postgres/timescaledb
1
0
Fork 0
mirror of https://github.com/timescale/timescaledb.git synced 2025-05-14 01:10:37 +08:00
Code Issues Packages Projects Releases Wiki Activity

Release 2.15.0

Browse Source 
This release contains performance improvements and bug fixes since
the 2.14.2 release. We recommend that you upgrade at the next
available opportunity.

In addition, it includes these noteworthy features:
* Support `time_bucket` with `origin` and/or `offset` on Continuous Aggregate
* Compression improvements:
  - Improve expression pushdown
  - Add minmax sparse indexes when compressing columns with btree indexes
  - Make compression use the defaults functions
  - Vectorize filters in WHERE clause that contain text equality operators and LIKE expressions

**Deprecation warning**
* Starting on this release will not be possible to create Continuous Aggregate using `time_bucket_ng` anymore and it will be completely removed on the upcoming releases.
* Recommend users to [migrate their old Continuous Aggregate format to the new one](https://docs.timescale.com/use-timescale/latest/continuous-aggregates/migrate/) because it support will be completely removed in next releases prevent them to migrate.
* This is the last release supporting PostgreSQL 13.

**For on-premise users and this release only**, you will need to run [this SQL script](https://github.com/timescale/timescaledb-extras/blob/master/utils/2.15.X-fix_hypertable_foreign_keys.sql) after running `ALTER EXTENSION`. More details can be found in the pull request [#6786](https://github.com/timescale/timescaledb/pull/6797).

**Features**
* #6382 Support for time_bucket with origin and offset in CAggs
* #6696 Improve defaults for compression segment_by and order_by
* #6705 Add sparse minmax indexes for compressed columns that have uncompressed btree indexes
* #6754 Allow DROP CONSTRAINT on compressed hypertables
* #6767 Add metadata table `_timestaledb_internal.bgw_job_stat_history` for tracking job execution history
* #6798 Prevent usage of deprecated time_bucket_ng in CAgg definition
* #6810 Add telemetry for access methods
* #6811 Remove no longer relevant timescaledb.allow_install_without_preload GUC
* #6837 Add migration path for CAggs using time_bucket_ng
* #6865 Update the watermark when truncating a CAgg

**Bugfixes**
* #6617 Fix error in show_chunks
* #6621 Remove metadata when dropping chunks
* #6677 Fix snapshot usage in CAgg invalidation scanner
* #6698 Define meaning of 0 retries for jobs as no retries
* #6717 Fix handling of compressed tables with primary or unique index in COPY path
* #6726 Fix constify cagg_watermark using window function when querying a CAgg
* #6729 Fix NULL start value handling in CAgg refresh
* #6732 Fix CAgg migration with custom timezone / date format settings
* #6752 Remove custom autovacuum setting from compressed chunks
* #6770 Fix plantime chunk exclusion for OSM chunk
* #6789 Fix deletes with subqueries and compression
* #6796 Fix a crash involving a view on a hypertable
* #6797 Fix foreign key constraint handling on compressed hypertables
* #6816 Fix handling of chunks with no contraints
* #6820 Fix a crash when the ts_hypertable_insert_blocker was called directly
* #6849 Use non-orderby compressed metadata in compressed DML
* #6867 Clean up compression settings when deleting compressed cagg
* #6869 Fix compressed DML with constraints of form value OP column
* #6870 Fix bool expression pushdown for queries on compressed chunks

**Thanks**
* @brasic for reporting a crash when the ts_hypertable_insert_blocker was called directly
* @bvanelli for reporting an issue with the jobs retry count
* @djzurawsk For reporting error when dropping chunks
* @Dzuzepppe for reporting an issue with DELETEs using subquery on compressed chunk working incorrectly.
* @hongquan For reporting a 'timestamp out of range' error during CAgg migrations
* @kevcenteno for reporting an issue with the show_chunks API showing incorrect output when 'created_before/created_after' was used with time-partitioned columns.
* @mahipv For starting working on the job history PR
* @rovo89 For reporting constify cagg_watermark not working using window function when querying a CAgg
This commit is contained in:
Fabrízio de Royes Mello 2024-05-06 11:52:12 -03:00
parent d8057c4326
commit defe4ef581
36 changed files with 519 additions and 495 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
.unreleased/RELEASE_NOTES_HEADER.md.j2
View File

