Idle AWS Resources: How NAT Gateways, Load Balancers and EC2 Instances Drain Budgets

·2 views

Cloud teams frequently overpay for AWS resources that are running but performing no useful work, yet 'idle' means something different for each resource type. An EC2 instance averaging below 5% CPU and negligible network traffic over 14 days can cost anywhere from $70 to $500 per month depending on instance size, while an unused NAT gateway costs $32.85 monthly before any data transfer charges. Idle Application and Network Load Balancers add over $16 each per month, and an EKS cluster with zero nodes still bills $73 a month for the control plane alone. Unconnected RDS databases are among the costliest idle resources, with Multi-AZ configurations doubling charges that can exceed $125 a month for mid-tier instances. Establishing specific, defensible thresholds — such as zero connections over 14 days for databases or zero requests over 7 days for load balancers — turns vague cost complaints into actionable cleanup targets.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Framer Templates Help Businesses Cut Website Launch Time Without Sacrificing Quality

Framer Templates offer businesses a pre-built website foundation, reducing the time and effort typically required to design and develop a site from scratch. Instead of building layouts, navigation, and visual elements from the ground up, teams can start with structured templates and focus on customization and content. This approach is particularly useful for startups validating ideas, marketing teams running campaigns, and companies needing landing pages on short notice. Templates can be tailored to reflect a brand's identity, covering elements like typography, color schemes, and page sections. Businesses are advised to evaluate factors such as responsiveness, performance, and flexibility for future updates before selecting a template.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

AWS Cost Tools Compared: What Trusted Advisor, Compute Optimizer and Cost Explorer Miss

AWS offers three main cost-management tools — Trusted Advisor, Compute Optimizer, and Cost Explorer — each designed to answer a different financial question rather than provide complete waste coverage. Trusted Advisor runs point-in-time hygiene checks on idle or underused resources, but its full cost features require a paid Business or Enterprise support plan. Compute Optimizer uses machine learning on utilization history to recommend rightsizing for compute resources like EC2, Lambda, and EBS, but ignores non-compute waste entirely. Cost Explorer excels at identifying commitment opportunities such as Reserved Instances and Savings Plans, while offering only a shallow rightsizing view. Significant waste categories — including non-production workloads running around the clock, orphaned snapshots, NAT gateway charges, cold data in hot storage, and Kubernetes over-provisioning — remain invisible to all three tools.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How to Rightsize AWS Cloud Resources Without Triggering Outages or Team Pushback

Cloud infrastructure often runs at just 15% average utilization because teams overprovision at setup and rarely scale back after incidents, making rightsizing a social challenge as much as a technical one. Migrating EBS volumes from gp2 to gp3 is recommended as a first step, offering roughly 20% cost savings with no downtime, though volumes over 1 TB require IOPS usage checks before migration. For EC2 instances, engineers should use 90 days of CPU history sized to p95/p99 peaks rather than averages, downsize one step at a time, and monitor user-facing latency rather than CPU alone. GPU instances such as the g5, p4, and p5 families carry costs ranging from about $735 to over $23,900 per month, making idle or oversized GPU nodes far more expensive than equivalent CPU waste. A disciplined rollout — changing one dimension per window, canarying one instance first, and keeping rollback a single API call away — is essential to sustaining a rightsizing program long-term.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Anchor Phrases Beat Verbatim Quotes for Reliable AI Fact Extraction

A developer tested two methods for extracting facts with source provenance from AI models: asking for verbatim quotes versus retrieving short anchor phrases of three to eight words. Verbatim quote accuracy varied widely across model tiers, ranging from just 37% to 93%, with some models silently fabricating plausible-sounding but incorrect quotes. The anchor phrase approach, which locates a real substring in the source document, achieved coverage of 93% to 100% across all tested model tiers. The anchor method also fails explicitly when no match is found, rather than returning a hallucinated citation. The findings highlight that extraction accuracy and extraction reliability are distinct problems, and that a model which quietly quotes incorrectly is more dangerous than one that admits uncertainty.

0 comments Read more at DEV Community