postgresql/src/test/regress/sql/create_am.sql

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

358 lines
14 KiB
MySQL
Raw Normal View History

--
-- Create access method tests
--
-- Make gist2 over gisthandler. In fact, it would be a synonym to gist.
CREATE ACCESS METHOD gist2 TYPE INDEX HANDLER gisthandler;
-- Verify return type checks for handlers
CREATE ACCESS METHOD bogus TYPE INDEX HANDLER int4in;
CREATE ACCESS METHOD bogus TYPE INDEX HANDLER heap_tableam_handler;
-- Try to create gist2 index on fast_emp4000: fail because opclass doesn't exist
CREATE INDEX grect2ind2 ON fast_emp4000 USING gist2 (home_base);
-- Make operator class for boxes using gist2
CREATE OPERATOR CLASS box_ops DEFAULT
FOR TYPE box USING gist2 AS
OPERATOR 1 <<,
OPERATOR 2 &<,
OPERATOR 3 &&,
OPERATOR 4 &>,
OPERATOR 5 >>,
OPERATOR 6 ~=,
OPERATOR 7 @>,
OPERATOR 8 <@,
OPERATOR 9 &<|,
OPERATOR 10 <<|,
OPERATOR 11 |>>,
OPERATOR 12 |&>,
FUNCTION 1 gist_box_consistent(internal, box, smallint, oid, internal),
FUNCTION 2 gist_box_union(internal, internal),
-- don't need compress, decompress, or fetch functions
FUNCTION 5 gist_box_penalty(internal, internal, internal),
FUNCTION 6 gist_box_picksplit(internal, internal),
FUNCTION 7 gist_box_same(box, box, internal);
-- Create gist2 index on fast_emp4000
CREATE INDEX grect2ind2 ON fast_emp4000 USING gist2 (home_base);
-- Now check the results from plain indexscan; temporarily drop existing
-- index grect2ind to ensure it doesn't capture the plan
BEGIN;
DROP INDEX grect2ind;
SET enable_seqscan = OFF;
SET enable_indexscan = ON;
SET enable_bitmapscan = OFF;
EXPLAIN (COSTS OFF)
SELECT * FROM fast_emp4000
WHERE home_base <@ '(200,200),(2000,1000)'::box
ORDER BY (home_base[0])[0];
SELECT * FROM fast_emp4000
WHERE home_base <@ '(200,200),(2000,1000)'::box
ORDER BY (home_base[0])[0];
EXPLAIN (COSTS OFF)
SELECT count(*) FROM fast_emp4000 WHERE home_base && '(1000,1000,0,0)'::box;
SELECT count(*) FROM fast_emp4000 WHERE home_base && '(1000,1000,0,0)'::box;
EXPLAIN (COSTS OFF)
SELECT count(*) FROM fast_emp4000 WHERE home_base IS NULL;
SELECT count(*) FROM fast_emp4000 WHERE home_base IS NULL;
ROLLBACK;
-- Try to drop access method: fail because of dependent objects
DROP ACCESS METHOD gist2;
-- Drop access method cascade
-- To prevent a (rare) deadlock against autovacuum,
-- we must lock the table that owns the index that will be dropped
BEGIN;
LOCK TABLE fast_emp4000;
DROP ACCESS METHOD gist2 CASCADE;
COMMIT;
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
--
-- Test table access methods
--
-- prevent empty values
SET default_table_access_method = '';
-- prevent nonexistent values
SET default_table_access_method = 'I do not exist AM';
-- prevent setting it to an index AM
SET default_table_access_method = 'btree';
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
-- Create a heap2 table am handler with heapam handler
CREATE ACCESS METHOD heap2 TYPE TABLE HANDLER heap_tableam_handler;
-- Verify return type checks for handlers
CREATE ACCESS METHOD bogus TYPE TABLE HANDLER int4in;
CREATE ACCESS METHOD bogus TYPE TABLE HANDLER bthandler;
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
SELECT amname, amhandler, amtype FROM pg_am where amtype = 't' ORDER BY 1, 2;
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
