Heroku Free-Tier Shutdown — SaaS Plan Impact Assessment Historical policy: November 28, 2022 · Assessment date: October 21, 2026 · Evidence: Heroku official documentation 1. Change Impact Summary Heroku discontinued free dynos, free PostgreSQL, and free Redis plans on November 28, 2022. The change affected application availability, persistent data, scheduled jobs, and infrastructure billing. Heroku Dev Center +1 The most consequential outcomes were: - CONFIRMED — Free dynos converted to Eco and were scaled to zero, requiring paid activation to resume operation. - CONFIRMED — Free PostgreSQL and Redis databases in personal and non-Enterprise Team accounts were queued for deletion. - CONFIRMED — Eligible Enterprise-owned free databases were automatically converted to Mini, with an application restart. - CONFIRMED — Scheduled jobs using free dynos stopped working until reconfigured. - INFERRED — Production downtime could interrupt customer transactions, authentication, scheduled processing, and other application-dependent services. - UNKNOWN — Actual customer impact, recoverability, contractual obligations, and completed migrations. No application inventory or execution records were provided. Evidence: Heroku retirement announcement and Enterprise guidance. Heroku Help +1 Critical historical distinction The specified migration is no longer executable as originally designed. Heroku Postgres Mini and Basic reached end-of-life on May 29, 2024. New Postgres deployments must use supported alternatives, such as Essential-tier plans. Heroku Dev Center +1 The primary decision is whether the assessment concerns historical compliance with the 2022 shutdown or recovery and modernization of applications that exist today. 2. Plan & Entitlement Diff Plan / Attribute Current (before shutdown) Proposed (2022) Classification Affected Users Evidence / Gap Dyno availability Free compute Paid Eco; inactive until enabled and scaled LOST / MODIFIED All free-dyno owners CONFIRMED Heroku Dev Center Compute allowance Free-tier limits Eco: 1,000 shared hours/month, $5/month MODIFIED Eco subscribers CONFIRMED Heroku App uptime Free dynos sleep Eco also sleeps after inactivity MODIFIED Production owners CONFIRMED; workload compatibility UNKNOWN PostgreSQL Free hobby-dev Mini: 10,000 rows, $5/month in 2022 MODIFIED Database owners CONFIRMED Heroku Help Redis Free hobby-dev Mini: 25 MB, $3/month MODIFIED Redis users CONFIRMED Heroku Help Data retention Existing free data Deletion for affected non-Enterprise resources LOST Personal / Team accounts CONFIRMED Heroku Dev Center Enterprise database Free hobby-dev Automatic Mini conversion MODIFIED Enterprise-owned resources CONFIRMED Heroku Help Scheduled jobs Free dyno execution Requires paid dyno configuration LOST until remediated Scheduler users CONFIRMED Heroku Help Customer contracts NOT PROVIDED No automatic contractual change established UNKNOWN SaaS subscribers Contracts and terms missing End-user entitlements NOT PROVIDED Must remain contractually valid UNKNOWN Application customers App-level authorization data missing A Heroku infrastructure plan is not equivalent to a SaaS customer's subscription plan. Changing Heroku resources does not, by itself, authorize changing end-user prices, subscription records, or contracted product access. 3. Customer Cohort Impact Matrix Customer Cohort Applicable Terms Price Treatment Access Treatment Transition Event Risk / Unknown Personal free-dyno owners Free tier retired Paid Eco or alternative Dynos scaled to zero Nov 28, 2022 Downtime Personal free-database owners No free retention protection Paid data plan required Deletion if not migrated From Nov 28, 2022 Permanent data loss Non-Enterprise Team owners Same free-data retirement Paid plan required Free databases queued for deletion From Nov 28, 2022 Data loss Enterprise-owned databases Enterprise policy applies No immediate contract invoice change Automatic Mini conversion in 2022 Nov 28, 2022 Restart and compatibility Enterprise members' personal apps Personal-account policy applies Separately billable No automatic Enterprise protection Nov 28, 2022 Incorrect ownership assumptions Scheduled-job users Free dynos discontinued Paid execution required Jobs fail until reconfigured Nov 28, 2022 Missed processing Production application owners Hosting obligations unknown Depends on selected resources Availability must be revalidated Resource migration Outage / SLA exposure SaaS end customers Contracts NOT PROVIDED No authorized price change established Existing entitlements require review Owner-defined Contractual exposure Confirmed account distinctions: Heroku Help +1 Grandfathering determination: - CONFIRMED — Heroku did not preserve the discontinued free infrastructure plans for ordinary personal and Team accounts. - CONFIRMED — Enterprise-owned free data services received a specific conversion treatment, not indefinite free-plan retention. - UNKNOWN — Any independently negotiated protections for the SaaS application's own paying customers. - NOT PROVIDED — Subscription records, contractual commitments, prepaid periods, and customer-specific exemptions. 4. Billing & Migration Blueprint Required resource mappings Original resource Historical 2022 migration Correct 2026 assessment free dyno Eco with paid activation Inspect current dyno state; select Eco or a production-suitable alternative Postgres hobby-dev Mini before deletion Use supported Essential-tier destination where recovery or migration is possible Redis hobby-dev Mini before deletion Verify current Key-Value Store plan, capacity, and data availability Scheduler on free dynos Reconfigure for paid dynos Verify every scheduled execution and missed-job recovery Enterprise-owned hobby-dev Automatic Mini conversion Inspect actual current resource and billing state The historical Eco and Mini billing start date was December 1, 2022, under Heroku's transition arrangements. That provision does not establish any billing treatment for migrations performed in 2026. Heroku Help Ordered execution plan 1. Inventory — Platform owner. Export account ownership, app IDs, dyno states, database identifiers, scheduled jobs, subscriptions, and invoices. Establish the actual post-2022 state before changing anything. 