Hosting migration

Fly.io → Railway

Move builds, previews, functions, routing, domains, environment configuration, stateful services, observability, and deployment controls from Fly.io into Railway with an explicit exception ledger and evidence-based cutover.

Typical timeline8–20 business days38–80 hours active work

Statusneeds review

Source testedFly.io documentation reviewed 2026-07-19

Destination testedRailway documentation reviewed 2026-07-19

Last reviewed2026-07-19

Sources2

Before you migrate

Should you make this move?

Both platforms have a case. Compare what you gain with what you give up before scheduling the cutover.

Current platform

Fly.io

Reasons to stay

Applications and databases can run close to users with container-level control
Managed deployments reduce infrastructure and release-management work

Reasons to leave

Networking, capacity, and operational debugging require more infrastructure skill
Runtime assumptions, platform services, and pricing models shape the architecture

New platform

Railway

What gets better

Fast service and database provisioning gives developers a low-friction application platform
Managed deployments reduce infrastructure and release-management work

What gets worse

Usage pricing and infrastructure abstraction can become limiting at scale
Runtime assumptions, platform services, and pricing models shape the architecture

Best of the move

Railway: Fast service and database provisioning gives developers a low-friction application platform. This removes a major source-side concern: Networking, capacity, and operational debugging require more infrastructure skill.

Worst of the move

What you lose: Applications and databases can run close to users with container-level control. What you inherit: Usage pricing and infrastructure abstraction can become limiting at scale.

01At a glance

Know the shape of the move.

Transfer outcome12 features audited

Clean1

Partial6

Manual4

Lost1

Mapping route9 of 9 fields have a destination path

SourceFly.io9 mapped fields

Automatic0

Transform4

Manual5

Unsupported0

DestinationRailway9 supported paths

This timeline assumes

Up to 20 projects, 100 functions, and 500 routes
The application source code and infrastructure dependencies are available for review.
A production-shaped destination preview runs before traffic changes.
The source stays intact and billable through the rollback window.
The migration team records evidence for every blocking verification check.

02Loss matrix

What survives the move.

“Partial” and “manual” are not footnotes. They are work that must be scheduled and verified.

Item	Outcome	Impact	What happens	Mitigation
Builds, releases, runtime behavior, configuration, routes, domains, and state	partial	critical	Core builds, releases, runtime behavior, configuration, routes, domains, and state can move, but Fly.io and Railway use different models, limits, identifiers, and import behavior.	Pilot every feature class, preserve source IDs, and reconcile accepted, transformed, rejected, and excluded items.
Provider-native runtime, managed services, history, and operational controls	manual	critical	Fly.io-specific runtime configuration and historical evidence do not become native Railway behavior through the core data transfer.	Inventory every active dependency, approve its destination disposition, and test the replacement before source writes stop.
Source code and static assets	clean	critical	Repository-owned application code can redeploy when the destination supports the framework and runtime.	Pin versions and compare reproducible build output.
Build configuration and dependency cache	partial	critical	Build images, commands, paths, caching, and framework detection differ.	Run clean production and preview builds with no source cache.
Environment variables and secrets	manual	critical	Secret values and environment scoping do not transfer with code.	Recreate least-privilege values and rotate them after cutover.
Functions, workers, and background jobs	partial	critical	Runtime APIs, regions, limits, concurrency, retries, and scheduling differ.	Port and integration-test every endpoint and job.
Databases and durable state	partial	critical	State requires a separate dump, restore, replication, or dual-write plan.	Rehearse consistency, downtime, restore time, checksums, and backups.
Object, key-value, and persistent storage	partial	critical	Managed stores and attached disks do not move with the application deployment.	Copy and verify every object, key, volume, and access policy.
Domains, DNS, and TLS	manual	critical	Custom domains and certificate ownership require destination validation and DNS changes.	Lower TTL, prevalidate domains, and retain exact rollback records.
Redirects, rewrites, headers, and routing	partial	critical	Rule syntax, ordering, matching, and proxy behavior differ.	Create a routing registry and test every rule and status code.
Preview, staging, and promotion workflow	manual	high	Branch deployments, approvals, aliases, and promotion models differ.	Rebuild and test the release path before production.
Observability, logs, alerts, and analytics	lost	critical	Historical telemetry remains in the source and alert integrations need new endpoints.	Export baselines and prove logs, metrics, traces, alerts, and retention.

03Field and feature mapping

Where each thing goes.

