Regarding publish_via_partiton_root with pglogical

Hi

Respected Team

I know the use case of implementing the partitions with publication and
subscription of built-in logical replication
CREATE PUBLICATION dbz_publication FOR TABLE betplacement.bet WITH
(publish_via_partition_root = true); This will use parent table to replica
data changes to target from source
But
Could you please provide a logic to implement the same with pglogical ?
Do i need to add this publish_via_partition_root to
pglogical.replication_set_add_table functionality ?
you response is valuable in this aspect

Regards,
Durga Mahesh

Hi Durgamahesh

In pglogical extension, there is no direct equivalent to the
publish_via_partition_root option found in the built-in logical replication

you can achieve similar results when a new partition is added; you must add
it to the replication set using pglogical.replication_set_add_table.
Similarly existing partitions are reorganized or split, ensure to update
the replication sets accordingly.

Thanks & Regards

*Muhammad Affan (*아판*)*

*PostgreSQL Technical Support Engineer** / Pakistan R&D*

Interlace Plaza 4th floor Twinhub office 32 I8 Markaz, Islamabad, Pakistan

On Sat, Jul 20, 2024 at 12:00 PM Durgamahesh Manne <
maheshpostgres9(at)gmail(dot)com> wrote:

> Hi
>
> Respected Team
>
> I know the use case of implementing the partitions with publication and
> subscription of built-in logical replication
> CREATE PUBLICATION dbz_publication FOR TABLE betplacement.bet WITH
> (publish_via_partition_root = true); This will use parent table to replica
> data changes to target from source
> But
> Could you please provide a logic to implement the same with pglogical ?
> Do i need to add this publish_via_partition_root to
> pglogical.replication_set_add_table functionality ?
> you response is valuable in this aspect
>
> Regards,
> Durga Mahesh
>

Hi

Any inputs from the PGDG team members regarding this ?

Regards
Durga Mahesh

On Mon, Jul 22, 2024 at 3:35 PM khan Affan <bawag773(at)gmail(dot)com> wrote:

> Hi Durgamahesh
>
> In pglogical extension, there is no direct equivalent to the
> publish_via_partition_root option found in the built-in logical replication
>
> you can achieve similar results when a new partition is added; you must
> add it to the replication set using pglogical.replication_set_add_table.
> Similarly existing partitions are reorganized or split, ensure to update
> the replication sets accordingly.
>
> Thanks & Regards
>
>
> *Muhammad Affan (*아판*)*
>
> *PostgreSQL Technical Support Engineer** / Pakistan R&D*
>
> Interlace Plaza 4th floor Twinhub office 32 I8 Markaz, Islamabad, Pakistan
>
> On Sat, Jul 20, 2024 at 12:00 PM Durgamahesh Manne <
> maheshpostgres9(at)gmail(dot)com> wrote:
>
>> Hi
>>
>> Respected Team
>>
>> I know the use case of implementing the partitions with publication and
>> subscription of built-in logical replication
>> CREATE PUBLICATION dbz_publication FOR TABLE betplacement.bet WITH
>> (publish_via_partition_root = true); This will use parent table to replica
>> data changes to target from source
>> But
>> Could you please provide a logic to implement the same with pglogical ?
>> Do i need to add this publish_via_partition_root to
>> pglogical.replication_set_add_table functionality ?
>> you response is valuable in this aspect
>>
>> Regards,
>> Durga Mahesh
>>
>

Please do not spam the mailing lists with requests for follow ups. In this
particular case, you received an answer two days after you posted it. It
went to both psql-general, pgsql-in-general, and to you directly, so there
seems little excuse for missing it.

Also note that pglogical is a third-party extension and as such you may
have more success in the future asking the maintainers of it via its github
page (as I see you have done, although in a somewhat impolite way:
https://github.com/2ndQuadrant/pglogical/issues/487)

Cheers,
Greg

Hey Greg

Thank you

Regards,
Durga Mahesh

On Sat, 28 Sept, 2024, 20:31 Greg Sabino Mullane, <htamfids(at)gmail(dot)com>
wrote:

> Please do not spam the mailing lists with requests for follow ups. In this
> particular case, you received an answer two days after you posted it. It
> went to both psql-general, pgsql-in-general, and to you directly, so there
> seems little excuse for missing it.
>
> Also note that pglogical is a third-party extension and as such you may
> have more success in the future asking the maintainers of it via its github
> page (as I see you have done, although in a somewhat impolite way:
> https://github.com/2ndQuadrant/pglogical/issues/487)
>
> Cheers,
> Greg
>
>