@ -1,3 +1,9 @@
{#
# Variables:
# release_current = the release version that has being builded
# release_previous = previous release version
# release_date = date of the release in format YYYY-MM-DD
#}
## {{ release_current }} ({{ release_date }})
This release contains performance improvements and bug fixes since
@ -5,16 +11,6 @@ the {{ release_previous }} release. We recommend that you upgrade at the next
available opportunity.
In addition, it includes these noteworthy features:
* Support `time_bucket` with `origin` and/or `offset` on Continuous Aggregate
* Compression improvements:
- Improve expression pushdown
- Add minmax sparse indexes when compressing columns with btree indexes
- Make compression use the defaults functions
- Vectorize filters in WHERE clause that contain text equality operators and LIKE expressions
**Deprecation warning**
* Starting on this release will not be possible to create Continuous Aggregate using `time_bucket_ng` anymore and it will be completely removed on the upcoming releases.
* Recommend users to [migrate their old Continuous Aggregate format to the new one](https://docs.timescale.com/use-timescale/latest/continuous-aggregates/migrate/) because it support will be completely removed in next releases prevent them to migrate.
* This is the last release supporting PostgreSQL 13.
**For on-premise users and this release only**, you will need to run [this SQL script](https://github.com/timescale/timescaledb-extras/blob/master/utils/2.15.X-fix_hypertable_foreign_keys.sql) after running `ALTER EXTENSION`. More details can be found in the pull request [#6786](https://github.com/timescale/timescaledb/pull/6797).
* feature 1
* feature 2
* feature N

1
.unreleased/auto_sparse_indexes
View File

@ -1 +0,0 @@
Implements: #6705 Add sparse minmax indexes for compressed columns that have uncompressed btree indexes

1
.unreleased/bugfix_6677
View File

@ -1 +0,0 @@
Fixes: #6677 Fix snapshot usage in CAgg invalidation scanner

1
.unreleased/bugfix_6717
View File

@ -1 +0,0 @@
Fixes: #6717 Fix handling of compressed tables with primary or unique index in COPY path

1
.unreleased/bugfix_6729
View File

@ -1 +0,0 @@
Fixes: #6729 Fix NULL start value handling in CAgg refresh

2
.unreleased/bugfix_6732
View File

@ -1,2 +0,0 @@
Fixes: #6732 Fix CAgg migration with custom timezone / date format settings
Thanks: @hongquan For reporting a 'timestamp out of range' error during CAgg migrations

1
.unreleased/bugfix_6752
View File

@ -1 +0,0 @@
Fixes: #6752 Remove custom autovacuum setting from compressed chunks

2
.unreleased/bugfix_6789
View File

@ -1,2 +0,0 @@
Fixes: #6789 Fix deletes with subqueries and compression
Thanks: @Dzuzepppe for reporting an issue with DELETEs using subquery on compressed chunk working incorrectly.

1
.unreleased/bugfix_6796
View File

@ -1 +0,0 @@
Fixes: #6796 Fix a crash involving a view on a hypertable

1
.unreleased/feature_6382
View File

@ -1 +0,0 @@
Implements: #6382 Support for time_bucket with origin and offset in CAggs

1
.unreleased/feature_6837
View File

@ -1 +0,0 @@
Implements: #6837 Add migration path for CAggs using time_bucket_ng

1
.unreleased/feature_6865
View File

@ -1 +0,0 @@
Implements: #6865 Update the watermark when truncating a CAgg

2
.unreleased/fix_6617
View File

@ -1,2 +0,0 @@
Fixes: #6617 Fix error in show_chunks
Thanks: @kevcenteno for reporting an issue with the show_chunks API showing incorrect output when 'created_before/created_after' was used with time-partitioned columns.

1
.unreleased/fix_6816
View File

@ -1 +0,0 @@
Fixes: #6816 Fix handling of chunks with no contraints

2
.unreleased/fix_6820
View File

@ -1,2 +0,0 @@
Fixes: #6820 Fix a crash when the ts_hypertable_insert_blocker was called directly
Thanks: @brasic for reporting a crash when the ts_hypertable_insert_blocker was called directly

2
.unreleased/pr_6621
View File

@ -1,2 +0,0 @@
Fixes: #6621 Remove metadata when dropping chunks
Thanks: @djzurawsk For reporting error when dropping chunks

1
.unreleased/pr_6696
View File

@ -1 +0,0 @@
Implements: #6696 Improve defaults for compression segment_by and order_by

2
.unreleased/pr_6698
View File

@ -1,2 +0,0 @@
Fixes: #6698 Define meaning of 0 retries for jobs as no retries
Thanks: @bvanelli for reporting an issue with the jobs retry count

