Seeking Guidance on Migrating Informatica PowerCenter Workflows to Databricks using Lakebridge

[chirag_nagar]

Hi everyone,

I hope you're doing well.

I'm currently exploring options to migrate a significant number of Informatica PowerCenter workflows and mappings to Databricks. During my research, I came across Lakebridge, especially its integration with BladeBridge, which appears to support automated conversion of Informatica logic into Databricks-native formats such as SQL, PySpark, and notebooks.

While the office documentation has been very helpful, I was hoping to get some additional insights or real-world experiences from the community to help guide our next steps.

I would truly appreciate your inputs on the following questions:
1. Has anyone here successfully migrated Informatica PowerCenter workflows using Lakebridge + BladeBridge

2. How was your experience regarding:

• Code quality
• Accuracy of transformations
• Workflow and orchestration logic coverage

3. Does BladeBridge support complex transformations and control flow logic in PowerCenter workflows?
Or is it primarily geared toward mapping conversions?

4. What does the licensing or pricing model look like for using BladeBridge via Lakebridge?
Is it included as part of Databricks support or offered separately?

5. Are there any demos or walkthroughs available that showcase the process of converting PowerCenter .xml files to Databricks code?

[MariuszK]

Good questions, I'm also curious. In the past, I supported this kind of migration using a custom solution to convert IPC to PySpark.

[Elyas]

 I'm also curious. Looking forward for the answers with a same kind of questions. And also I could see there is a code repo shared by databricks for the lakebridge GitHub - databrickslabs/lakebridge: Accelerates migrations to Databricks by automating key migration.... Please share the experience if anyone used the code repo for migration.

[thelogicplus]

@chirag_nagar You should definitely check out Travinto Tools—it's one of the most reliable and effective solutions for migrating Informatica (PowerCenter/IDMC) workloads to Databricks. I've been using it for the past three years during my time at Deloitte, and it's consistently delivered great results.

[dinshows15]

Dear all , 

Is there any tool can convert Informatica BDM mappings to Pyspark code. 

Lakebridge is not able to do it , it has limitation and it can work only for Powercenter mappings. 

Down the line they may come up but as of now only till Power center. 

While compairng xmls of powercenter and BDM its huge diff in format and overall structure of xml.

Informatica made it quite complex to make it difficult to get converted to other tech like Pyspark by using these kind of tools. 

[Kalyan4]

Hi, Currently there’s no mature tool that can fully convert Informatica BDM mappings into PySpark. Most converters (like Lakebridge) support PowerCenter only, since BDM is metadata-driven and its XML structure is very different. BDM → PySpark typically needs partial automation plus manual refactoring. Sharing the BDM complexity (pushdown, CDC, mapping types) can help identify the best approach.

[thelogicplus]

Hi @dinshows15  You can use Travinto Technologies Code converter tool. They have very good tool and it supports informatica BDM as well as.  Please check with Travinto .

Lake bridge(Blade bridge) is the worst tool that I have used .

[dinshows15]

Thansk for the update @thelogicplus  I will have look and try to connect with team of Travinto.

[AnnaKing]

Hi Chirag. At Kanerika Inc,, we've built a migration accelerator that automates 80% of the Informatica to Databricks migration process, saving you significant time, effort, and resources. You can check out the demo video of the same here - https://www.youtube.com/watch?v=Fc46f1LzV-4
You can also explore  - https://kanerika.com/services/migration/informatica-to-databricks/
To discuss about the solution, please reach out to us at https://kanerika.com/meet

[dinshows15]

Thankyou for your update. 

How can we convert the informatica bdm or dei mappings or workflows by using any tool like Flip to Pyspark to migrate business logic written in informatica to Databricks.

Pl let me know.

Thanks 

Dinesh 

[AnnaKing]

While I check this for DEI Mappings, would it be possible to give me more info regarding your current setup and BDM workflows? I am sure our migration tool can be enhanced to process BDE logic.

[dinshows15]

Thanks @AnnaKing , we are having informatica bdm mappings and workflows. dei is just new name given to bdm by informatica. My question is which tool can convert these bdm mappings in pyspark code.as we know lakebridge tool only can convert from powercenter not from bdm.

[AnnaKing]

Hi @dinshows15. Kanerika;s Informatica->Databricks using FLIP can handle BDE Mappings as well. It covers both PowerCenter as well as Informatica Developer (BDE/Java Designer) repositories. Let us know if you would like to setup a separate session for one-on-one discussion with our Solution architect.

[devangpanchal]

Hi,

We've helped teams work through this exact decision (Synapse dedicated pool plus Spark ingestion, Fabric-only vs keeping Databricks). A few points that usually settle it:

1. Split the decision by workload, not by platform.

- T-SQL warehouse workloads (stored procedures, CTAS patterns, views feeding Power BI) map closely to Fabric Warehouse. The Migration Assistant can take a DACPAC from the dedicated pool. Note that distribution settings (HASH, REPLICATE, ROUND_ROBIN) don't exist in Fabric, so that DDL has to be rewritten.
- Spark-based API ingestion can run in Fabric notebooks, but check library dependencies and networking first. Private endpoints and managed VNet behaviour differ from Synapse.

2. Plan for shared capacity, which is the part teams miss most often.

In Fabric, Spark jobs, warehouse queries and Power BI all draw from the same capacity. A heavy ingestion run can throttle report users at the same moment. Before sizing a SKU, migrate one subject area and watch the Capacity Metrics app over a full business cycle, including month-end.

3. On mirroring and Unity Catalog security.

As you noted, mirroring does not carry Unity Catalog permissions into Fabric. If you go hybrid, decide upfront which platform owns access control, and re-implement permissions on the Fabric side for anything users query there. Otherwise you end up maintaining two security models that drift apart.

4. A simple rule of thumb.

- Team is mostly T-SQL and Power BI, no ML or streaming roadmap: Fabric-only is a reasonable target.
- ML, streaming or a multi-cloud requirement: keep Databricks for engineering and use mirroring to serve BI in Fabric.

Synapse to Fabric versus Databricks is one of the most common questions that comes up in the data warehouse consulting work Bacancy Technology provides. The platform choice is rarely the hard part. Capacity planning and the security model are where projects slip, so it is worth settling both before committing.

Happy to go deeper if you can share rough data volumes and how much of the Synapse estate is T-SQL vs Spark.