-- First create tables employing the new AM using USING
-- plain CREATE TABLE
CREATE TABLE tableam_tbl_heap2(f1 int) USING heap2;
INSERT INTO tableam_tbl_heap2 VALUES(1);
SELECT f1 FROM tableam_tbl_heap2 ORDER BY f1;
-- CREATE TABLE AS
CREATE TABLE tableam_tblas_heap2 USING heap2 AS SELECT * FROM tableam_tbl_heap2;
SELECT f1 FROM tableam_tbl_heap2 ORDER BY f1;
-- SELECT INTO doesn't support USING
SELECT INTO tableam_tblselectinto_heap2 USING heap2 FROM tableam_tbl_heap2;
-- CREATE VIEW doesn't support USING
CREATE VIEW tableam_view_heap2 USING heap2 AS SELECT * FROM tableam_tbl_heap2;
-- CREATE SEQUENCE doesn't support USING
CREATE SEQUENCE tableam_seq_heap2 USING heap2;
-- CREATE MATERIALIZED VIEW does support USING
CREATE MATERIALIZED VIEW tableam_tblmv_heap2 USING heap2 AS SELECT * FROM tableam_tbl_heap2;
SELECT f1 FROM tableam_tblmv_heap2 ORDER BY f1;
-- CREATE TABLE .. PARTITION BY supports USING.
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
CREATE TABLE tableam_parted_heap2 (a text, b int) PARTITION BY list (a) USING heap2;
SELECT a.amname FROM pg_class c, pg_am a
WHERE c.relname = 'tableam_parted_heap2' AND a.oid = c.relam;
DROP TABLE tableam_parted_heap2;
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
CREATE TABLE tableam_parted_heap2 (a text, b int) PARTITION BY list (a);
-- new partitions will inherit from the current default, rather the partition root
SET default_table_access_method = 'heap';
CREATE TABLE tableam_parted_a_heap2 PARTITION OF tableam_parted_heap2 FOR VALUES IN ('a');
SET default_table_access_method = 'heap2';
CREATE TABLE tableam_parted_b_heap2 PARTITION OF tableam_parted_heap2 FOR VALUES IN ('b');
RESET default_table_access_method;
-- but the method can be explicitly specified
CREATE TABLE tableam_parted_c_heap2 PARTITION OF tableam_parted_heap2 FOR VALUES IN ('c') USING heap;
CREATE TABLE tableam_parted_d_heap2 PARTITION OF tableam_parted_heap2 FOR VALUES IN ('d') USING heap2;
-- List all objects in AM
SELECT
pc.relkind,
pa.amname,
CASE WHEN relkind = 't' THEN
(SELECT 'toast for ' || relname::regclass FROM pg_class pcm WHERE pcm.reltoastrelid = pc.oid)
ELSE
relname::regclass::text
END COLLATE "C" AS relname
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
FROM pg_class AS pc,
pg_am AS pa
WHERE pa.oid = pc.relam
AND pa.amname = 'heap2'
ORDER BY 3, 1, 2;
-- Show dependencies onto AM - there shouldn't be any for toast
SELECT pg_describe_object(classid,objid,objsubid) AS obj
FROM pg_depend, pg_am
WHERE pg_depend.refclassid = 'pg_am'::regclass
AND pg_am.oid = pg_depend.refobjid
AND pg_am.amname = 'heap2'
ORDER BY classid, objid, objsubid;
-- ALTER TABLE SET ACCESS METHOD
CREATE TABLE heaptable USING heap AS
SELECT a, repeat(a::text, 100) FROM generate_series(1,9) AS a;
SELECT amname FROM pg_class c, pg_am am
WHERE c.relam = am.oid AND c.oid = 'heaptable'::regclass;
-- Switching to heap2 adds new dependency entry to the AM.
ALTER TABLE heaptable SET ACCESS METHOD heap2;
SELECT pg_describe_object(classid, objid, objsubid) as obj,
pg_describe_object(refclassid, refobjid, refobjsubid) as objref,
deptype
FROM pg_depend
WHERE classid = 'pg_class'::regclass AND
objid = 'heaptable'::regclass
ORDER BY 1, 2;
-- Switching to heap should not have a dependency entry to the AM.
