Developing Applications On Managed Service For Apache Flink

MSF Pricing Calculator

Overview

Estimate MSF application monthly cost for sizing decisions, budgeting, and optimization analysis. Prices, tables, and worked examples below are us-east-1 and show the billing model; they are not current rates.

Get Current Rates

Before quoting a dollar figure, fetch the rates for the application's region from the Price List API (the pricing endpoint is called in us-east-1 for every region):

aws pricing get-products --region us-east-1 --service-code AmazonKinesisAnalytics \
  --filters Type=TERM_MATCH,Field=regionCode,Value=<region> \
  --query 'PriceList[]' --output json

Use these usage types (prefixed with a region code, e.g. USE1-, EUW1-):

Apply the live rates to the formula below. If the API is unavailable, use the MSF pricing page, and say which source and region the estimate uses.

Pricing Components

There is no separate charge for application code in S3 (your S3 bucket cost only) and no per-API-call charge for Flink REST API or kinesisanalyticsv2 operations.

Formula

KPU_count   = Parallelism / ParallelismPerKPU
Total_KPUs  = KPU_count + 1                                    # orchestration

KPU_cost     = Total_KPUs × $0.11 × hours_per_month              # Total_KPUs for the KPU charge only
storage_cost = KPU_count × 50 × $0.10 × hours_per_month / 730
backup_cost  = total_snapshot_GB × $0.023

monthly_total = KPU_cost + storage_cost + backup_cost

24/7 = 730 hours/month.

To check an estimate against a real bill: in Cost Explorer (service Amazon Kinesis Analytics), RunningApplicationStorage GB-Mo × 730 / 50 equals application KPU-hours, and total KPU-Hour minus that figure equals orchestration KPU-hours.

Sizing → Cost Reference (24/7, KPU + running storage)

Monthly = KPU charge for all KPUs + $5 running storage per application KPU.

These assume stateless-to-moderate stateful workloads. Stateful workloads (keyed windows with large state) typically require PPK=1, doubling KPU count and cost vs. the stateless equivalent.

Worked Examples

Small / development (Parallelism=2, PPK=1, 24/7, negligible state, no snapshots):

• KPU count: 2 / 1 = 2 application KPUs, plus 1 orchestration KPU = 3 billed KPUs

• KPU cost: 3 × $0.11 × 730 = $240.90

• Running application storage: 2 × 50 × $0.10 = $10.00

• Snapshot cost: $0 (none retained)

• Total: ~$251/month

The orchestration KPU is the most common sizing mistake at this scale: it raises the bill from $171 to $251, a 47% increase, and is 32% of the total. Always include it.

Medium production (8 parallel, PPK=1, 24/7, 5 GB state):

• KPUs: 8 + 1 = 9 → 9 × $0.11 × 730 = $722.70

• Storage: 8 × 50 × $0.10 = $40.00

• Snapshots: 5 GB × $0.023 = $0.12

• Total: ~$763/month

Enterprise with autoscaling (avg 16 KPU, peak 32 KPU 20% of time, 60 GB state):

• Peak: 33 × $0.11 × 146 = $530

• Off-peak: 17 × $0.11 × 584 = $1,092

• Storage: (32 × 146 + 16 × 584) × 50 / 730 × $0.10 = $96.00

• Snapshots: 60 GB × $0.023 = $1.38

• Total: ~$1,719/month

Cost Optimization Levers

Ranked by typical impact:

1. Right-size PPK for workload: Stateless apps at PPK=2 cost half of PPK=1 at the same parallelism. Rule of thumb: stateless → PPK=2; stateful → PPK=1.

2. Stop non-prod when idle: Dev/test apps run 200h/month vs 730h is a 73% reduction. Snapshot before stop, restore on start.

3. Enable autoscaling for variable load: 30–50% average reduction for spiky workloads with predictable off-peaks.

4. Match Parallelism to source partitions: Idle subtasks waste KPUs. Parallelism > shard/partition count creates them.

5. Snapshot retention strategy: Without retention, a streaming app's snapshot footprint grows monotonically.

6. Always remember the orchestration KPU: At 1–2 application KPU sizing it's ~32–48% of the bill — easy to under-estimate.

Common Mistakes

References

• MSF Pricing

• resource-optimization.md — KPU sizing inputs for the calculator

• scaling-decisions.md — cost impact of scaling actions