Source	Destination	Method	Notes
Fly.io project, app, or service	Railway project or service	manual	Map topology, runtime, region, and ownership.
Build command and output	Destination build configuration	transform	Pin runtime and dependency versions.
Environment variable or secret	Destination environment variable or secret	manual	Recreate by environment and rotate.
Function, worker, or job	Destination compute unit	transform	Port APIs, limits, retries, and schedules.
Database or durable store	Destination data service	transform	Migrate and verify separately from code.
Domain and certificate	Destination domain and certificate	manual	Prevalidate before DNS changes.
Redirect, rewrite, or header	Destination routing rule	transform	Test precedence, conditions, and status.
Preview or staging environment	Destination preview or staging environment	manual	Rebuild branch and access behavior.
Log, metric, alert, or analytics stream	Destination observability	manual	Reset history after exporting baselines.

04Before you begin

Make the move recoverable.

Backup procedure

Create the source-of-truth backup

Preserve Fly.io data, configuration, and operating evidence before any destination write.

Export Fly.io build settings, environments, domains, routes, functions, stateful services, access, integrations, and recent releases.
Create application, data, storage, DNS, certificate, and observability inventories.
Record traffic, error, latency, build, job, and capacity baselines.
Preserve the last known-good release and hash all exports.

Proof to capture: A signed manifest reconciles every scoped record class, runtime dependency, export file, count, and hash.

Transformation · Infrastructure manifest and route tester

Runtime and routing manifest

Translate application topology, builds, compute, domains, and rules.

Inventory source values and exceptions.
Define explicit destination mappings.
Reject unmapped critical records.

Proof to capture: Save the input, output, command or tool settings, warnings, and final item counts.

Transformation · Provider tools, checksums, and restore scripts

State and configuration migration

Move data, objects, secrets, and observability deterministically.

Normalize encoding, dates, identifiers, and blanks.
Run a representative pilot.
Reconcile accepted, rejected, and transformed rows.

Proof to capture: Save the input, output, command or tool settings, warnings, and final item counts.

05Handle with care

The things most likely to hurt.

These are operating limits. Treat every “Stop if” condition as a blocked migration, not a suggestion.

Preview

The app builds but runtime behavior differs

criticalpossible likelihood

Routes render while functions, jobs, state, caching, or retries behave differently.

Consequence: Users lose data or critical actions.
Mitigation: Test complete business paths under production-shaped load.

Stop if: Any critical path lacks durable evidence.

Cutover

Both platforms write durable state

criticalpossible likelihood

Source and destination workers, schedules, or web traffic modify the same data.

Consequence: State diverges or actions duplicate.
Mitigation: Freeze source writers and transfer one owner at a time.

Stop if: An unapproved source writer remains active.

Domain

DNS rollback is unavailable

criticalpossible likelihood

Certificates, proxy settings, or DNS records cannot return traffic quickly.

Consequence: An outage outlives the application fix.
Mitigation: Prevalidate both directions and preserve exact prior records.

Stop if: Rollback cannot be completed within the approved objective.

06Precise timeline

Do the work in this order.

Estimate forUp to 20 projects, 100 functions, and 500 routes

Total elapsed8–20 business days

Active work38–80 hours

BufferAdd time for large exports, unsupported Fly.io features, destination limits, identity exceptions, regulated data, or strict downtime requirements.

Days 1–3Inventory and decisions2–3 days

Days 3–5Backup and reconcile1–3 days

Days 5–12Map and pilot3–8 days

Days 10–20Final delta and switch1–5 days

Days 12–40Observe and close3–10 days

1
Days 1–3 · inventory
Inventory and decisions
6–12 hours active2–3 days elapsedOwner review waiting
- Inventory Fly.io data, features, users, domains, and integrations.
- Approve scope, owners, mappings, and exclusions.
Depends on: Fly.io and Railway administrator access
Stop / go checkpoint
Export?
Go when: Every critical item has an owner and disposition.
Stop when: Consent, billing, access, or system ownership is unclear.
2
Days 3–5 · backup
Backup and reconcile
5–10 hours active1–3 days elapsedExport processing waiting
- Create immutable exports and configuration evidence.
- Reconcile counts, totals, and hashes.
Depends on: Approved inventory
Stop / go checkpoint
Transform?
Go when: Source totals and export manifests agree.
Stop when: Any critical dataset or configuration is absent.
3
Days 5–12 · pilot
Map and pilot
10–35 hours active3–8 days elapsedDestination processing and review waiting
- Configure Railway and transform representative data.
- Run a pilot containing normal records and every critical edge case.
Depends on: Verified backup
Stop / go checkpoint
Scale?
Go when: Pilot mappings, behavior, access, and evidence pass.
Stop when: Any critical check fails or produces unexplained variance.
4
Days 10–20 · cutover
Final delta and switch
5–25 hours active1–5 days elapsedDNS, import, or sync propagation waiting
- Freeze production writes in Fly.io.
- Apply the final delta, switch ownership, and run blocking checks.
Depends on: Passed pilot and approved rollback
Stop / go checkpoint
Open production?
Go when: Counts reconcile and one destination system owns production.
Stop when: A source writer remains active or a blocking check fails.
5
Days 12–40 · observe
Observe and close
3–18 hours active3–10 days elapsedOperating-cycle evidence waiting
- Monitor one complete operating cycle.
- Sign the verification report and close rollback only after stability.
Depends on: Verified cutover
Stop / go checkpoint
Close rollback?
Go when: No trigger occurs during the agreed observation period.
Stop when: Data, access, delivery, routing, or business results regress.

