Infor OS as a Data Architecture Backbone

Hello,

My organization is working toward building a data catalog and getting serious about data architecture. I would like feedback from Infor customers, staff, and partners about how useful Infor OS might be as the backbone of this new catalog and architecture.

Thanks, Nick

Find more posts tagged with

Infor Data Fabric

Comments

Andrew Sowden

We too are creating a comprehensive Data Catalogue - we are calling it the "Gilberts Graph" as it will have all the data sources including outside of Infor, including derived fields and how they are calculated in a standard way.

It would be interesting to see how other people are managing this too.

John Hart

We're also working on a project like this. We're trying to bring the data from all our sites to Snowflake. We have 16 sites on 5 different ERP systems. The biggest problem I've been having is finding a way to reliably extract data from the Data Lake so it can be ingested into Snowflake. We're on a multu-tenant, IGS system and it seems like every technology Infor has is geared towards commercial clients and not available for govcloud systems. It's been very frustrating.

Enrico Menotti

Well, I'm working on Birst reporting from both LN (Cloud) and an old SAP R/3 (on prem). For LN I'm stepping through the Infor Data Lake. However, for SAP I decided to use a direct connection from Birst to the Sybase db (Connect 2.0, aka Agent). There are two main reasons: first, while LN is actively publishing data to the Data Lake, I don't have any way for SAP to do the same. In particular, I see no way to mark deleted records. The second reason, considering that the Data Lake would not add any value to the process (for the first reason), is the sake of keeping things lean - and by connecting directly from Birst I'm saving one step in the process.

Kevin Heiman

The IGS environment supports;

the Data Lake APIs so the ION ETL tool can be used to pull data or any API capable client application can pull data.
Other methods could consider ION data flows to send to File FTP location
Access by Compass is not supported in IGS.
Streaming Pipelines is not supported in IGS

Have you explored these options? Please reach out if needed.

It is customer responsibility to manage the IGS data boundaries.

Nick James

Hey Enrico,

So you feel that trying to build out the SAP data in the Infor Data Lake would have too much overhead? Are you satisfied with the way that Birst interacts with the SAP data?

Thanks,

Nick

Enrico Menotti

Hi Nick,

As I said, the Data Lake, IMHO, makes sense when you keep all variations (including deletions) of records, which is only possible if the source system is actively pushing. In my case, the most I can do is to query SAP regularly, once a day let's say, and check what's there. This way, I'm missing all deleted records and all variations which happen more often than my query period. So I'm going to only look at the current data, and keep those only, which is something doable without the Data Lake.

Birst is capable of querying directly Sybase (some setup needed... along with quite a bit of headache), and so long I'm happy with how this goes - but I'm still developing. On the other hand, the only way I have yet for moving data to the Data Lake is by ABAP exports to .csv files (there should be some more direct way but I didn't investigate yet). Then the Data Lake keeps those raw files in Atlas as its original storage, and creates sort of a cache in Compass (queryable area they call it) - which looks to me very inefficient, due to the sub-optimal storage, the slowness of the first query (when Compass is still empty) and the multiple steps involved. Maintenance is sometimes required (notably, when new fields are added), which means purging both Atlas and Compass and publishing again - and loosing previous variations. All of this, again IMHO, only makes sense for LN, where the source system is actively pushing data - and there is no other way to get our hands directly on the LN database.

Copyright © 2025 Infor. All rights reserved.