database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-06-11 20:28:21 +03:00
Code Activity

logical decoding: Fix handling of large old tuples with replica identity full.

Browse Source 
When decoding the old version of an UPDATE or DELETE change, and if that
tuple was bigger than MaxHeapTupleSize, we either Assert'ed out, or
failed in more subtle ways in non-assert builds.  Normally individual
tuples aren't bigger than MaxHeapTupleSize, with big datums toasted.
But that's not the case for the old version of a tuple for logical
decoding; the replica identity is logged as one piece. With the default
replica identity btree limits that to small tuples, but that's not the
case for FULL.

Change the tuple buffer infrastructure to separate allocate over-large
tuples, instead of always going through the slab cache.

This unfortunately requires changing the ReorderBufferTupleBuf
definition, we need to store the allocated size someplace. To avoid
requiring output plugins to recompile, don't store HeapTupleHeaderData
directly after HeapTupleData, but point to it via t_data; that leaves
rooms for the allocated size.  As there's no reason for an output plugin
to look at ReorderBufferTupleBuf->t_data.header, remove the field. It
was just a minor convenience having it directly accessible.

Reported-By: Adam Dratwiński
Discussion: CAKg6ypLd7773AOX4DiOGRwQk1TVOQKhNwjYiVjJnpq8Wo+i62Q@mail.gmail.com
This commit is contained in:
Andres Freund
2016-03-05 18:02:20 -08:00
parent 0bda14d54c
commit c8f621c43a
5 changed files with 223 additions and 64 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
contrib/test_decoding/sql/toast.sql
View File

@ -265,4 +265,41 @@ ALTER TABLE toasted_copy ALTER COLUMN data SET STORAGE EXTERNAL;
203 untoasted200
\.
SELECT substr(data, 1, 200) FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL, 'include-xids', '0', 'skip-empty-xacts', '1');
-- test we can decode "old" tuples bigger than the max heap tuple size correctly
DROP TABLE IF EXISTS toasted_several;
CREATE TABLE toasted_several (
id serial unique not null,
toasted_key text primary key,
toasted_col1 text,
toasted_col2 text
);
ALTER TABLE toasted_several REPLICA IDENTITY FULL;
ALTER TABLE toasted_several ALTER COLUMN toasted_key SET STORAGE EXTERNAL;
ALTER TABLE toasted_several ALTER COLUMN toasted_col1 SET STORAGE EXTERNAL;
ALTER TABLE toasted_several ALTER COLUMN toasted_col2 SET STORAGE EXTERNAL;
INSERT INTO toasted_several(toasted_key) VALUES(repeat('9876543210', 2000));
SELECT regexp_replace(data, '^(.{100}).*(.{100})$', '\1..\2') FROM pg_logical_slot_peek_changes('regression_slot', NULL, NULL, 'include-xids', '0', 'skip-empty-xacts', '1');
-- test update of a toasted key without changing it
UPDATE toasted_several SET toasted_col1 = toasted_key;
UPDATE toasted_several SET toasted_col2 = toasted_col1;
SELECT regexp_replace(data, '^(.{100}).*(.{100})$', '\1..\2') FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL, 'include-xids', '0', 'skip-empty-xacts', '1');
/*
* update with large tuplebuf, in a transaction large enough to force to spool to disk
*/
BEGIN;
INSERT INTO toasted_several(toasted_key) SELECT * FROM generate_series(1, 10234);
UPDATE toasted_several SET toasted_col1 = toasted_col2 WHERE id = 1;
DELETE FROM toasted_several WHERE id = 1;
COMMIT;
DROP TABLE toasted_several;
SELECT regexp_replace(data, '^(.{100}).*(.{100})$', '\1..\2') FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL, 'include-xids', '0', 'skip-empty-xacts', '1')
WHERE data NOT LIKE '%INSERT: %';
SELECT pg_drop_replication_slot('regression_slot');