Arunprasad Rajkumar

@gokhangulbiz Thanks for your contribution. Could you please rebase against latest main?

@gokhangulbiz Is this PR on your radar still?

@dcupif As an workaround, we use [exclude-table-data], which avoids copying initial data and avoids errors during CDC. Once the migration is over, the unwanted table can be dropped. This might...

The same problem we have encountered while using pgcopydb for Postgres to Timescale migration as well. The source would have plain Postgres table, but the same table on target will...

@dimitri If I understand correctly, we should `TRUNCATE` if the table on the target is partitioned but not on the source. If the table remains partitioned on both the source...

I see `table missing in catalog error` due to quoting added by https://github.com/dimitri/pgcopydb/pull/595. i.e. the catalog doesn't escape the nspname, relname; but it has been escaped for wal2json by https://github.com/dimitri/pgcopydb/pull/595.

> Lots of changes in ld_transform and I'm not sure why... do we really have to change all the internal function signatures in the context of this bugfix? @dimitri We...

@dimitri > > I think the alternate approach can be accessing the catalog from ld_wal2json.c & ld_test_decoding.c, ignore the generated column while transforming to the data object. > > This...

@dimitri It would be useful if can review this. Thank you.

@dimitri, Did you find some time to review this? Thank you.