ALTER TABLE heaptable SET ACCESS METHOD heap;
SELECT pg_describe_object(classid, objid, objsubid) as obj,
pg_describe_object(refclassid, refobjid, refobjsubid) as objref,
deptype
FROM pg_depend
WHERE classid = 'pg_class'::regclass AND
objid = 'heaptable'::regclass
ORDER BY 1, 2;
ALTER TABLE heaptable SET ACCESS METHOD heap2;
SELECT amname FROM pg_class c, pg_am am
WHERE c.relam = am.oid AND c.oid = 'heaptable'::regclass;
SELECT COUNT(a), COUNT(1) FILTER(WHERE a=1) FROM heaptable;
-- DEFAULT access method
BEGIN;
SET LOCAL default_table_access_method TO heap2;
ALTER TABLE heaptable SET ACCESS METHOD DEFAULT;
SELECT amname FROM pg_class c, pg_am am
WHERE c.relam = am.oid AND c.oid = 'heaptable'::regclass;
SET LOCAL default_table_access_method TO heap;
ALTER TABLE heaptable SET ACCESS METHOD DEFAULT;
SELECT amname FROM pg_class c, pg_am am
WHERE c.relam = am.oid AND c.oid = 'heaptable'::regclass;
ROLLBACK;
-- ALTER MATERIALIZED VIEW SET ACCESS METHOD
CREATE MATERIALIZED VIEW heapmv USING heap AS SELECT * FROM heaptable;
SELECT amname FROM pg_class c, pg_am am
WHERE c.relam = am.oid AND c.oid = 'heapmv'::regclass;
ALTER MATERIALIZED VIEW heapmv SET ACCESS METHOD heap2;
SELECT amname FROM pg_class c, pg_am am
WHERE c.relam = am.oid AND c.oid = 'heapmv'::regclass;
SELECT COUNT(a), COUNT(1) FILTER(WHERE a=1) FROM heapmv;
-- No support for multiple subcommands
ALTER TABLE heaptable SET ACCESS METHOD heap, SET ACCESS METHOD heap2;
ALTER TABLE heaptable SET ACCESS METHOD DEFAULT, SET ACCESS METHOD heap2;
ALTER MATERIALIZED VIEW heapmv SET ACCESS METHOD heap, SET ACCESS METHOD heap2;
DROP MATERIALIZED VIEW heapmv;
DROP TABLE heaptable;
-- Partition hierarchies with access methods
BEGIN;
SET LOCAL default_table_access_method = 'heap';
CREATE TABLE am_partitioned(x INT, y INT) PARTITION BY hash (x);
-- pg_class.relam is 0, no dependency recorded between the AM and the
-- partitioned table.
SELECT relam FROM pg_class WHERE relname = 'am_partitioned';
SELECT pg_describe_object(classid, objid, objsubid) AS obj,
pg_describe_object(refclassid, refobjid, refobjsubid) as refobj
FROM pg_depend, pg_am
WHERE pg_depend.refclassid = 'pg_am'::regclass
AND pg_am.oid = pg_depend.refobjid
AND pg_depend.objid = 'am_partitioned'::regclass;
-- New default is set, with dependency added.
ALTER TABLE am_partitioned SET ACCESS METHOD heap2;
SELECT a.amname FROM pg_class c, pg_am a
WHERE c.relname = 'am_partitioned' AND a.oid = c.relam;
SELECT pg_describe_object(classid, objid, objsubid) AS obj,
pg_describe_object(refclassid, refobjid, refobjsubid) as refobj
FROM pg_depend, pg_am
WHERE pg_depend.refclassid = 'pg_am'::regclass
AND pg_am.oid = pg_depend.refobjid
AND pg_depend.objid = 'am_partitioned'::regclass;
-- Default is set, with dependency updated.
SET LOCAL default_table_access_method = 'heap2';
ALTER TABLE am_partitioned SET ACCESS METHOD heap;
SELECT a.amname FROM pg_class c, pg_am a
WHERE c.relname = 'am_partitioned' AND a.oid = c.relam;
-- Dependency pinned, hence removed.
