sites.fmind.dev

Portfolio 07 · telemetry economics

AI observability cost and cardinality calculator

Convert observed event volume, retained sink copies, label combinations, instances, scrape rate, and histogram accounting into separate log and metric ledgers.

Decision this answers: Which label or duplicated sink drives the next bill step, and does the account-scoped free allowance actually absorb it?

Scenario

Observed average during active time.

measured[1]

Measured serialized size, not source object size.

measured[1]

Every documented duplicate ingestion or storage copy.

measured[2]

Families affected by the selected labels.

measured[3]

Observed combinations, not the theoretical Cartesian upper bound.

measured[1]

Editable effective rate after the free allowance.

published list[1]

The complete default scenario is server rendered. Interactions stay in this browser.

Decision summary

Monthly log volume

1.91 TiBexact-derived

Active seconds times retained sink copies.

[1]

Observed active series

230400exact-derived

Families times observed combinations times instances.

[3]

Monthly samples

39813120000exact-derived

Series multiplied by the selected scrape schedule.

[2]

Log ingestion bill

$952.67exact-derived

Free allowance applied once at account scope.

[1]
Direct-labeled result profile

The tables below are the accessible source of truth; bar lengths never carry identity alone.

Provider counting boundaries

CandidateLog meterMetric meterHistogram treatmentFree allowance scopeFuture price policyEvidence
Google Cloud ObservabilityGiB ingested/retainedSamples or bytes by productProvider ruleBilling accountDisplay, do not apply earlypublished list[1]
Azure MonitorGB ingestion/retentionTime-series dependentProvider ruleCommitment/account scopeDisplay, do not apply earlypublished list[2]
Amazon CloudWatchGB ingestion/archiveCustom metrics/APIProvider ruleAccount and regionDisplay, do not apply earlypublished list[3]

Official surface versus this site

This offline workbench

Official pages define provider meters. This page starts from one AI workload shape and exposes the cardinality multiplier, duplicated sinks, and account-scoped allowances before applying a meter.

Sources and methodology

Every result-affecting reference is visible here without JavaScript and is retained in the JSON export.

[1] Google Cloud Observability pricing
Authority
Google Cloud
Native identifier
Cloud Logging and Monitoring meters
Region
global
Unit
GiB and metric samples
Evidence
published list
Price kind
list
As of
2026-08-01
Effective from
2026-08-01
Retrieved
2026-08-01
Freshness
fresh
Normalization
The opening log rate and allowance are a dated scenario snapshot; native scopes remain visible.

Open the authoritative source

[2] Azure Monitor Logs cost model
Authority
Microsoft
Native identifier
Azure Monitor Logs
Unit
GB ingestion and retention
Evidence
published list
As of
2026-08-01
Retrieved
2026-08-01
Freshness
reference
Normalization
Counting and duplicate destination charges are represented separately.

Open the authoritative source

[3] Amazon CloudWatch pricing
Authority
Amazon Web Services
Native identifier
CloudWatch metrics and logs
Unit
metric, sample, GB
Evidence
published list
As of
2026-08-01
Retrieved
2026-08-01
Freshness
reference
Normalization
Provider-specific distribution and histogram rules are not flattened into one universal meter.

Open the authoritative source

Limits and non-claims

  • The default histogram contributes 14 bucket/sum/count observations per sample event.
  • Theoretical Cartesian cardinality is an upper bound, never substituted for observed combinations.
  • Provider-specific account aggregation and effective dates must be refreshed before a production budget.