2
.unreleased/pr_6726
View File

@ -1,2 +0,0 @@
Fixes: #6726 Fix constify cagg_watermark using window function when querying a CAgg
Thanks: @rovo89 For reporting constify cagg_watermark not working using window function when querying a CAgg

1
.unreleased/pr_6754
View File

@ -1 +0,0 @@
Implements: #6754 Allow DROP CONSTRAINT on compressed hypertables

2
.unreleased/pr_6767
View File

@ -1,2 +0,0 @@
Implements: #6767 Add metadata table `_timestaledb_internal.bgw_job_stat_history` for tracking job execution history
Thanks: @mahipv For starting working on the job history PR

1
.unreleased/pr_6770
View File

@ -1 +0,0 @@
Fixes: #6770 Fix plantime chunk exclusion for OSM chunk

1
.unreleased/pr_6797
View File

@ -1 +0,0 @@
Fixes: #6797 Fix foreign key constraint handling on compressed hypertables

1
.unreleased/pr_6798
View File

@ -1 +0,0 @@
Implements: #6798 Prevent usage of deprecated time_bucket_ng in CAgg definition

1
.unreleased/pr_6810
View File

@ -1 +0,0 @@
Implements: #6810 Add telemetry for access methods

1
.unreleased/pr_6811
View File

@ -1 +0,0 @@
Implements: #6811 Remove no longer relevant timescaledb.allow_install_without_preload GUC

1
.unreleased/pr_6849
View File

@ -1 +0,0 @@
Fixes: #6849 Use non-orderby compressed metadata in compressed DML

1
.unreleased/pr_6867
View File

@ -1 +0,0 @@
Fixes: #6867 Clean up compression settings when deleting compressed cagg

1
.unreleased/pr_6869
View File

@ -1 +0,0 @@
Fixes: #6869 Fix compressed DML with constraints of form value OP column

1
.unreleased/pr_6870
View File

@ -1 +0,0 @@
Fixes: #6870 Fix bool expression pushdown for queries on compressed chunks

64
CHANGELOG.md
View File

@ -4,6 +4,70 @@
`psql` with the `-X` flag to prevent any `.psqlrc` commands from
accidentally triggering the load of a previous DB version.**
## 2.15.0 (2024-05-06)
This release contains performance improvements and bug fixes since
the 2.14.2 release. We recommend that you upgrade at the next
available opportunity.
In addition, it includes these noteworthy features:
* Support `time_bucket` with `origin` and/or `offset` on Continuous Aggregate
* Compression improvements:
- Improve expression pushdown
- Add minmax sparse indexes when compressing columns with btree indexes
- Improve compression setting defaults
- Vectorize filters in WHERE clause that contain text equality operators and LIKE expressions
**Deprecation warning**
* Starting on this release will not be possible to create Continuous Aggregate using `time_bucket_ng` anymore and it will be completely removed on the upcoming releases.
* Recommend users to [migrate their old Continuous Aggregate format to the new one](https://docs.timescale.com/use-timescale/latest/continuous-aggregates/migrate/) because it support will be completely removed in next releases prevent them to migrate.
* This is the last release supporting PostgreSQL 13.
**For on-premise users and this release only**, you will need to run [this SQL script](https://github.com/timescale/timescaledb-extras/blob/master/utils/2.15.X-fix_hypertable_foreign_keys.sql) after running `ALTER EXTENSION`. More details can be found in the pull request [#6786](https://github.com/timescale/timescaledb/pull/6797).
**Features**
* #6382 Support for time_bucket with origin and offset in CAggs
* #6696 Improve defaults for compression segment_by and order_by
* #6705 Add sparse minmax indexes for compressed columns that have uncompressed btree indexes
* #6754 Allow DROP CONSTRAINT on compressed hypertables
* #6767 Add metadata table `_timestaledb_internal.bgw_job_stat_history` for tracking job execution history
* #6798 Prevent usage of deprecated time_bucket_ng in CAgg definition
* #6810 Add telemetry for access methods
* #6811 Remove no longer relevant timescaledb.allow_install_without_preload GUC
* #6837 Add migration path for CAggs using time_bucket_ng
* #6865 Update the watermark when truncating a CAgg
**Bugfixes**
* #6617 Fix error in show_chunks
* #6621 Remove metadata when dropping chunks
* #6677 Fix snapshot usage in CAgg invalidation scanner
* #6698 Define meaning of 0 retries for jobs as no retries
* #6717 Fix handling of compressed tables with primary or unique index in COPY path
* #6726 Fix constify cagg_watermark using window function when querying a CAgg
* #6729 Fix NULL start value handling in CAgg refresh
* #6732 Fix CAgg migration with custom timezone / date format settings
* #6752 Remove custom autovacuum setting from compressed chunks
* #6770 Fix plantime chunk exclusion for OSM chunk
* #6789 Fix deletes with subqueries and compression
* #6796 Fix a crash involving a view on a hypertable
* #6797 Fix foreign key constraint handling on compressed hypertables
* #6816 Fix handling of chunks with no contraints
* #6820 Fix a crash when the ts_hypertable_insert_blocker was called directly
* #6849 Use non-orderby compressed metadata in compressed DML
* #6867 Clean up compression settings when deleting compressed cagg
* #6869 Fix compressed DML with constraints of form value OP column
* #6870 Fix bool expression pushdown for queries on compressed chunks
**Thanks**
* @brasic for reporting a crash when the ts_hypertable_insert_blocker was called directly
* @bvanelli for reporting an issue with the jobs retry count
* @djzurawsk For reporting error when dropping chunks
* @Dzuzepppe for reporting an issue with DELETEs using subquery on compressed chunk working incorrectly.
* @hongquan For reporting a 'timestamp out of range' error during CAgg migrations
* @kevcenteno for reporting an issue with the show_chunks API showing incorrect output when 'created_before/created_after' was used with time-partitioned columns.
* @mahipv For starting working on the job history PR
* @rovo89 For reporting constify cagg_watermark not working using window function when querying a CAgg
## 2.14.2 (2024-02-20)
This release contains bug fixes since the 2.14.1 release.

