What is the Best Way to Share Data Using ION?

Good morning,

I received a request from the CEO's office to share an API Key for our On-Prem Infor Public Sector (IPS) "Customer Call" data. We are sharing data with a third party software that does data mining. I know there are several ways to do this directly from our IPS Rest Services, but I thought this would be a perfect opportunity to use ION. My IPS on-prem has the CustomerCall BOD built-in, so I was thinking it would just be a matter of building a data flow and dropping the data into the Data Lake. I am not 100% sure how to achieve this, or how to share the data out in a secure manner once it is available.

I am still relatively new to creating flows in ION.

Any and all input is appreciated!
Thanks,

Nick

Find more posts tagged with

General Discussions

Accepted answers

All comments

Nick James

OK, I was able to get the data into the lake, I just had to fix a connection issue... What do I do with the data next? Or can I just bypass the lake and serve it out to the vendor directly?

Thanks,

Nick

ION Data Lake Flow

unknown

Nick:

If you put the data into Data Lake, your vendor still cannot access it. You are public sector so I presume you are licensed for the Enterprise Connector, so you can define a File Connection Point, that is one option; the other option are ION Back End Services.
Which ERP is in scope here, FSM, EAM, LN M3 ?

unknown

Hey Nick,

IPS supports publishing data to Data Lake but I would recommend checking out the documentation for IPS on how and what is supported in this context.

Data Lake is a distributed data platform with one part storage and another part compute. You've put some data into the Data Lake but the kind of data you've put in is meaningful. While Data Lake can store BODs - canonical XML documents - you cannot use the Compass query platform to query BODs today.

However, IPS does replicate data as .ndjson and when stored in the Data Lake, you can use the Compass query platform to query that data and provide applications and clients with access to do the same through services like the Compass JDBC driver or Compass APIs.

The Data Fabric playlist on the Infor OS YouTube channel might be a bit helpful here in understanding this better as well.

You can see what you're storing in your Data Lake by using the Atlas feature under the Data Lake navigation menu in ION Desk.

I'm not 100% sure on your use case here and if you're only just trying to integrate with a 3rd party vendor but if it's that simple, you may wish to instead use the ION integration capabilities to publish data from IPS (as you've done already) and register a connection point for your vendor based on their technology capabilities (e.g. file server, database connection, queue, RESTful API, etc.)

unknown

Mike, if he needs to share the data, he can opt to publish from Datalake through say a File Connection Point and publish to an SFTP, and provide key access to the SFTP server, or created a BackEnd Service for the Vendor and provide the credentials to the Vendor, right ? You assign the correct roles through Mingle to narrow the Vendors access to the Business Classes. The vendor would need to build the API calls to fetch the data.

unknown

You could actually do both but which approach you take typically aligns with the use case at hand. Duh, didn't need to say that But I suspect there's more detail here so I'll be overly exhaustive.

You can use ION as an event-driven, pub-sub integration platform to deliver those files to some target destination like an sFTP.

Alternatively, you could store and stage the data in Data Lake and then either use ION's Data Lake Flows to schedule retrieval of data objects to then be pushed to some target or the Data Lake's Payload APIs to have an API client retrieve it. This is usually oriented with batch-based integrations.

In both cases, this is ultimately delivering some payload or file. If the data is in a queryable format (e.g. ndjson, CSV, TSV, etc.), then you could use the Compass platform to process queries against files in storage.

You can register an Authorized App for an outside, 3rd party client to handshake with the gateway and run API calls where they're permitted (oAuth scopes).

Back to your original question - yeah, the option you mentioned on a backend service (Authorized App) would be viable but security restrictions might be a bigger topic that is specific to the solution being leveraged. If you're calling the application's APIs (e.g. IPS, CSF, EAM, etc.), then they may have security configuration and ACL settings to determine whether someone can tap into a particular data model, API method, and so on.

That's where specifics play a bigger role in determining the right approach to solutioning here but at least a rough guide on different capabilities and approaches.

Rees Howel

Let's try again. If you are familiar with Python, I believe that you can use the Scripting feature to transform a BOD to ndjson as an additional step in the dataflow. Then, as I think Mike suggested, you can query the ndjson in the datalake to extract to your 3rd party. (*Now* I can touch Return :-) )

Nick James

ReesH,

Unfortunately, I have not done much Python scripting. I am typically a Copy & Paste coder, but typically if I have a good base of code to start with I can figure out how to make it sing. The problem is that Infor doesn't seem to put out a whole lot of tutorials, and the Help files will only get you to a certain point. My initial training in ION was back in 2017. I don't think scripting was even available then, and I believe that Data Lake was Business Vault (which we never even configured beyond the ICS install).

As you can see from the image I posted above, I was able to get the BOD into the Lake but I honestly do not know what the next step would be...

Thanks for the input,

Nick

unknown

Nick:

You have to decide which is the best connection option for you vendor. If you read through Mike's suggestions,

[quote userid="16617" url="~/infor-ion/f/general/24133/what-is-the-best-way-to-share-data-using-ion/64708"]You can register an Authorized App for an outside, 3rd party client to handshake with the gateway and run API calls where they're permitted (oAuth scopes)

This is an option if you encourage the 3rd party to connect to the tenant, they are granted access to the tenant through ionapi credentials. If you do this then the 3rd arty is responsible for the solution

The other option is using an SFTP server and keys. You can define a File Connection Point through the ION Enterprise Connector, create an ION data flow to move the documents from the datalake to the server running the connector, and publish interface files to a folder on the server running the connector. From there you can move the files to the SFTP server.

Another option is using the JDBC driver in a similar fashion. You need to define a back end service for the driver, connect it to a Mingle ID, download the inapi credentials. You can use the JDBC driver on any server you wish, e.g. an SFTP server, build an interface solution to extract the records to the SFTP server, and provide 3rd party connectivity to the SFTP server.

Yet another JDBC option is creating a back end service and accounts dedicated to the 3rd party, providing them with the driver and the ionapi credentials, then it is up to the 3rd party to build a JDBC client e.g. java package to read the documents you are publishing to Datalake. Security is an issue with this approach.

Another option is allowing the 3rd party to make an API call directly to Hansen. That is actually faster than the Datalake (because you don't have to convert the S3 storage class) but you better understand your Hansen security model. I grant that type of access for mature FSM solutions ; ClincalBridge or MHC

Once you decide on the 3rd party connection, I can make solution suggestions.

Nick James

Thanks for the response. The Authorized App approach is what I would like to explore.

Quick Links

Developer Community
Infor OS Platform Forums
Hackathon
Developer Portal
Developer Newsletter

Copyright © 2025 Infor. All rights reserved.