Migration
Talend to Microsoft Fabric migration: replacing Talend jobs with Fabric pipelines
How to move Talend Studio jobs to Microsoft Fabric: inventory components from job XML, choose Dataflow Gen2, Warehouse SQL or Spark notebooks, map tMap and other components, and replace contexts and schedules with Data Factory pipelines.
Hemant Walchale
· 2 min read
Talend often moves data from source systems into a Microsoft SQL Server or Azure warehouse. When the organization standardizes on Fabric, those jobs can be replaced with Fabric’s own pipelines, Dataflows and notebooks, removing a separate ETL platform and keeping data, transformation and Power BI together.
This is the approach we recommend.
1. Inventory
Count components across all Talend jobs from their XML .item files:
grep -rhoE 'componentName="[^"]+"' --include=*.item . | sort | uniq -c | sort -rn
Collect context groups, joblets, routines, child jobs, schedules (Talend Administration Center or Talend Cloud) and run history. Retire jobs that no longer run.
2. Choose a Fabric engine per job
| Job profile | Fabric engine |
|---|---|
| Extract and load with light transformation | Copy activity plus Dataflow Gen2 |
| Set-based transformation of large tables | T-SQL in a Fabric Warehouse |
Complex logic, large files, heavy tJava use |
Spark notebooks in a Lakehouse |
3. Map components
| Talend component | In Fabric |
|---|---|
tDBInput, tFileInput*, SaaS connectors |
Copy activity, mirroring or Dataflow Gen2 connectors |
tMap |
SQL joins and expressions, merge queries in Dataflow Gen2, or DataFrame logic |
tFilterRow, tJoin, tAggregateRow |
WHERE, JOIN, GROUP BY |
tSortRow, tUniqRow |
Window functions such as ROW_NUMBER() |
tDBOutput with upsert |
MERGE in the Warehouse where supported, or Delta merge in Spark |
tJava, tJavaRow, routines |
SQL or Python in notebooks |
tRunJob, joblets |
Invoke Pipeline activities, shared notebooks or procedures |
| Context variables | Pipeline parameters and deployment pipeline settings |
| TAC or Talend Cloud schedules | Pipeline schedules and event triggers |
4. Reconcile and cut over
Run Talend and Fabric side by side and compare every target table. Watch nulls, string handling, rounding, date parsing and Fabric Warehouse’s case-sensitive comparisons. Cut over by domain and disable Talend schedules the same day.
The short version
Count components from the job XML, choose Dataflow Gen2, Warehouse SQL or Spark per job, translate tMap and friends, rewrite Java code, replace contexts and schedules with pipeline parameters and triggers, reconcile and switch off Talend by domain.
Compare with Talend to Databricks and Talend to Snowflake, or browse all migration guides.
Not sure where your own platform stands? Take the free 3-minute data platform scorecard. Unfamiliar terms are explained in the data engineering glossary.