The database migration

, , ,
1

Hi!

Are there any ways to move or converting the ADABASE database to another database (MySQL, PostgreSQL, MSSQL, Oracle etc.)? Maybe there are some tools?

Thanks in advance!

2

There is an add-on product for Adabas, the Event Replicator for Adabas that can be used for this purpose.

1 Like
3

Is this a one time conversion? Or are the Adabas files to be kept in synch with other files?

1 Like
4

Hi Steve!

Now, the database is in use. I thought moved to another database, and start writing new software. And when the software is will written - migrate ADABAS database again (new Data migrate) and migrate users to the new software.

I will be very happy if there is a way to implement synchronization for time until written software. This will greatly facilitate the work.
Well, if it is highly difficult task, then I will have to implement the first version (although I still do not know how…).

Wolfgang written about Event Replicator for Adabas, but I just started using ADABAS and I do not know where to find this tool. If this a part of ADABAS then how I can run it?

Thanks in advance!

Best Regards,

5

There are several software packages from 3rd parties, as well as the Adabas replicator that will “map” an Adabas file to an SQL file.

However, this might not be a trivial undertaking. If you are new to Adabas, as mentioned in your last post, you are probably not aware of MU (multiple valued) fields and PE (periodic) groups.

For example. Suppose I have an employee file. An employee might be assigned to more than one project. This might be implemented via an MU field. Basically, Adabas does a lot of the bookkeeping for you. In order to convert to an SQL file, you will have to ascertain how you want this implemented.

Same employee file. An employee could have multiple addresses. Instead of a single field repeating (as described above for project number), you now have multiple fields repeating as a “group” of fields. Again, you will have to decide how you want the “mapping” to be defined.

It sounds as if synchronizing the two files would be “overkill”. I would start by creating an SQL “play file” by mapping selected records from the Adabas file. Then I would use the play file to write and test the new code. This would also help you decide if you did the “proper” mapping for MUs and PEs (oh yes, there can be multiple valued fields within periodic groups). Then, when you are certain the new code works, I would map the Adabas file to SQL, creating a “production” version of the Adabas file as an SQL file.

1 Like
6

Thank you very much for your help, Steve!

I am familiar with the MU and PE and they are largely present in my database.

I wanted to go this way. But I do not know how to create a SQL file from ADABAS file. Maybe you can tell me?

I’m thinking of doing this:

  • by NATURAL make a text file table
  • from a text file into a new database

But if you can create a SQL file, it will be the best solution in my opinion.

Thanks in advance!

Best Regards,

P.S. Sorry for my English:)

7

If you are not going to acquire software to do the job for you, your suggested work plan sounds good, namely writing a Natural program to read the Adabas file and write a work file.

Read an Adabas file record

Use a work file record as your output.

Format the workfile record appropriately so you can import data into your SQL package.

HOWEVER, the various software packages would cut out a lot of work on your part.

steve

1 Like
8

The other option would be to open the world’s best database (Adabas) for new applications rather than replacing both the database and existing applications completely.

There are a number of options for doing this, Software AG offers

  • Adabas SOA Gateway - for SOAP / REST access to Adabas (including MUs and PEs)
  • Adabas SQL Gateway - for SQL access to Adabas
1 Like