SELECT pg_describe_object(classid, objid, objsubid) AS obj,
pg_describe_object(refclassid, refobjid, refobjsubid) as refobj
FROM pg_depend, pg_am
WHERE pg_depend.refclassid = 'pg_am'::regclass
AND pg_am.oid = pg_depend.refobjid
AND pg_depend.objid = 'am_partitioned'::regclass;
-- Default and AM set in the clause are the same, relam should be set.
SET LOCAL default_table_access_method = 'heap2';
ALTER TABLE am_partitioned SET ACCESS METHOD heap2;
SELECT a.amname FROM pg_class c, pg_am a
WHERE c.relname = 'am_partitioned' AND a.oid = c.relam;
-- Reset to default
ALTER TABLE am_partitioned SET ACCESS METHOD DEFAULT;
SELECT relam FROM pg_class WHERE relname = 'am_partitioned';
-- Upon ALTER TABLE SET ACCESS METHOD on a partitioned table, new partitions
-- will inherit the AM set. Existing partitioned are unchanged.
SELECT relam FROM pg_class WHERE relname = 'am_partitioned';
SET LOCAL default_table_access_method = 'heap';
CREATE TABLE am_partitioned_0 PARTITION OF am_partitioned
FOR VALUES WITH (MODULUS 10, REMAINDER 0);
SET LOCAL default_table_access_method = 'heap2';
CREATE TABLE am_partitioned_1 PARTITION OF am_partitioned
FOR VALUES WITH (MODULUS 10, REMAINDER 1);
SET LOCAL default_table_access_method = 'heap';
ALTER TABLE am_partitioned SET ACCESS METHOD heap2;
CREATE TABLE am_partitioned_2 PARTITION OF am_partitioned
FOR VALUES WITH (MODULUS 10, REMAINDER 2);
ALTER TABLE am_partitioned SET ACCESS METHOD DEFAULT;
SELECT relam FROM pg_class WHERE relname = 'am_partitioned';
CREATE TABLE am_partitioned_3 PARTITION OF am_partitioned
FOR VALUES WITH (MODULUS 10, REMAINDER 3);
-- Partitioned table with relam at 0
ALTER TABLE am_partitioned SET ACCESS METHOD DEFAULT;
CREATE TABLE am_partitioned_5p PARTITION OF am_partitioned
FOR VALUES WITH (MODULUS 10, REMAINDER 5) PARTITION BY hash(y);
-- Partitions of this partitioned table inherit default AM at creation
-- time.
CREATE TABLE am_partitioned_5p1 PARTITION OF am_partitioned_5p
FOR VALUES WITH (MODULUS 10, REMAINDER 1);
-- Partitioned table with relam set.
ALTER TABLE am_partitioned SET ACCESS METHOD heap2;
CREATE TABLE am_partitioned_6p PARTITION OF am_partitioned
FOR VALUES WITH (MODULUS 10, REMAINDER 6) PARTITION BY hash(y);
-- Partitions of this partitioned table inherit its AM.
