Validate prior to implementation

Hi,

Looking for feedback on what other customers have done to validate data in M3? We are in our user testing and are researching different approaches to provide the business with the raw data to do validation of M3 information as well as our Mongoose applications that have been built.

Different options we have discussed is adhoc reports against base tables exporting to excel, compass (but that has 100 record limit and is only tables sent to the data lake), birst (but then we would have to bring every table into the data lake and model the data in order for them to get the data), install data lake jdbc driver and use dbeaver to connect (risk of bad queries being created).

How has others here done this for their implementations? Any feedback is appreciated

I am assuming it is NOT recommended to push every table to the data lake, correct due to space as well as throughput?

Find more posts tagged with

Comments

Legacy Contributor

Hi,

Which data do you want to validate? I supouse that you are implementing M3, ¿it is?

In our case, we validate all the info prior to live at M3 directly, except transaccional information, that was upload at the final moment and was validate after go live, but only because was a little part and no sensible data.

We did it by sql extractions and Excels with some formulas. But after this experience I notice that it could no be the best option.
Also, the IT team did some validations first.
Regards,

Andres.

tdouglass

Thank you for your feedback. What tool did you use for sql extractions. Unfortunately, our business person wants access to ALL tables, but we told him that was not an option. otherwise, you just used the M3 Screens? (yes it is M3). When you say upload, do you mean legacy data or data that you need to add for go live?

Legacy Contributor

We did extractions with Microsoft SQL Service Studio and pasting it on EXCEL by IT team after that we gave it to the business person.

Maybe you could use also SSIS because there are problems doing copying and past to excel by cells formats.. by experience it is sensitive to have human errors, you need to have it by hand to do not make mistakes, but it's possible, we did it and it was successful for us. Another alternative is use ODBC qrys to excel with a format that business person can undestand.

Obviously access to all tables of M3 structure is a crazy thing.. there are thousands of tables and the format for a non it person is not udestandable.

With upload, I'm saying all the data needed for go live. We had2 scenarios, data validated by M3 screens and data validated by sql extractions.

In our case we did a pre migration 1 week early to go-live, and no new data was add to te previus ERP (only transactional), so during this week all the key users validate the data by M3 screens manly and some SQL extractions to EXCEL. After this week we only add or adjust transaction data, like stock or financial part.

If you need more info maybe your consultant partners could help you, in any case feel free to ask me antihing.

tdouglass

I really appreciate your feedback Are you Multi Tenant? This may be a question for your IT team, but did you pull from the data lake, or the M3 tables directly? if M3 data directly, how did they connect Sql Service Studio to M3 tables? Or did they use API calls? It's okay if you don't have the answer and you don't want to follow up with IT. We have also posed the question to our consultant parner on how others have validated data during user testing (pre go live).

Copyright © 2025 Infor. All rights reserved.