Horton

Re: adding partitioned tables to publications

amit <amitlangote09@gmail.com>

View thread
From: Amit Langote <amitlangote09@gmail.com>
To: Peter Eisentraut <peter.eisentraut@2ndquadrant.com>
Cc: Rafia Sabih <rafia.pghackers@gmail.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2020-03-26T14:23:30Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Fix RELCACHE_FORCE_RELEASE issue

  2. Allow publishing partition changes via ancestors

  3. Add logical replication support to replicate into partitioned tables

  4. Refactor code to look up local replication tuple

  5. Some refactoring of logical/worker.c

  6. Prepare to support non-tables in publications

  7. Support adding partitioned tables to publication

On Wed, Mar 25, 2020 at 9:29 PM Peter Eisentraut
<peter.eisentraut@2ndquadrant.com> wrote:
> On 2020-03-23 06:02, Amit Langote wrote:
> > Okay, added some tests.
> >
> > Attached updated patches.
>
> I have committed the worker.c refactoring patch.
>
> "Add subscription support to replicate into partitioned tables" still
> has lacking test coverage.  Your changes in relation.c are not exercised
> at all because the partitioned table branch in apply_handle_update() is
> never taken.  This is critical and tricky code, so I would look for
> significant testing.

While trying some tests around the code you  mentioned, I found what
looks like a bug, which looking into now.

> The code looks okay to me.  I would remove this code
>
> +       memset(entry->attrmap->attnums, -1,
> +              entry->attrmap->maplen * sizeof(AttrNumber));
>
> because the entries are explicitly filled right after anyway, and
> filling the bytes with -1 has an unclear effect.  There is also
> seemingly some fishiness in this code around whether attribute numbers
> are zero- or one-based.  Perhaps this could be documented briefly.
> Maybe I'm misunderstanding something.

Will check and fix as necessary.

--
Thank you,

Amit Langote
EnterpriseDB: http://www.enterprisedb.com