Magento migration: moving off Magento 1, and between 2.x releases
Magento migration means three different projects: rebuilding from Magento 1, upgrading within 2.x, or leaving the platform, each with distinct data, code and redirect work.
One word, three different projects
“Magento migration” can mean rebuilding a Magento 1 store, upgrading an existing 2.x installation, or moving the business to another platform entirely. These are not variations of the same technical job. One replaces unsupported software, one preserves an established codebase while updating it, and one replaces the platform, integrations and operating model.
The distinction matters because the greatest cost is rarely the transfer itself. It comes from work that is discovered late: incompatible extensions, incomplete customer records, bespoke checkout logic and old URLs that nobody mapped before the launch.
Whatever the route, migration should begin with an inventory of URLs, data and dependencies. That inventory determines whether the project is manageable and provides the acceptance criteria for the new store.
Magento 1 to Magento 2: a rebuild, not a move
Magento 1 reached end of life on 30 June 2020. A store still running it has been without vendor support for years, and the practical options are to rebuild on Magento Open Source or Adobe Commerce or to leave Magento. Carrying the existing installation forward is not a credible way to retain the old codebase.
The distinction is simple: business data can be migrated, but the theme, extensions and custom code generally cannot. Magento 2 is a different application with a different database model, administration system, rendering architecture and extension ecosystem. Treating the exercise as a file transfer produces a fragile result.
What the Data Migration Tool covers
Adobe’s Data Migration Tool is designed for structured data from Magento 1. It handles core entities such as products, categories, customers, quotes and orders, along with supporting catalogue and content information where the entities and mappings permit.
The tool is strongest when the source store used Magento’s standard data model and the destination has a clean configuration. Problems arise where merchants have extended products, customers or orders with custom attributes, or where store scopes, customer groups and catalogue rules do not have direct equivalents.
A successful import does not prove that the data is usable. Taxonomy may need rebuilding around navigation rather than the old category tree. Product attributes may have changed meaning, order status mappings may not reflect the new business process, and customer records may need to be reconciled with the marketing system used by the new store.
The tool does not convert a Magento 1 theme into a Magento 2 theme. Nor does it make an incompatible extension compatible. Custom tables, bespoke modules, payment logic, shipping rules and integration code need individual decisions.
What has to be rebuilt
The frontend needs a new implementation. Luma is the default Magento storefront theme, with Blank as its parent theme, but installing it is only the starting point. Templates, layout changes, JavaScript behaviour and any custom responsive or accessibility work from Magento 1 need to be recreated and tested.
Extensions need equivalents rather than copied files. For every Magento 1 extension, establish whether a supported replacement exists, whether its data can be migrated, whether it uses a current API, and whether it remains compatible with the chosen release. An extension that merely imports its tables into Magento 2 is not a supported solution.
Server-side customisation needs a port rather than a transplant. Magento 1 code will not run safely on Magento 2 without review and usually without modification. This includes checkout customisation, observers, cron jobs, API integrations, tax logic, promotions and operational scripts.
Payment providers, carriers, ERP connections, accounting exports, search, email, loyalty and subscription services also form part of the rebuild. A data report showing thousands of imported products can conceal the fact that orders cannot be fulfilled or invoices cannot be reconciled.
The practical sequence is to capture the source catalogue and behaviour, build the destination configuration, import data into a test environment, then rebuild integrations and user-facing functionality around it. A full data import repeatedly during development is not a substitute for representative test data.
Moving between Magento 2.x releases
An upgrade within Magento 2.x is usually a controlled software change rather than a new store. Core customer, order and catalogue data remains in place while the application, database schema, dependencies and extension code move to the target release.
This is the most common migration now, but it should not be confused with a routine server update. Magento releases change compatibility requirements, database structures and extension contracts. The target release must work with the merchant’s catalogue logic and third-party stack, not merely install successfully.