07The point of change

Cut over with a way back.

Go live

Cutover

Make Railway the only production system without losing the final Fly.io delta.

Recommended window: A low-volume weekday morning with platform, data, DNS, and business owners available.

Freeze production writes and scheduled actions in Fly.io.
Export, transform, and reconcile the final delta.
Apply the approved delta to Railway.
Switch domains, forms, integrations, sending, or sync ownership as applicable.
Run every blocking verification check and keep the source intact.

Proof to capture: Railway owns production, totals reconcile, and every blocking check has durable evidence.

Return to safety

Rollback

Return production ownership to Fly.io without losing destination-era changes.

Deadline: Within seven days and before source data, plans, credentials, domains, or billing are changed.

Stop new writes and scheduled actions in Railway.
Restore the prior Fly.io routing, forms, integrations, sending, or sync ownership.
Export the Railway post-cutover delta.
Review and apply safe destination-era changes to Fly.io.
Run the same blocking checks against the restored source.

Proof to capture: Fly.io again owns production with current data and no duplicate destination action.

Rollback immediately when

Unexplained critical count or value variance
Missing or exposed critical data
Duplicate production action
Failed access, routing, delivery, or integration check
A critical feature has no safe destination replacement

08Verification report

Prove the migration worked.

Every blocking check must pass. Capture the evidence before cleanup begins.

ID	Check	Method	Expected result	Evidence
`V-01`Blocking	Reproducible release	Build a pinned source revision from a clean environment.	Output and dependency versions match acceptance.	Build log and artifact hash.
`V-02`Blocking	HTTP behavior	Crawl routes, redirects, rewrites, headers, errors, and assets.	Approved status, body, and header parity.	Route diff.
`V-03`Blocking	Functions and jobs	Run every endpoint, queue path, schedule, retry, and timeout.	One correct durable result per test.	Runtime log.
`V-04`Blocking	Data and storage reconciliation	Compare counts, checksums, invariants, and backup restore.	Every scoped item is accounted for and recoverable.	State report.
`V-05`Blocking	Secrets and access	Test representative users, services, previews, and production actions.	Least privilege with no secret exposure.	Access evidence.
`V-06`Blocking	DNS, TLS, and rollback	Resolve from multiple networks and exercise rollback records.	Valid TLS and routing within approved objectives.	DNS and certificate report.
`V-07`Blocking	Logs, metrics, traces, and alerts	Generate known failures and load.	Signals arrive with correct context and thresholds.	Observability evidence.
`V-08`Blocking	Single production platform	Inspect traffic, writers, jobs, webhooks, and deployment triggers.	Railway alone owns production.	Ownership checklist.

09Post-migration cleanup

Remove the scaffolding safely.

Safe after: One complete operating cycle, at least seven stable days, and owner sign-off on every blocking check.

Create final Fly.io exports and archive verification evidence.
Revoke temporary credentials, API keys, webhooks, and elevated roles.
Remove obsolete embeds, forms, jobs, integrations, and DNS records.
Keep the source intact through the approved retention window.
Cancel paid plans only after billing, legal, and recovery review.
Schedule the next Railway backup, access, and migration-playbook review.

Sources and maintenance

Built to be reviewed.

Tested 2026-07-19. Next scheduled review: 2026-10-19.

Fly.io: official migration documentationAccessed 2026-07-19
Railway: official migration documentationAccessed 2026-07-19

Know the shape of the move.

This timeline assumes

What survives the move.

Where each thing goes.

Make the move recoverable.

Create the source-of-truth backup

Runtime and routing manifest

State and configuration migration

The things most likely to hurt.

The app builds but runtime behavior differs

Both platforms write durable state

DNS rollback is unavailable

Do the work in this order.

Inventory and decisions

Export?

Backup and reconcile

Transform?

Map and pilot

Scale?

Final delta and switch

Open production?

Observe and close

Close rollback?

Cut over with a way back.

Cutover

Rollback

Prove the migration worked.

Remove the scaffolding safely.

Built to be reviewed.