On 5/14/25 23:09, veem v wrote:
> Hi,
> Its postgres database behind the scenes.
>
> We have a use case in which the customer is planning to migrate data
> from an older version (V1) to a newer version (V2). For V2, the tables
> will be new, but their structure will be similar to the V1 version with
> few changes in relationship might be there. We want to have this
> migration approach happen in multiple phases in which each time the
> delta data from version V1 will be moved to version- V2 and then final
> cutover will happen to V2 if all looks good or else rollback to V1. The
> tables are smaller in size like max ~100K records in tables.
>
> My question is, is it a good idea to have an approach in which we will
> have procedures created to move the delta data in every phase and
> schedule those using some tasks for each table. Or any other strategy
> should we follow?
This is what Sqitch(https://sqitch.org/) was designed for.
The biggest issue is that the data will be incrementing while you do the
structural changes. How you handle that is going to depend on the
question raised by Peter J. Holzer:
Is this being done in place on one Postgres instance or between
separate Postgres instances?
>
> Also another thing to note , we have used sequences as primary keys in
> some tables and they have FK relationships with other tables, so the
> same sequence number in version V2 will cause issues/conflict, so how
> should we handle this scenario? Should we just create new sequences with
> higher start values?
>
> Regards
> Veem
--
Adrian Klaver
adrian.klaver@aklaver.com