Support dates determine the order of work
As at 13 November 2026, the support position is clear:
| Release | Support position |
|---|---|
| 2.4.9 | Current general release. Regular support runs to 31 May 2029. |
| 2.4.8 | Regular support runs to 31 May 2028. |
| 2.4.7 | Regular support runs to 31 May 2027, with extended support to 31 May 2028. |
| 2.4.6 | Regular support ended on 11 August 2026. |
A 2.4.6 store is therefore facing an unsupported platform before migration starts. The remedy is not a cosmetic patch to the existing installation. It is an upgrade project with its own compatibility testing, deployment plan and rollback decision.
Magento 2.4.9 was released on 12 May 2026. The 2.4.10-alpha1 release that appeared on 13 October 2026 is a pre-release and is not suitable for production. An alpha may help identify early compatibility issues, but it does not provide an appropriate migration destination for an operating store.
Why the requirements list can enlarge the project
Adobe lists only the following supported combinations for Magento 2.4.9 on-premises:
- PHP 8.5
- MariaDB 12.3 or MySQL 8.4
- OpenSearch 3
- Valkey 9 for cache and sessions
- RabbitMQ 4.3
- Composer 2.10
This is a substantive change. Elasticsearch and Redis are not listed as supported for 2.4.9. A store that assumes it can keep its current search, cache or session services without checking may find itself replacing infrastructure as part of the upgrade.
The jump can therefore include more than Magento code. PHP may need application changes, the database may need a version change, search indexing must work with OpenSearch 3, and queues or cache behaviour may need review. Hosting must satisfy the target combination rather than merely continue to run the old release.
Composer dependency constraints deserve particular attention. Current versions of extensions, payment modules and custom packages must support the target release and its PHP requirements. An extension that installed cleanly on an older release may not be compatible with 2.4.9.
Upgrade work is more than replacing files
Custom modules must be tested against changed APIs and framework behaviour. Admin workflows, frontend rendering, checkout, asynchronous jobs and third-party webhooks can fail in ways that a storefront page request does not reveal. Use representative orders, including refunds, failed payments, partial shipments and customer account actions.
Database migrations, indexers, cron jobs, message queues, static content and cache invalidation also need explicit verification. A successful deployment script is not evidence that the store is healthy. Order placement, invoice creation, email delivery, search indexing and administration should all form part of the release gate.
Take a full backup and know how it will be restored before changing production. Test the upgrade on a copy of production data and infrastructure, record the exact dependency versions, and rehearse the cutover. The launch window should allow for troubleshooting rather than assuming every failure can be corrected immediately.
Leaving Magento entirely
Leaving Magento is legitimate. A platform change can be justified by operational fit, extension availability, internal skills, hosting requirements or the cost of maintaining custom work. It is not automatically cheaper, and the commercial case depends on the licences, infrastructure, support and development burden of the destination.
For a merchant considering how to migrate from Magento to Shopify, or comparing another platform, the platform’s import tools should be treated as assistants rather than proof of completeness. They may handle conventional product and customer files, but they cannot infer the behaviour of a Magento installation.
What can transfer
Products are usually the easiest category to transfer once identifiers, variants, options, images, stock and categories are cleaned. The destination may accept standard product fields, but the merchant still has to decide how bundles, configurable products, custom options, unit-of-measure rules and marketplace feeds should behave.
Customer records can generally be exported and imported, subject to consent, privacy requirements and field mapping. Historical order records can often be imported as order history rather than as live operational orders. That distinction matters for refunds, claims, invoices, fulfilment records and accounting.
Collections, promotions, content and media may need separate exports or recreation. A clean Magento catalogue can save effort, but a catalogue carrying inconsistent categories, duplicate values and missing metadata will reproduce those problems in the new platform.
What does not transfer
Customer passwords do not transfer to another platform. The destination will normally require customers to set a new password or use a verification flow. Account records can survive; authenticated access cannot simply be copied across the migration.
URL structures do not transfer automatically. Magento URLs may encode product identifiers, category paths or store views that the destination handles differently. Existing search visibility depends on preserving those entry points through deliberate redirects.
