TechPrufer
Get started
Atlassian
Atlassian logoAtlassian · Feature trail

Forge Billing Platform

7 posts · Jan 2025 – Aug 2026

What this trail covers

Atlassian's Forge Billing Platform is evolving to include consumption-based pricing for hosted Forge features, effective January 1, 2026. This includes new pricing for Forge SQL compute duration, compute requests, and data stored, alongside updates to existing pricing for Compute Functions, KVS, and Logs. A generous free tier will be provided for experimentation, with developers only paying for usage above this threshold. New tools like a cost dashboard and cost estimator are being launched to h

All Atlassian trails
  1. More ways to Build on Forge: Container Services is Now in PreviewThis post announces the Preview release of Forge container services, allowing apps to run container images directly on Forge infrastructure as persistent web services. This expands Forge's compute capabilities beyond serverless functions, supporting long-running processes and multi-language backends. It introduces a private container repository, manifest-driven deployment, observability, and a REST API. Security features include blue/green deployments and restricted file system permissions. The service is now a billable capability, with pricing details provided.
  2. Engineering the Forge Billing Platform for Reliability and ScaleThis post details the engineering of Atlassian's Forge Billing Platform, focusing on the reliable measurement, attribution, and billing of usage for the Forge serverless app development platform. It outlines the architecture from Forge services emitting usage events through Streamhub to a shared usage pipeline (UTS) that normalizes and verifies data before handing it off to internal billing systems. Key technical challenges addressed include handling duplicates and late events for counter-style and gauge-style metrics (like invocations and storage), ensuring idempotency and determinism in data processing, and translating billing state into entitlements and enforcement. The system leverages Plato (Databricks/S3) for data lake storage and StarRocks for query-optimized hot tier tables, with Flink jobs for data refinement. The goal is to provide developers with visibility and control over their spend through the Developer Console and APIs.
  3. Getting Ready for Forge BillingThis post announces the official start of Forge billing on January 1, 2026, with the first invoices due in February. It details the setup of developer spaces for billing, introduces financially backed uptime SLAs (99.9% for Forge Functions, KVS, and SQL) with service credits for downtime, and outlines updates to platform quotas and limits, removing several usage-based quotas (Functions invocations/runtime/data, KVS capacity/transfer, resource update quotas). It also highlights faster support response times for Platinum, Gold, and Silver partners and 24x5 help for technical issues via ECOHELP tickets, with billing support now available for paid users.
  4. Updates to Forge Pricing: Effective January 2026This post announces the introduction of consumption-based pricing for hosted Forge features, effective January 1, 2026. It details the new pricing model for Forge Functions, KVS, Logs, and SQL, including free usage allowances and prices above the threshold. It also introduces the concept of developer spaces for invoicing, new tools like a cost dashboard and cost estimator, and outlines the timeline for billing implementation, including staged billing and the first invoice issuance in February 2026. SLAs and support for paid users are also discussed.
  5. Optimizing Forge SQL on a 600K database with TiDB EXPLAINThis post details how to optimize Forge SQL queries for large datasets (600K+ rows) by leveraging TiDB's EXPLAIN and EXPLAIN ANALYZE tools. It demonstrates a step-by-step optimization process, starting with identifying performance bottlenecks caused by full table scans and expensive joins in a pagination query. The author then introduces adding indexes to target foreign keys and sorting columns, which improved execution time by nearly 50% but still hit memory limits. The final optimization involves rewriting the query using Common Table Expressions (CTEs) to retrieve a small subset of rows first, significantly reducing memory usage and enabling scalability beyond Forge's memory and time constraints. The post also provides practical setup instructions and code examples for reproducing the optimizations.
  7. A Preview of Future Forge PricingIntroduces the consumption-based pricing model for Forge, effective January 1, 2026. Details pricing for Compute Functions, KVS (data read, written, stored), and Logs (data written), with specific free thresholds and per-metric prices. Mentions potential for financially backed SLAs and enhanced support for apps exceeding the free tier. Outlines future tooling for usage measurement and forecasting.