specification: FinOps Framework specificationVersion: '1.0' schema: https://www.finops.org/framework/ alignedWith: framework: FinOps Foundation Framework frameworkUrl: https://www.finops.org/framework/ dataSpec: FOCUS dataSpecVersion: '1.3' dataSpecUrl: https://focus.finops.org/focus-specification/v1-3/ provider: Cribl providerId: cribl publisherName: Cribl, Inc. serviceCategory: Observability Pipeline created: '2026-05-04' # Provenance stamped 2026-08-11: this artifact was written by the API Evangelist # bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35. method: generated modified: '2026-05-05' reconciled: true tags: - Configuration - Data Lake - Data Pipelines - Data Routing - Edge Computing - Observability - Search - Security Data - Stream Processing - Telemetry - FinOps - Cost Management - FOCUS description: FOCUS-aligned FinOps for Cribl - consumption-based licensing on processed-data volume (TB/day) plus tier-bound capacity (worker groups, worker processes, edge nodes, Lake capacity, search executors). Free tier covers 1 TB/day; Standard and Enterprise are negotiated commercially. sources: - https://cribl.io/pricing/ - https://docs.cribl.io/api-reference/ - https://focus.finops.org/focus-specification/v1-3/ billingModel: pricingCategory: Consumption-Based + Subscription billingFrequency: Annual billingCurrency: USD chargeCategories: - Usage - Purchase - Tax - Adjustment - Refund - Credit chargeFrequency: Recurring focusColumns: ServiceName: Cribl ServiceCategory: Observability Pipeline ProviderName: Cribl PublisherName: Cribl, Inc. InvoiceIssuerName: Cribl, Inc. PricingCategory: Usage-Based BillingCurrency: USD ChargeCategory: Usage meters: - name: data_processed description: Volume of telemetry data processed through Cribl Stream / Edge per day unit: TB aggregation: sum dimensions: - source - destination - pipeline - worker_group - environment - name: worker_processes description: Worker process count concurrently running across worker groups unit: worker_process aggregation: max dimensions: - worker_group - environment - name: edge_nodes description: Cribl Edge agents deployed across the fleet unit: edge_node aggregation: max dimensions: - fleet - environment - name: lake_storage description: Cribl Lake storage consumed unit: GB-month aggregation: avg dimensions: - dataset - retention_class - name: search_executors description: Concurrent Cribl Search executors unit: executor aggregation: max dimensions: - workspace - environment - name: subscription_fee description: Annual / monthly subscription baseline tied to the platform tier unit: month aggregation: max dimensions: - tier - workspace principles: - name: Visibility description: Use the Cribl Cloud dashboard, Cribl Search, and per-pipeline metrics to track TB/day processed by source, destination, pipeline, and worker group. Export usage telemetry into the same observability stack you bill for. - name: Allocation description: Allocate cost by source system (security, infra, app), destination (SIEM, lake, archive), and team. Cribl's pipeline-level metrics make per-team chargeback feasible for shared deployments. - name: Optimization description: Cribl's core ROI is volume reduction - use Reduce, Suppress, Sample, and Drop functions in pipelines to shrink data before downstream-tool ingest cost; route low-value telemetry to Cribl Lake instead of premium SIEM ingest; right-size worker process count to match TB/day; use Cribl As Code (Terraform / SDK) to enforce the optimization patterns across environments. - name: Accountability description: Assign each pipeline an owner; alert on TB/day burn-rate at the worker-group level so over-budget telemetry sources are caught before a renewal cycle. Tie Cribl ROI metrics (pre- vs post-Cribl downstream cost) into FinOps reviews. maintainers: - FN: Kin Lane email: kin@apievangelist.com