3
sql/CMakeLists.txt
View File

@ -45,7 +45,8 @@ set(MOD_FILES
updates/2.13.0--2.13.1.sql
updates/2.13.1--2.14.0.sql
updates/2.14.0--2.14.1.sql
updates/2.14.1--2.14.2.sql)
updates/2.14.1--2.14.2.sql
updates/2.14.2--2.15.0.sql)
# The downgrade file to generate a downgrade script for the current version, as
# specified in version.config

440
sql/updates/2.14.2--2.15.0.sql Normal file
View File

@ -0,0 +1,440 @@
-- Remove multi-node CAGG support
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_cagg_log_add_entry(integer,bigint,bigint);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_hyper_log_add_entry(integer,bigint,bigint);
DROP FUNCTION IF EXISTS _timescaledb_internal.materialization_invalidation_log_delete(integer);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_process_cagg_log(integer,integer,regtype,bigint,bigint,integer[],bigint[],bigint[]);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_process_cagg_log(integer,integer,regtype,bigint,bigint,integer[],bigint[],bigint[],text[]);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_process_hypertable_log(integer,integer,regtype,integer[],bigint[],bigint[]);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_process_hypertable_log(integer,integer,regtype,integer[],bigint[],bigint[],text[]);
DROP FUNCTION IF EXISTS _timescaledb_internal.hypertable_invalidation_log_delete(integer);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_cagg_log_add_entry(integer,bigint,bigint);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_hyper_log_add_entry(integer,bigint,bigint);
DROP FUNCTION IF EXISTS _timescaledb_functions.materialization_invalidation_log_delete(integer);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_process_cagg_log(integer,integer,regtype,bigint,bigint,integer[],bigint[],bigint[]);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_process_cagg_log(integer,integer,regtype,bigint,bigint,integer[],bigint[],bigint[],text[]);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_process_hypertable_log(integer,integer,regtype,integer[],bigint[],bigint[]);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_process_hypertable_log(integer,integer,regtype,integer[],bigint[],bigint[],text[]);
DROP FUNCTION IF EXISTS _timescaledb_functions.hypertable_invalidation_log_delete(integer);
-- Remove chunk metadata when marked as dropped
CREATE FUNCTION _timescaledb_functions.remove_dropped_chunk_metadata(_hypertable_id INTEGER)
RETURNS INTEGER LANGUAGE plpgsql AS $$
DECLARE
_chunk_id INTEGER;
_removed INTEGER := 0;
BEGIN
FOR _chunk_id IN
SELECT id FROM _timescaledb_catalog.chunk
WHERE hypertable_id = _hypertable_id
AND dropped IS TRUE
AND NOT EXISTS (
SELECT FROM information_schema.tables
WHERE tables.table_schema = chunk.schema_name
AND tables.table_name = chunk.table_name
)
AND NOT EXISTS (
SELECT FROM _timescaledb_catalog.hypertable
JOIN _timescaledb_catalog.continuous_agg ON continuous_agg.raw_hypertable_id = hypertable.id
WHERE hypertable.id = chunk.hypertable_id
-- for the old caggs format we need to keep chunk metadata for dropped chunks
AND continuous_agg.finalized IS FALSE
)
LOOP
_removed := _removed + 1;
RAISE INFO 'Removing metadata of chunk % from hypertable %', _chunk_id, _hypertable_id;
WITH _dimension_slice_remove AS (
DELETE FROM _timescaledb_catalog.dimension_slice
USING _timescaledb_catalog.chunk_constraint
