BizTalk Migration
November 4, 2026
If you run a BizTalk Server estate, BizTalk migration has stopped being a someday project. Microsoft has confirmed that BizTalk Server 2020 is the last version it will ship, new sales are expected to stop on 31 March 2027, and mainstream support ends in April 2028. There is no next release coming to reset the clock
For most integration teams, the hard part isn’t deciding whether to move. It’s moving without breaking the order flows, EDI exchanges and finance interfaces that have run quietly on BizTalk for a decade. Many of those interfaces were built by people who have since moved on, documented lightly, and patched in place more times than anyone can count.
This guide is for the integration lead or architect who owns that estate. We’ll walk through the verified lifecycle dates, how to inventory what you actually run, how to choose between lift-and-shift and re-architecture, and how to stage the cutover with a rollback plan and a hybrid period that keeps production safe. A BizTalk migration checklist at the end turns it all into a working plan.
In This Blog
BizTalk Server 2020 follows Microsoft's Fixed Lifecycle Policy, and the milestones below come straight from Microsoft's own lifecycle and migration pages.
| Milestone | Date | What it means for you |
| BizTalk Server 2020 released | February 2020 | Final version of the product. No successor release is planned. |
| End of sale (expected) | 31 March 2027 | You can't buy new BizTalk licenses after this date, which matters if you need to add capacity or environments. |
| End of mainstream support | April 2028 | No new fixes or enhancements after this point unless you buy extended coverage. |
| Paid extended-mainstream support | 13 April 2028 to 10 April 2030 | Optional, paid and limited to eligible customers. Microsoft positions it as temporary migration cover, not a long-term home. |
The last row is the one teams most often misread. Extended support to 2030 is not a blanket safety net. Microsoft expects it to exclude several capabilities, including Business Activity Monitoring (BAM), the ESB Toolkit, industry accelerators, legacy enterprise adapters, BizTalk-specific ALM tooling, and Visual Studio versions newer than 2022. If your estate leans on any of these, your real deadline is April 2028, not 2030.
Working back from those dates, the window is tighter than it looks. A mid-sized estate with a few hundred interfaces typically needs time for discovery, design, a pilot, several migration waves and a parallel-run period for each. Teams that start planning in 2026 can migrate on their own schedule. Teams that start in 2028 will be migrating under pressure, often while paying for extended support at the same time.

Every successful BizTalk migration starts with an honest inventory, because you can't estimate, sequence or de-risk what you haven't counted. The BizTalk Administration Console tells you what's deployed. It doesn't tell you what still matters, who depends on it, or what breaks if it stops.
Build one register per BizTalk application and capture at least the following.
| Artifact | What to record | Why it matters |
| Orchestrations | Long-running or short-lived, correlation sets, compensation logic, convoys | Long-running and convoy patterns need the most redesign effort |
| Maps and schemas | XSLT maps, custom and scripting functoids, flat file and EDI schemas | Plain maps often carry over; custom functoids usually need rework |
| Pipelines | Custom pipeline components, disassemblers, encryption and signing steps | Hidden .NET code often lives here |
| Adapters | Built-in, line-of-business (SAP, Oracle, mainframe), WCF-Custom, third-party | Each needs a confirmed equivalent connector on the target platform |
| Ports and bindings | Receive locations, send ports, filters, schedules, retry settings | This is where routing logic and partner endpoints actually live |
| Business rules | Business Rules Engine policies and vocabularies | Some can be reused, others need refactoring |
| BAM and ESB Toolkit | Activities, views, itineraries | Expected to be excluded from extended support, so plan these early |
| Custom code and config | Helper assemblies, SSO config store entries, certificates | Secrets and certificates are a frequent cutover blocker |
Then add the runtime picture: message volumes, peak windows, payload sizes, SLAs, the business owner, and every upstream and downstream system. Tracking data and host instance metrics will show you which interfaces are busy and which have been silent for months.
With that register in hand, give every interface one of four labels:
Microsoft's Azure Logic Apps Migration Agent can help with discovery and analysis if Azure is your likely target, but treat any automated output as a first draft. The business context, ownership and risk ratings still have to come from your team.