CREATE TABLE am_partitioned_6p1 PARTITION OF am_partitioned_6p
FOR VALUES WITH (MODULUS 10, REMAINDER 1);
SELECT c.relname, a.amname FROM pg_class c, pg_am a
WHERE c.relam = a.oid AND
c.relname LIKE 'am_partitioned%'
UNION ALL
SELECT c.relname, 'default' FROM pg_class c
WHERE c.relam = 0
AND c.relname LIKE 'am_partitioned%' ORDER BY 1;
DROP TABLE am_partitioned;
COMMIT;
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
-- Second, create objects in the new AM by changing the default AM
BEGIN;
SET LOCAL default_table_access_method = 'heap2';
-- following tests should all respect the default AM
CREATE TABLE tableam_tbl_heapx(f1 int);
CREATE TABLE tableam_tblas_heapx AS SELECT * FROM tableam_tbl_heapx;
SELECT INTO tableam_tblselectinto_heapx FROM tableam_tbl_heapx;
CREATE MATERIALIZED VIEW tableam_tblmv_heapx USING heap2 AS SELECT * FROM tableam_tbl_heapx;
CREATE TABLE tableam_parted_heapx (a text, b int) PARTITION BY list (a);
CREATE TABLE tableam_parted_1_heapx PARTITION OF tableam_parted_heapx FOR VALUES IN ('a', 'b');
-- but an explicitly set AM overrides it
CREATE TABLE tableam_parted_2_heapx PARTITION OF tableam_parted_heapx FOR VALUES IN ('c', 'd') USING heap;
-- sequences, views and foreign servers shouldn't have an AM
CREATE VIEW tableam_view_heapx AS SELECT * FROM tableam_tbl_heapx;
CREATE SEQUENCE tableam_seq_heapx;
CREATE FOREIGN DATA WRAPPER fdw_heap2 VALIDATOR postgresql_fdw_validator;
CREATE SERVER fs_heap2 FOREIGN DATA WRAPPER fdw_heap2 ;
CREATE FOREIGN table tableam_fdw_heapx () SERVER fs_heap2;
-- Verify that new AM was used for tables, matviews, but not for sequences, views and fdws
SELECT
pc.relkind,
pa.amname,
CASE WHEN relkind = 't' THEN
(SELECT 'toast for ' || relname::regclass FROM pg_class pcm WHERE pcm.reltoastrelid = pc.oid)
ELSE
relname::regclass::text
END COLLATE "C" AS relname
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
FROM pg_class AS pc
LEFT JOIN pg_am AS pa ON (pa.oid = pc.relam)
WHERE pc.relname LIKE 'tableam_%_heapx'
ORDER BY 3, 1, 2;
-- don't want to keep those tables, nor the default
ROLLBACK;
-- Third, check that we can neither create a table using a nonexistent
-- AM, nor using an index AM
CREATE TABLE i_am_a_failure() USING "";
CREATE TABLE i_am_a_failure() USING i_do_not_exist_am;
CREATE TABLE i_am_a_failure() USING "I do not exist AM";
CREATE TABLE i_am_a_failure() USING "btree";
-- Other weird invalid cases that cause problems
CREATE FOREIGN TABLE fp PARTITION OF pg_am DEFAULT SERVER x;
tableam: introduce table AM infrastructure. This introduces the concept of table access methods, i.e. CREATE ACCESS METHOD ... TYPE TABLE and CREATE TABLE ... USING (storage-engine). No table access functionality is delegated to table AMs as of this commit, that'll be done in following commits. Subsequent commits will incrementally abstract table access functionality to be routed through table access methods. That change is too large to be reviewed & committed at once, so it'll be done incrementally. Docs will be updated at the end, as adding them incrementally would likely make them less coherent, and definitely is a lot more work, without a lot of benefit. Table access methods are specified similar to index access methods, i.e. pg_am.amhandler returns, as INTERNAL, a pointer to a struct with callbacks. In contrast to index AMs that struct needs to live as long as a backend, typically that's achieved by just returning a pointer to a constant struct. Psql's \d+ now displays a table's access method. That can be disabled with HIDE_TABLEAM=true, which is mainly useful so regression tests can be run against different AMs. It's quite possible that this behaviour still needs to be fine tuned. For now it's not allowed to set a table AM for a partitioned table, as we've not resolved how partitions would inherit that. Disallowing allows us to introduce, if we decide that's the way forward, such a behaviour without a compatibility break. Catversion bumped, to add the heap table AM and references to it. Author: Haribabu Kommi, Andres Freund, Alvaro Herrera, Dimitri Golgov and others Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql https://postgr.es/m/20190107235616.6lur25ph22u5u5av@alap3.anarazel.de https://postgr.es/m/20190304234700.w5tmhducs5wxgzls@alap3.anarazel.de
2019-03-06 18:54:38 +01:00
-- Drop table access method, which fails as objects depends on it
DROP ACCESS METHOD heap2;
-- we intentionally leave the objects created above alive, to verify pg_dump support