagura neram romba adhigam irukkuradhu
  • Sessions-a server-kulla store pannuradhu
  • Health check illama irukkuradhu, so bad server-a replace aagaadhu
  • Festival-ku mundhaiye test pannama irukkuradhu
  • Sale mudinja piragu ASG settings-a thirumba paakkaama vittuduradhu

Seekiram checklist

  • [ ] Launch template ready-a?
  • [ ] Min, Desired, Max set pannirukkeengala?
  • [ ] Rendu Availability Zones-la servers irukka?
  • [ ] ELB health check on-aa?
  • [ ] Target tracking policy (CPU 50% madhiri) set pannirukkeengala?
  • [ ] Festival schedule-ku Scheduled scaling add pannirukkeengala?
  • [ ] Load test pannitteengala?
  • [ ] Budget alert vechirukkeengala?

Cleanup

Practice mudinja piragu delete pannunga:
  1. Auto Scaling Group (idha delete pannaa instances-um delete aagum)
  2. Application Load Balancer
  3. Target groups
  4. Launch template
  5. Use pannaadha security groups
Aduthu naal Billing → Cost Explorer-la check pannunga.

Oru line-la

Auto Scaling = traffic ethukku thagundha maadhiri servers-a thaanaave add, remove pannum. Festival-la fast, sale mudinja piragu cost kammi.

Key takeaways

  • Auto Scaling traffic-ku thagundha madhiri EC2 servers-a thaanaave maathum
  • Min, Desired, Max mukkiyam. Max bill-a kaapaathum
  • Beginners-ku Target Tracking, theriyaadha festival-ku Scheduled Scaling useful
  • App stateless-a irukkanum, data shared place-la irukkanum
  • Eppavum budget set panni, cleanup pannunga

References

  • Amazon EC2 Auto Scaling: https://aws.amazon.com/ec2/autoscaling/
  • EC2 Auto Scaling User Guide: https://docs.aws.amazon.com/autoscaling/ec2/userguide/what-is-amazon-ec2-auto-scaling.html
# amazon-ec2# auto-scaling# scalability# aws-backup# backend
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article