WHERE dimension_slice.id = chunk_constraint.dimension_slice_id
AND chunk_constraint.chunk_id = _chunk_id
RETURNING _timescaledb_catalog.dimension_slice.id
)
DELETE FROM _timescaledb_catalog.chunk_constraint
USING _dimension_slice_remove
WHERE chunk_constraint.dimension_slice_id = _dimension_slice_remove.id;
DELETE FROM _timescaledb_internal.bgw_policy_chunk_stats
WHERE bgw_policy_chunk_stats.chunk_id = _chunk_id;
DELETE FROM _timescaledb_catalog.chunk_index
WHERE chunk_index.chunk_id = _chunk_id;
DELETE FROM _timescaledb_catalog.compression_chunk_size
WHERE compression_chunk_size.chunk_id = _chunk_id
OR compression_chunk_size.compressed_chunk_id = _chunk_id;
DELETE FROM _timescaledb_catalog.chunk
WHERE chunk.id = _chunk_id
OR chunk.compressed_chunk_id = _chunk_id;
END LOOP;
RETURN _removed;
END;
$$ SET search_path TO pg_catalog, pg_temp;
SELECT _timescaledb_functions.remove_dropped_chunk_metadata(id) AS chunks_metadata_removed
FROM _timescaledb_catalog.hypertable;
--
-- Rebuild the catalog table `_timescaledb_catalog.continuous_aggs_bucket_function`
--
CREATE OR REPLACE FUNCTION _timescaledb_functions.cagg_get_bucket_function(
mat_hypertable_id INTEGER
) RETURNS regprocedure AS '@MODULE_PATHNAME@', 'ts_continuous_agg_get_bucket_function' LANGUAGE C STRICT VOLATILE;
-- Since we need now the regclass of the used bucket function, we have to recover it
-- by parsing the view query by calling 'cagg_get_bucket_function'.
CREATE TABLE _timescaledb_catalog._tmp_continuous_aggs_bucket_function AS
SELECT
mat_hypertable_id,
_timescaledb_functions.cagg_get_bucket_function(mat_hypertable_id),
bucket_width,
origin,
NULL::text AS bucket_offset,
timezone,
false AS bucket_fixed_width
FROM
_timescaledb_catalog.continuous_aggs_bucket_function
ORDER BY
mat_hypertable_id;
ALTER EXTENSION timescaledb
DROP TABLE _timescaledb_catalog.continuous_aggs_bucket_function;
DROP TABLE _timescaledb_catalog.continuous_aggs_bucket_function;
CREATE TABLE _timescaledb_catalog.continuous_aggs_bucket_function (
mat_hypertable_id integer NOT NULL,
-- The bucket function
bucket_func regprocedure NOT NULL,
-- `bucket_width` argument of the function, e.g. "1 month"
bucket_width text NOT NULL,
-- optional `origin` argument of the function provided by the user
bucket_origin text,
-- optional `offset` argument of the function provided by the user
bucket_offset text,
-- optional `timezone` argument of the function provided by the user
bucket_timezone text,
-- fixed or variable sized bucket
bucket_fixed_width bool NOT NULL,
-- table constraints
CONSTRAINT continuous_aggs_bucket_function_pkey PRIMARY KEY (mat_hypertable_id),
CONSTRAINT continuous_aggs_bucket_function_mat_hypertable_id_fkey FOREIGN KEY (mat_hypertable_id) REFERENCES _timescaledb_catalog.hypertable (id) ON DELETE CASCADE
);
INSERT INTO _timescaledb_catalog.continuous_aggs_bucket_function
SELECT * FROM _timescaledb_catalog._tmp_continuous_aggs_bucket_function;
DROP TABLE _timescaledb_catalog._tmp_continuous_aggs_bucket_function;
SELECT pg_catalog.pg_extension_config_dump('_timescaledb_catalog.continuous_aggs_bucket_function', '');
GRANT SELECT ON TABLE _timescaledb_catalog.continuous_aggs_bucket_function TO PUBLIC;
ANALYZE _timescaledb_catalog.continuous_aggs_bucket_function;
ALTER EXTENSION timescaledb DROP FUNCTION _timescaledb_functions.cagg_get_bucket_function(INTEGER);
