Hello,

We are starting LN in new agency and ofcourse we have many data to fill in : item list, customers list...

We are tryind to avoid to type all by hand.

I tried Exchange but i find it a bit difficult when you have to prepare the input text file. You have to create the same structure as the structure of the table.
I wonder if it could induce data corruption for some tables that have many fields replicated in many other tables.

Is exchange safe for all data ?

Is there other way than Exchange to import data into LN ?

Is MS Office integration tool valid to do such operation ?

Regards.

Hello,

first of all I would NOT use an Office Integration Tool to integrate bulk data into any erp system. This tool isn't intended to do such tasks at all.

You can use Exchange indeed. You sure must have a plan or sequence which data you should insert first, which next and so on. E.g. first insert common data like the units and so on then check other requirements of the table tcibd001 (Item Main) and then insert all item data. Normaly on this way you become errors if some constraints aren't correct.

Alternativly you can use some SQL tool to insert data directly into the database. But here it can happen easierly that you don't see a specific constrain and run into errors later on. But I experienced this method is faster from the amount of work you have to do.

Regards,
Litrax

I would try to avoid going directly to the DB. Although, it would be very possible, using the exchange module is nice because it will not let inconsistent data go in your system. (in the long run that can potentially create a bigger issue that would be more painful to fix, you can save yourself hours of hassle)
It checks for data integrity as well as it does all the DAL checks as if you were using the sessions to enter data (if you select it to do so and it is always recommended to do so)
It definitely is designed for this kind of task. Just make sure you do a lot of testing beforehand.

Hi

I agree with others, and my experience has been that Data Migration with exchange is the simplest and safest(esp. with DAL) way for migrating data.

The only thing to take into consideration is that sometimes the error messages could be misleading.

Good Luck.

Thank you all for helping.
Indeed, exchange seems to be good for migrating from Baan to Baan but not from "non Baan" to "Baan".
Have a nice day.

I have used exchange to migrate from a non-Baan environment to a Baan environment with no issues. Obviously you have to understand where the data is coming from and going to, but it's the same concept as from Baan to Baan.

Vince

Thank you Vincent. It's very interesting.
Was it easy for you to prepare the ACSI files you imported into BaaN ?
Dit it took a long time ?
For me, the problem is to translate an MS Excel file with item data to an ASCII file that follows the right structure of the item's table on LN. I already exported a file with the table's structure . But ofcourse not all the fields are mandatory. So, some fields have to be empty, others have not... Moreover each field has its own lenght which must be respected in the ASCII file...
Thanks.

Medtun,

Exchange is a very useful tool.

We used it to export Items, Bills of Material, Routings, Routing Operations, Tools, Departments, Work Centers, Machines, Business Partners, Purchase Orders, Serialized Items, Clusters and a whole bunch more from a legacy IBM Mainframe and import it all into LN.

As mentioned before, make sure that you use DALs when importing. On the downside, some of the error messages generated on the import are a bit misleading.

We also use Exchange on a regular basis to make mass updates to certain tables and fields.

Mike

Thank you mike.

Medtun,

To use Exchange you need to understand it.

In two of your above threads, you mentioned that you have to put the fields in table field order. This is not true. Exchange gives you the option of pulling in fields in table order to make it easier for you later when you map the fields to the actual table fields. You can name the fields what you want. For example, if a field is not used, I give it a different name or "filler1", "filler2", to declare the space in the file, but not use the data.

When you map the fields to table fields, you can use default falues for the fields that are not mandatory.

Also, Exchange processes the fields sequentially. First it will do the fields that are in the primary index. Then it processes them in order of the field mappings. Therefore, if a field is dependant on another field, that other field should be mapped first. If that other fields value is calculated or converted to another value, you will have the pre and post values to work with for your dependant field.

Infor has a nice tool called Any2LN for migration data from a non-Baan System into ERP LN. Basically it´s a pre-defined Access Database that need to be filled. From the Access DB a dump is created that will be imported into ERP with the Migration Studio. I have used it in two projects and I found it very usefull.

It´s worth a look.

regards

Dieter

Thank you Dieter. Sorry for the late reply. A such product would be perfect. I'll contact Infor to ask for more informations about it.

Regards.

A2LN is good for a quick load of clean, perfect data. I have worked with A2LN, MANMAN2LN, and Baan2LN. While Baan2LN and MANMAN2LN worked ok, but far from perfect, I have 3 observations about A2LN.

1. Any modifications that have to be done, has to be done by Infor.
2. They don't handle duplicates correctly. When doing City Codes from addresses, it takes the first eight characters as a City Code. When two different Cities have the same first 8 characters, only one gets created and all the addresses for both cities will reference only one of them.
3. They don't use any DAL checks. One example is when you import items and purchase or sales item data. The general item had a valid Unit of Measure, and so did the Purchase Item Detail. But there was no coversion between the two. A2LN loaded it, but unit testing did not find it. A load with Exchange showed the problem from the DAL.

I know of 3 recent implementations, 2 with A2LN, 1 with Exchange. The 2 with A2LN had issues, and Infor was brought in to make changes. The 1 with Exchange was done by a person in the company and had excellent results.

Thank you William. Your point of view is also useful.
Concerning my need, i'm not decided yet. I'm a functionnal Admin. And even if i have the necessary technical background, i'm not sure to find enought time to spend it on Exchange. So A2LN would be interesting if it would be easier to use. But, if it could be not safe ! it won't be funny too...
Anyway thank you all for helping. I remains interested by any other advice or comment.
Regards.

Time is not a factor for the development.

If I use A2LN, I have to write an export from my old system in the field order of A2LN.

If I use Exchange, I export using the tools of my legacy system as is. In exchange I define my ASCII file definition in field order of my legacy system.

In this scenario, Exchange could be faster for development.

One more advantage of Exchange is that you can overwrite. If you find an error later, you can run an exchange to fix field data from the original data. I mentioned in a earlier post that you do not need to have your fields in table order.

With A2LN you can not rerun, you can only append records or truncate the table and start over.

Hi everybody,

I contacted Infor about Any2LN. they said it's a tool included in a package called ‘DataLoad’. It is not for sale. It's only used by infor teams when delivering missions of data migration.
Concerning my need, i'm trying to learn how to import files with exchange.

Thank you all.

A2LN is good for a quick load of clean, perfect data. I have worked with A2LN, MANMAN2LN, and Baan2LN. While Baan2LN and MANMAN2LN worked ok, but far from perfect, I have 3 observations about A2LN.

1. Any modifications that have to be done, has to be done by Infor.
2. They don't handle duplicates correctly. When doing City Codes from addresses, it takes the first eight characters as a City Code. When two different Cities have the same first 8 characters, only one gets created and all the addresses for both cities will reference only one of them.
3. They don't use any DAL checks.

Hello Toolswizard,

your observations are correct, but they are not specific to Any2LN. The same applies for the migraton studio (Baan2LN ?) for migrating data from Baan IV to ERP LN.

Any2LN can help a lot to migrate the master data very quick into ERP LN without knowing to many details about the data structure in ERP LN. Most probably some data need to be converted with Exchange, but this is then the exception. It´s the old 80/20 rule :-)

I have used Any2LN in two projects. One project went very smooth and Any2LN was absolut sufficient, in the other project a lot of work need to be done during and after the migration with Exchange. It´s depending on the quality of data in the old system and how much business processes will change.

Regards

Dieter