2. Recovery assessment — Database administrator. Identify surviving databases, verified external backups, missing datasets, and restoration feasibility. Protect remaining data before provisioning or changing resources. 3. Policy and target selection — Product and finance owners. Approve current plans, spending limits, customer obligations, and intended downtime windows. 4. Migration rehearsal — Engineering. Test supported resource transitions on representative isolated environments. Verify database integrity, connectivity, background jobs, and billing identifiers. 5. Controlled cutover — Platform operations. Transition recoverable workloads, validate dependency readiness, restore process capacity, and verify end-to-end application behavior. 6. Reconciliation — Finance and operations. Compare actual resource states, invoices, scheduled executions, database records, and customer access against expected outcomes. Critical failure scenarios Scenario Preventive Control Recovery Database already deleted Verify retention and backups before migration Restore from independently verified backup if available Dyno active but database missing Dependency readiness checks Disable affected writes; restore verified data Scheduler executions missed Durable job ledger and gap analysis Idempotent replay of recoverable jobs Duplicate provisioning or billing Stable resource IDs and preflight reconciliation Reconcile charges; resolve provider-side exceptions Enterprise ownership mismatch Verify actual app/add-on owner Correct account-specific migration plan Eco sleeping or quota exhaustion Validate workload and capacity requirements Move critical workloads to suitable always-on resources Partial application recovery Dependency-level health checks Revert reversible configuration and resume only healthy services Heroku explicitly states that data already removed from its platform cannot be restored through Heroku. External backups must therefore be treated as a separate recovery source, not an assumed safety net. Heroku Help Essential verification tests All tests below are PROPOSED. None are EXECUTED or VERIFIED. - Verify database restoration using record counts, referential integrity, and representative application queries. - Test application startup after database and resource transitions. - Simulate duplicate, interrupted, and delayed provisioning/billing events. - Verify Scheduler execution, recovery, and prevention of duplicate job effects. - Compare actual invoices with approved resources and billing periods. - Verify existing customer access and authorization after recovery. The SaaS subscription system, if any, must be reconciled independently of Heroku infrastructure charges. 5. Release Decision Gate Status Supporting Evidence Required Resolution Policy CONDITIONAL Heroku retirement rules confirmed Obtain owner authorization and customer contracts Entitlement UNKNOWN Infrastructure effects known Establish actual application access and dependencies Billing UNKNOWN Historical list prices and billing dates confirmed Verify current resource pricing and invoices Customer UNKNOWN General cohorts identified Establish affected population and obligations Recovery FAIL Historical deletion risk; obsolete Postgres Mini destination Establish surviving data, supported targets, and tested recovery FINAL VERDICT BLOCKED The original 2022 migration plan must not be executed unchanged in 2026. PRIMARY REASON: The shutdown has already occurred, PostgreSQL Mini is obsolete, and actual resource and recovery states have not been established. CRITICAL BLOCKERS: Unknown database survival, missing resource inventory, unsupported historical target plan, missing financial authorization, and no recovery verification. REQUIRED APPROVALS: Infrastructure owner, database owner, finance owner, and contractual/customer-policy owner where applicable. WHAT WOULD CHANGE THE VERDICT: A verified inventory, supported target architecture, demonstrated recovery, approved billing treatment, and successful integration and reconciliation tests. 6. Next 3 Actions Priority Action Owner Expected Evidence 1 Audit every historical free Heroku app, database, and job to establish its current state Platform owner Resource inventory and ownership/status report 2 Establish data recoverability and select supported 2026 resource plans Database + infrastructure leads Backup verification, target mappings, recovery rehearsal 3 Approve financial and operational transitions; test and reconcile end-to-end Finance + engineering leads Approved migration plan, test results, access validation, billing reconciliation Final assessment The highest-consequence failure in this case is irreversible data loss, followed by prolonged application downtime and missed scheduled processing. An Enterprise affiliation alone does not establish protection for personal-account resources. No customer counts, real financial exposure, migration execution evidence, or recovery results were supplied. Consequently, the documented findings establish the vendor-policy impact, not proof that any particular application was successfully or unsuccessfully migrated. The release remains blocked until the actual resource state and recoverability are established.