#Cost
35 posts
CloudWatch Metrics and Alarm Costs: The Next Leak After Logs
Decomposing the CloudWatch bill beyond Logs: custom metrics accruing $0.30 a month each and how dimension combinations explode the count, the rates for alarms, dashboards, and API calls (GetMetricData in particular), how third-party observability tools generate metric API charges, cutting counts with the embedded metric format and dimension design, and the inspection order.
Choosing Where to Analyze Logs on AWS: Logs Insights vs Athena vs OpenSearch
A working standard for where to search and analyze logs: Logs Insights querying what already sits in CloudWatch at $0.005 per GB scanned, Athena asking SQL of logs archived in S3 at $5 per TB with partitioning and compression deciding the bill, OpenSearch and its fixed cost for always-on dashboards and full-text search, and the practical setup that layers all three by query frequency.
Kinesis vs MSK vs SQS: Choosing Between Streams and Queues
A working standard for the three services that receive event data: the model split between queues that delete on consumption and streams that replay, SQS's simplicity at $0.40 per million, Kinesis's provisioned shards and on-demand mode, the Kafka-ecosystem requirements that justify MSK, its two branches of Express brokers and Serverless, and a decision order driven by scale and team capability.
Spot Instances in Practice: What the 90% Discount Requires, and Designing for Interruption
What Spot demands in exchange for up to 90% off with no commitment: how the 2-minute interruption notice and rebalance recommendation actually behave, the line between interruption-tolerant workloads and everything else, how the price-capacity-optimized allocation strategy and instance diversification reduce interruption frequency, where Fargate Spot fits, and completing the three cost axes alongside RIs, Savings Plans, and Graviton.
Route 53 Routing Policies Compared: Choosing Among the 8 Policies
A scenario-driven standard for Route 53's eight routing policies: what simple, weighted, latency, failover, geolocation, geoproximity, multivalue, and IP-based are each for, the query rates that vary by policy ($0.40–$0.80 per million) and the effect of free alias queries, health check pricing and pairing rules, and the answers for canary deploys, DR, and multi-region.
Step Functions vs SQS Chains: Choosing a Workflow Orchestration Model
Whether multi-step backend processing should be conducted by Step Functions or passed along SQS-and-Lambda chains: the structural split between orchestration and choreography, the price gap between Standard at $25 per million state transitions and Express at $1 per million executions, what an SQS chain gains in simplicity and loses in visibility, and the deciding factors of retries, compensation, and long runs, plus the patterns that combine both.
Graviton vs x86: Choosing an EC2 Instance Architecture
A working standard for ARM (Graviton) versus x86 on EC2: the hourly rate gap between m8g and m7i and the larger price-performance gap behind it, where Graviton5 (m9g, GA in 2026) fits, why managed services migrate first, the difficulty gap between interpreted languages and native dependencies, the workloads where x86 remains, and the decision order.
Cutting CloudWatch Logs Costs: Ingestion Is the Bill, Not Storage
Decomposing the CloudWatch Logs bill: why ingestion at $0.50/GB is 17 times the $0.03 storage rate, what the half-price Infrequent Access log class gives up, volume tiers and the intelligent tiering introduced in 2026, routing Lambda logs to S3 and Firehose instead, retention settings, and the order in which to reduce what you ingest.
Aurora Serverless v2 vs Provisioned: Pricing Structure and When to Use Which
A working standard for Aurora's two capacity models: the ACU billing unit at about $0.12 per hour, what scale-to-zero auto-pause buys and the resume delay it costs, the break-even where provisioned plus commitments wins for steady load, the trap of a high ACU floor turning serverless into fixed cost, and the hybrid of a provisioned writer with serverless readers.
Choosing an ElastiCache Engine: Valkey vs Redis OSS vs Memcached
A working standard for the three ElastiCache engines: how Valkey, forked after the Redis license change, became the default for new clusters, the 20% node-based and 33% serverless price gap, the serverless minimum that splits at 100MB for Valkey versus 1GB for Redis OSS, the narrow niche where Memcached still fits, and the node-based versus serverless decision.
API Gateway vs ALB: Choosing What Sits in Front of Lambda
A working standard for what to put in front of Lambda: the pricing-model split between per-request billing (HTTP API at $1.00 per million) and hourly-plus-LCU billing (ALB at about $16/month fixed), the crossover near 16 million requests a month, the feature set that justifies REST API at $3.50, where the free Lambda Function URL fits, and a decision order that includes VPC-internal targets and WAF.
Fixing Lambda Cold Starts: The Cost Math of Provisioned Concurrency
Lambda cold starts and their remedies from a cost angle: how the init phase creates the latency, the near-free fixes like smaller packages and SnapStart, the two-part pricing of Provisioned Concurrency ($0.0000041667/GB-s to keep warm plus a discounted execution rate), the crossover around 60% utilization where it becomes cheaper than on-demand, scheduling, and the decision order.