Most estates end up using more than one approach, so the real question is which approach fits each group of interfaces. Here's how the three main paths compare.
| Approach | Best when | Watch out for |
| Stay and stabilise on BizTalk 2020 | You need time to plan and older versions are your bigger risk | It buys time to 2028 at best. It is not a destination. |
| Lift-and-shift to a modern platform | Interfaces are stable, well understood and mostly maps, schemas and simple routing | You can carry old design debt into a new platform and pay for it twice |
| Re-architect | Flows are business-critical, brittle, or built around MessageBox publish-subscribe and long-running orchestrations | Takes longer and needs strong business involvement to confirm requirements |
A practical rule: lift-and-shift the many simple interfaces to build momentum, and re-architect the few that carry real business risk. Translating a fragile orchestration line by line usually just moves the fragility.
Microsoft names Azure Logic Apps Standard as BizTalk's successor, and for Microsoft-centric teams it's the natural first look. It isn't the only credible option, though. Your choice should follow your hosting, skills and integration strategy, not just your current vendor.
Score each option against the same criteria: where workloads must run (cloud, on-premises or both), EDI and B2B depth, the connectors your adapters map to, your team's skills, the licensing model, and how well the platform supports your wider API, data and automation roadmap. A BizTalk migration is a rare chance to fix the integration foundation, not just replace it.
If Azure is your target, plan on Logic Apps Standard rather than Consumption, since Microsoft's BizTalk migration capabilities only support Standard workflows. The two platforms are built differently, so think in terms of equivalents, not one-to-one copies.
| BizTalk component | Logic Apps Standard equivalent | Typical effort |
| Orchestrations | Stateful or stateless workflows | Refactor the process logic |
| MessageBox publish-subscribe | Azure Service Bus topics (or RabbitMQ, Kafka) | Redesign as a separate messaging layer |
| Schemas and maps | XSLT maps, Liquid templates, Data Mapper | Reuse where compatible, then validate |
| Pipelines and helper code | Workflow actions, .NET local functions, custom connectors | Reuse or refactor |
| Business Rules Engine | Azure Logic Apps Rules Engine | Reuse supported policies |
| EDI and B2B | Integration accounts, B2B connectors | Reassess partners, agreements and certificates |
| Adapters | Built-in, managed or custom connectors | Confirm availability and throughput |
| BAM and tracking | Business Process Tracking, Azure Monitor, Application Insights | Redesign monitoring and support model |
A few design shifts catch teams out:
The fear of breaking production is legitimate, and the answer is to never cut over anything you can't reverse in minutes. Big-bang migrations are where BizTalk projects go wrong. Waves, parallel runs and pre-agreed rollback triggers are where they go right.
This takes longer on paper than a single weekend cutover. In practice it's faster, because you avoid the incident reviews, emergency fixes and loss of business confidence that follow a failed big-bang go-live.

For most estates, BizTalk and the new platform will run together for months, so design the coexistence period on purpose rather than letting it happen by accident. These are the patterns that work well.
Budget for the overlap as well. You'll pay for BizTalk, SQL Server and Windows licences, the new platform's capacity, and the people supporting both for the length of the transition. A shorter, well-planned hybrid period is almost always cheaper than a long, unplanned one.
Use this BizTalk migration checklist to turn the steps above into a plan your team and leadership can sign off.
Plan
Discover
Design
Build and test
Cut over
Close out
BizTalk end of life is a fixed date, but how you get there is still your call. The teams that come through it well are the ones that inventory early, choose the right path for each interface, and treat rollback and coexistence as part of the design rather than an afterthought.
Tellestia has delivered more than 100 data and integration projects over six-plus years, with hands-on expertise across Azure, Boomi, MuleSoft, WSO2 and Workato. That platform range means we can recommend the target that fits your estate, not the one we happen to sell, and then help you migrate to it in safe, staged waves.
Explore our Legacy Middleware Migration services to see how we take integration estates from BizTalk and other legacy middleware to a modern platform, or talk to our team about an assessment of your BizTalk estate.
