Horton

Re: row filtering for logical replication

Peter Smith <smithpb2250@gmail.com>

View thread
From: Peter Smith <smithpb2250@gmail.com>
To: Amit Kapila <amit.kapila16@gmail.com>
Cc: "houzj.fnst@fujitsu.com" <houzj.fnst@fujitsu.com>, Ajin Cherian <itsajin@gmail.com>, Euler Taveira <euler@eulerto.com>, Greg Nancarrow <gregn4422@gmail.com>, vignesh C <vignesh21@gmail.com>, "tanghy.fnst@fujitsu.com" <tanghy.fnst@fujitsu.com>, Dilip Kumar <dilipbalaut@gmail.com>, Rahila Syed <rahilasyed90@gmail.com>, Peter Eisentraut <peter.eisentraut@enterprisedb.com>, Önder Kalacı <onderkalaci@gmail.com>, japin <japinli@hotmail.com>, Michael Paquier <michael@paquier.xyz>, David Steele <david@pgmasters.net>, Craig Ringer <craig@2ndquadrant.com>, Amit Langote <amitlangote09@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2021-12-21T23:56:31Z
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. Release cache tuple when no longer needed

  2. Add some additional tests for row filters in logical replication.

  3. Fix one of the tests introduced in commit 52e4f0cd47.

  4. Allow specifying row filters for logical replication of tables.

  5. Move scanint8() to numutils.c

  6. Replace Test::More plans with done_testing

  7. Reduce relcache access in WAL sender streaming logical changes

  8. Small cleanups related to PUBLICATION framework code

  9. Add a view to show the stats of subscription workers.

  10. Allow publishing the tables of schema.

  11. Doc: improve documentation of CREATE/ALTER SUBSCRIPTION.

  12. Add PublicationTable and PublicationRelInfo structs

  13. Remove unused argument "txn" in maybe_send_schema().

  14. Add prepare API support for streaming transactions in logical replication.

  15. Unify PostgresNode's new() and get_new_node() methods

  16. Use l*_node() family of functions where appropriate

  17. Add support for prepared transactions to built-in logical replication.

  18. Restore the portal-level snapshot after procedure COMMIT/ROLLBACK.

  19. Rename a parse node to be more general

  20. Remove unused column atttypmod from initial tablesync query

  21. SEARCH and CYCLE clauses

On Mon, Dec 20, 2021 at 9:30 PM Amit Kapila <amit.kapila16@gmail.com> wrote:
>
> On Mon, Dec 20, 2021 at 8:41 AM houzj.fnst@fujitsu.com
> <houzj.fnst@fujitsu.com> wrote:
> >
> > Thanks for the comments, I agree with all the comments.
> > Attach the V49 patch set, which addressed all the above comments on the 0002
> > patch.
> >
>
> Few comments/suugestions:
> ======================
> 1.
> + Oid publish_as_relid = InvalidOid;
> +
> + /*
> + * For a partition, if pubviaroot is true, check if any of the
> + * ancestors are published. If so, note down the topmost ancestor
> + * that is published via this publication, the row filter
> + * expression on which will be used to filter the partition's
> + * changes. We could have got the topmost ancestor when collecting
> + * the publication oids, but that will make the code more
> + * complicated.
> + */
> + if (pubform->pubviaroot && relation->rd_rel->relispartition)
> + {
> + if (pubform->puballtables)
> + publish_as_relid = llast_oid(ancestors);
> + else
> + publish_as_relid = GetTopMostAncestorInPublication(pubform->oid,
> +    ancestors);
> + }
> +
> + if (publish_as_relid == InvalidOid)
> + publish_as_relid = relid;
>
> I think you can initialize publish_as_relid as relid and then later
> override it if required. That will save the additional check of
> publish_as_relid.
>

Fixed in v51* [1]

> 2. I think your previous version code in GetRelationPublicationActions
> was better as now we have to call memcpy at two places.
>

Fixed in v51* [1]

> 3.
> +
> + if (list_member_oid(GetRelationPublications(ancestor),
> + puboid) ||
> + list_member_oid(GetSchemaPublications(get_rel_namespace(ancestor)),
> + puboid))
> + {
> + topmost_relid = ancestor;
> + }
>
> I think here we don't need to use braces ({}) as there is just a
> single statement in the condition.
>

Fixed in v51* [1]

> 4.
> +#define IDX_PUBACTION_n 3
> + ExprState    *exprstate[IDX_PUBACTION_n]; /* ExprState array for row filter.
> +    One per publication action. */
> ..
> ..
>
> I think we can have this define outside the structure. I don't like
> this define name, can we name it NUM_ROWFILTER_TYPES or something like
> that?
>

Partly fixed in v51* [1], I've changed the #define name but I did not
move it. The adjacent comment talks about these ExprState caches and
explains the reason why the number is 3. So if I move the #define then
half that comment would have to move with it. I thought it is better
to keep all the related parts grouped together with the one
explanatory comment, but if you still want the #define moved please
confirm and I can do it in a future version.

> I think we can now merge 0001, 0002, and 0005. We are still evaluating
> the performance for 0003, so it is better to keep it separate. We can
> take the decision to merge it once we are done with our evaluation.
>

Merged as suggested in v51* [1]

------
[1] https://www.postgresql.org/message-id/CAHut%2BPs%2BdACvefCZasRE%3DP%3DDtaNmQvM3kiGyKyBHANA0yGcTZw%40mail.gmail.com

Kind Regards,
Peter Smith.
Fujitsu Australia