Horton
Menu

RE: Perform streaming logical transactions by background workers and parallel apply

Zhijie Hou (Fujitsu) <houzj.fnst@fujitsu.com>

View thread
From: "houzj.fnst@fujitsu.com" <houzj.fnst@fujitsu.com>
To: "kuroda.hayato@fujitsu.com" <kuroda.hayato@fujitsu.com>, "wangw.fnst@fujitsu.com" <wangw.fnst@fujitsu.com>
Cc: Amit Kapila <amit.kapila16@gmail.com>, Dilip Kumar <dilipbalaut@gmail.com>, Masahiko Sawada <sawada.mshk@gmail.com>, "shiy.fnst@fujitsu.com" <shiy.fnst@fujitsu.com>, Peter Smith <smithpb2250@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2022-08-29T11:31:49Z
Lists: pgsql-hackers 
On Tues, Aug 24, 2022 16:41 PM Kuroda, Hayato/黒田 隼人 <kuroda.hayato@fujitsu.com> wrote:
> Dear Wang,
> 
> Followings are my comments about v23-0003. Currently I do not have any 
> comments about 0002 and 0004.

Thanks for your comments.

> 09. general
> 
> It seems that logicalrep_rel_mark_parallel_apply() is always called 
> when relations are opened on the subscriber-side, but is it really 
> needed? There checks are required only for streaming parallel apply, 
> so it may be not needed in case of streaming = 'on' or 'off'.

Improved.
This check is only performed when using apply background workers.

> 10. commit message
> 
>     2) There cannot be any non-immutable functions used by the subscriber-side
>     replicated table. Look for functions in the following places:
>     * a. Trigger functions
>     * b. Column default value expressions and domain constraints
>     * c. Constraint expressions
>     * d. Foreign keys
> 
> "Foreign key" should not be listed here because it is not related with 
> the mutability. I think it should be listed as 3), not d..

Improved.

> 11. create_subscription.sgml
> 
> The constraint about foreign key should be described here.
> 
> 11. relation.c
> 
> 11.a
> 
> +       CacheRegisterSyscacheCallback(PROCOID,
> +                                                                 logicalrep_relmap_reset_parallel_cb,
> +                                                                 
> + (Datum) 0);
> 
> Isn't it needed another syscache callback for pg_type?
> Users can add any constraints via ALTER DOMAIN command, but the added 
> constraint may be not checked.
> I checked AlterDomainAddConstraint(), and it invalidates only the 
> relcache for pg_type.
> 
> 11.b
> 
> +               /*
> +                * If the column is of a DOMAIN type, determine whether
> +                * that domain has any CHECK expressions that are not
> +                * immutable.
> +                */
> +               if (get_typtype(att->atttypid) == TYPTYPE_DOMAIN)
> +               {
> 
> I think the default value of *domain* must be also checked here.
> I tested like followings.
> 
> ===
> 1. created a domain that has a default value CREATE DOMAIN tmp INT 
> DEFAULT 1 CHECK (VALUE > 0);
> 
> 2. created a table
> CREATE TABLE foo (id tmp PRIMARY KEY);
> 
> 3. checked pg_attribute and pg_class
> select oid, relname, attname, atthasdef from pg_attribute, pg_class 
> where pg_attribute.attrelid = pg_class.oid and pg_class.relname = 
> 'foo' and attname = 'id';
>   oid  | relname | attname | atthasdef
> -------+---------+---------+-----------
>  16394 | foo     | id      | f
> (1 row)
> 
> Tt meant that functions might be not checked because the if-statement 
> `if (att-
> >atthasdef)` became false.
> ===

Fixed.
In addition, to reduce duplicate validation, only the flag "parallel_apply_safe" is reset when pg_proc and pg_type changes.

> 12. 015_stream.pl, 016_stream_subxact.pl, 022_twophase_cascade.pl, 
> 023_twophase_stream.pl
> 
> -       my ($node_publisher, $node_subscriber, $appname, $is_parallel) = @_;
> +       my ($node_publisher, $node_subscriber, $appname) = @_;
> 
> Why the parameter is removed? I think the test that waits the output 
> from the apply background worker is meaningful.

Revert this change.
In addition, made some modifications to the logs confirmed in these test files to
ensure the streamed transactions complete as expected using apply background worker.

> 13. 032_streaming_apply.pl
> 
> The filename seems too general because apply background workers are 
> tested in above tests.
> How about "streaming_apply_constraint" or something?

Renamed to 032_streaming_parallel_safety.

Best regards,
Hou zj

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 invalid memory access during the shutdown of the parallel apply worker.

  2. Fix assertion failure in apply worker.

  3. Use elog to report unexpected action in handle_streamed_transaction().

  4. Use appropriate wait event when sending data in the apply worker.

  5. Allow the logical_replication_mode to be used on the subscriber.

  6. Rename GUC logical_decoding_mode to logical_replication_mode.

  7. Display the leader apply worker's PID for parallel apply workers.

  8. Improve the code to decide and process the apply action.

  9. Document the newly added wait events added by commit 216a784829.

  10. Perform apply of large transactions by parallel workers.

  11. Wake up a subscription's replication worker processes after DDL.

  12. Add copyright notices to meson files

  13. Better document logical replication parameters

  14. Add a common function to generate the origin name.

  15. Harmonize parameter names in storage and AM code.

  16. Avoid using list_length() to test for empty list.

  17. Improve two comments related to a boolean DefElem's value

  18. Fix partition table's REPLICA IDENTITY checking on the subscriber.

  19. Fix data inconsistency between publisher and subscriber.

  20. Fix cache look-up failures while applying changes in logical replication.