Cloud costsfour expensive defaults

Four AWS architecture defaults that inflate the bill, priced with real list prices for your Region, and how we fix each one.

Cloud cost scenarios

Prices change with the Region: our clients run in Paris, Frankfurt, the UAE and Bahrain.

01

S3 traffic going through a NAT Gateway

Workloads in private subnets read large volumes from Amazon S3: a data lake, model training, log processing. By default that traffic leaves through the NAT Gateway, which bills every gigabyte it processes.

Default configuration

Private subnet → NAT Gateway → S3. Every GB pays the NAT Gateway data processing fee.

$5,000 / month

Optimised architecture

An S3 gateway VPC endpoint and one route in the subnet route tables: S3 traffic goes straight to S3 inside the Region. AWS charges nothing extra for gateway endpoints.

$0 / month

Your figures

Price used: $0.05 / GB (NAT Gateway)

Estimated savings

$5,000 / month

$60,000 / year

How we implement it

  • Create the S3 gateway endpoint (Terraform) and associate it with the private route tables.
  • Switch bucket policies that rely on aws:SourceIp to aws:SourceVpce.
  • Check in Cost Explorer / CUR that NatGateway-Bytes drops.

Trade-offs to plan for

  • Works for S3 and DynamoDB in the same Region only.
  • Interface endpoints (PrivateLink) are a different product: billed per hour and per GB.
  • The NAT Gateway hourly fee remains if other traffic still needs it.

02

Microservices talking across Availability Zones

In a Kubernetes (EKS) cluster spread over several Availability Zones, services call each other without regard to zone. AWS bills traffic between instances in different zones $0.01 per GB in each direction, so $0.02 per GB for a call that crosses.

Default configuration

Pods spread evenly over N zones: about (N−1)/N of internal calls cross a zone.

$6,667 / month

Optimised architecture

Kubernetes Services with trafficDistribution: PreferSameZone (or Istio locality load balancing) keep calls in the caller's zone whenever a healthy endpoint exists there.

$1,333 / month

Your figures

Price used: $0.01 / GB × 2 (out + in)

Estimated savings

$5,333 / month

$64,000 / year

How we implement it

  • Measure cross-zone traffic first (VPC Flow Logs or the CUR line DataTransfer-Regional-Bytes).
  • Enable PreferSameZone on the chattiest Services and spread replicas per zone (topology spread constraints).
  • Keep cluster-wide fallback: if a zone has no healthy endpoint, traffic goes elsewhere.

Trade-offs to plan for

  • Each zone needs enough replicas, otherwise load becomes uneven.
  • Not every call can stay local; the share depends on your traffic.
  • AWS App Mesh reached end of support on 30 September 2026: do not start new designs on it.

03

Stateful batch jobs kept on On-Demand

CI runners or batch jobs of several hours keep state on disk, so teams avoid Spot Instances, which can be interrupted with a two-minute notice, and pay On-Demand prices around the clock.

Default configuration

A fleet of m5.2xlarge On-Demand instances (Linux), running all month.

$32,704 / month

Optimised architecture

The same fleet on Spot, with state on EBS or checkpoints on S3. On the two-minute Spot interruption notice, an EventBridge rule triggers automation that saves the job, moves the volume to a replacement instance and resumes.

$13,033 / month

Your figures

Price used: m5.2xlarge: $0.448/h On-Demand · $0.1486/h Spot

Estimated savings

$19,671 / month

$236,047 / year

How we implement it

  • Checkpoint jobs (or keep their working directory on EBS).
  • EventBridge rule on EC2 Spot Instance Interruption Warning → Lambda / Step Functions to drain, detach and re-attach.
  • Several instance types (m5, m6i, m7i) and an On-Demand fallback in the Auto Scaling group.

Trade-offs to plan for

  • An EBS volume lives in one Availability Zone: the replacement instance must be in the same zone (or use S3 checkpoints to move across zones).
  • Spot prices change by zone and over time, and capacity is not guaranteed.
  • Short jobs without state are simpler: just retry them.

04

A DynamoDB table with a hot partition key

A leaderboard or ad-tech table writes thousands of items per second under one key per day (for example 2026-11-18). A single DynamoDB partition accepts at most 1,000 write units per second: past that, writes are throttled whatever the capacity you pay for. Raising provisioned capacity costs more and does not fix the hot key.

Default configuration

One partition key per day, and capacity over-provisioned to try to absorb the throttling.

$22,542 / month

Optimised architecture

Write sharding: a suffix 0…N−1 added to the key spreads writes over at least peak ÷ 1,000 partitions, and you provision only the capacity you need (or use on-demand mode).

$11,271 / month

Your figures

Price used: $0.000772 / WCU-hour

Estimated savings

$11,271 / month

$135,254 / year

Minimum shards needed: 20

How we implement it

  • Choose N with headroom (often 2× the minimum) and a random or hash-based suffix.
  • Read by querying the N shards in parallel, or maintain a periodic rollup.
  • Watch WriteThrottleEvents and the CloudWatch Contributor Insights hot keys.

Trade-offs to plan for

  • Reads become N queries (or a rollup to maintain).
  • One write unit covers an item up to 1 KB: larger items consume more.
  • Without sharding, more than 1,000 writes per second on one key is not possible.

What is your bill hiding?

In two weeks, we audit your AWS account (network, compute, storage, data) and hand you a prioritised list of savings with the effort each one takes.

AWS public list prices in USD, excluding tax, Free Tier and negotiated discounts (Savings Plans, EDP), checked on 2026-11-18 in the AWS Price List and pricing pages. Spot: lowest price in the Region published by AWS that day; it changes by Availability Zone and over time. Illustrative scenarios, not a client system.

Sources

FAQ / 03

Cloud costs

About these figures

Are these the prices on my bill?

They are AWS public list prices for the Region you select, checked on the date shown. Your bill can be lower with Savings Plans, Reserved Instances or a negotiated discount, and Spot prices move every day.

The audit uses your Cost and Usage Report, so the figures are yours, not estimates.

What does the cloud cost audit deliver?

A review of your AWS account over two weeks: network (NAT, endpoints, cross-zone traffic), compute (sizing, Spot, Savings Plans), storage and databases.

You receive a prioritised list of savings with the effort, the risk and the expected monthly amount for each.

Do you also work on Azure and Google Cloud?

Yes. The same kinds of defaults exist there (NAT, egress, zone traffic, idle capacity); the scenarios on this page use AWS because most of our clients run on it.

Can you guarantee the savings?

No honest engineer can guarantee an amount before seeing the account. What we commit to is a measured baseline, changes you can review in code (Terraform), and a before/after comparison on your bill.

Write to us

Describe your context, even in two lines. We come back with precise questions, then a proposal.

Address

Ennasr

Tunis

Tunisia

Form

Project form

Reply within 24 business hours