Custom checkout logic, membership rules, subscription behaviour, gift cards, wish lists, loyalty balances and bespoke pricing need rebuilding or replacement. Payment tokens, stored credentials and integration-specific data may be tied to the old platform and cannot be assumed to be portable.
Extensions do not transfer as extensions. Their functionality may be available through another app or may need custom development, and the destination’s operating rules may produce different behaviour. Integrations with ERP, PIM, tax, fraud, customer service, email and accounting systems must be mapped again.
The SEO cost can outweigh the platform saving
For an established store, a careless migration can cost more in lost organic visibility and delayed recovery than the new platform saves in licence or operating cost. Search engines need time to process a substantial URL change, even when every old address has a valid destination.
Migration is also an opportunity to improve the catalogue, but changing URLs, titles, descriptions and category structure simultaneously makes causes difficult to isolate. Keep strategic SEO improvements separate from the minimum changes required to move the catalogue safely.
The redirect map: do not skip it
Across all three migrations, the redirect map is the item most often deferred and the item most expensive to repair after launch. A redirect cannot be reconstructed reliably once the old catalogue, rewrite rules and store configuration have disappeared.
Begin by exporting every indexable URL from the old store. This should include product, category, tag, CMS and landing-page addresses, along with the store scope and canonical URL where relevant. Check the server logs and search data for traffic that does not appear in the catalogue navigation.
Map each old URL to one deliberate destination. Use a permanent redirect for a permanent move. Where a product has a close equivalent, point to that product. Where it has no equivalent, choose the nearest relevant category or a genuinely useful landing page rather than sending every removed product to the homepage.
Watch for redirect chains, loops and redirecting URLs that themselves redirect. These weaken the signal passed to the destination and can create crawling and indexing problems. Preserve valuable parameters only when they have a defined meaning and do not create duplicate destinations.
Test the map in a non-production environment, then test it again after DNS changes and CDN configuration. Sample redirects manually, crawl the new site, confirm that destination pages return the expected response, and monitor server logs and Search Console after launch.
Keep the redirect file outside the new application’s editable content. A deployment, import or CMS edit should not quietly remove or overwrite it. Record why each significant mapping was chosen, especially where the old page has no direct replacement.
Plan the migration as an inventory
Before choosing a route, produce three inventories: URLs, business data and system dependencies. The URL inventory provides the redirect map. The data inventory records fields that must be cleaned, transformed or deliberately abandoned. The dependency inventory identifies extensions, custom code, integrations, infrastructure and operational processes.
For Magento 1, the first decision is whether the unsupported implementation can be rebuilt within the available window. For 2.x, compare the current release with the supported upgrade target and test the complete stack against its requirements. For a platform exit, assess the destination’s native behaviour before assuming an import tool can bridge the difference.
Define the checks that prove completion. They should include checkout, payment authorisation, tax, fulfilment, refunds, account access, search, email, reporting, integrations, redirects and representative product and customer journeys.
The common mistake is treating migration as a cutover. It is better managed as the replacement of a connected system: the catalogue continues to earn traffic, orders continue to move through operations and customers continue to expect their accounts and history to work.
Done properly, the technical transfer is only part of the result. The store remains usable, URLs retain their value, operational processes continue to function and the chosen platform has a clear reason for existing.
Questions this page answers
- What does Magento migration mean?
- It can mean rebuilding a Magento 1 store, upgrading an existing 2.x installation, or moving the business to another platform entirely.
- What can Adobeu2019s Data Migration Tool transfer?
- It handles structured Magento 1 data, including products, categories, customers, quotes and orders, where entities and mappings permit.
- What must be rebuilt from Magento 1?
- Frontend implementation, extensions, server-side customisation, payment providers, carriers, ERP connections and other integrations generally need rebuilding or porting.
- Which Magento 2.4 releases have support?
- As at 13 November 2026, 2.4.9 is the current general release. Regular support runs to 31 May 2029 for 2.4.9, 31 May 2028 for 2.4.8 and 31 May 2027 for 2.4.7.