DROP FUNCTION IF EXISTS _timescaledb_functions.cagg_get_bucket_function(INTEGER);
--
-- End rebuild the catalog table `_timescaledb_catalog.continuous_aggs_bucket_function`
--
-- Convert _timescaledb_catalog.continuous_aggs_bucket_function.bucket_origin to TimestampTZ
UPDATE _timescaledb_catalog.continuous_aggs_bucket_function
SET bucket_origin = bucket_origin::timestamp::timestamptz::text
WHERE length(bucket_origin) > 1;
-- Historically, we have used empty strings for undefined bucket_origin and timezone
-- attributes. This is now replaced by proper NULL values. We use TRIM() to ensure we handle empty string well.
UPDATE _timescaledb_catalog.continuous_aggs_bucket_function SET bucket_origin = NULL WHERE TRIM(bucket_origin) = '';
UPDATE _timescaledb_catalog.continuous_aggs_bucket_function SET bucket_timezone = NULL WHERE TRIM(bucket_timezone) = '';
-- So far, there were no difference between 0 and -1 retries. Since now on, 0 means no retries. Updating the retry
-- count of existing jobs to -1 to keep the current semantics.
UPDATE _timescaledb_config.bgw_job SET max_retries = -1 WHERE max_retries = 0;
DROP FUNCTION IF EXISTS _timescaledb_functions.get_chunk_relstats;
DROP FUNCTION IF EXISTS _timescaledb_functions.get_chunk_colstats;
DROP FUNCTION IF EXISTS _timescaledb_internal.get_chunk_relstats;
DROP FUNCTION IF EXISTS _timescaledb_internal.get_chunk_colstats;
-- In older TSDB versions, we disabled autovacuum for compressed chunks
-- to keep the statistics. However, this restriction was removed in
-- #5118 but no migration was performed to remove the custom
-- autovacuum setting for existing chunks.
DO $$
DECLARE
chunk regclass;
BEGIN
FOR chunk IN
SELECT pg_catalog.format('%I.%I', schema_name, table_name)::regclass
FROM _timescaledb_catalog.chunk c
JOIN pg_catalog.pg_class AS pc ON (pc.oid=format('%I.%I', schema_name, table_name)::regclass)
CROSS JOIN unnest(reloptions) AS u(option)
WHERE
dropped = false
AND osm_chunk = false
AND option LIKE 'autovacuum_enabled%'
LOOP
EXECUTE pg_catalog.format('ALTER TABLE %s RESET (autovacuum_enabled);', chunk::text);
END LOOP;
END
$$;
--
-- Rebuild the catalog table `_timescaledb_catalog.continuous_agg`
--
-- (1) Create missing entries in _timescaledb_catalog.continuous_aggs_bucket_function
CREATE OR REPLACE FUNCTION _timescaledb_functions.cagg_get_bucket_function(
mat_hypertable_id INTEGER
) RETURNS regprocedure AS '@MODULE_PATHNAME@', 'ts_continuous_agg_get_bucket_function' LANGUAGE C STRICT VOLATILE;
-- Make sure function points to the new version of TSDB
CREATE OR REPLACE FUNCTION _timescaledb_functions.to_interval(unixtime_us BIGINT) RETURNS INTERVAL
AS '@MODULE_PATHNAME@', 'ts_pg_unix_microseconds_to_interval' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
-- We need to create entries in continuous_aggs_bucket_function for all CAggs that were treated so far
-- as fixed indicated by a bucket_width != -1
INSERT INTO _timescaledb_catalog.continuous_aggs_bucket_function
SELECT
mat_hypertable_id,
_timescaledb_functions.cagg_get_bucket_function(mat_hypertable_id),
-- Intervals needs to be converted into the proper interval format
-- Function name could be prefixed with 'public.'. Therefore LIKE instead of starts_with is used
CASE WHEN _timescaledb_functions.cagg_get_bucket_function(mat_hypertable_id)::text LIKE '%time_bucket(interval,%' THEN
_timescaledb_functions.to_interval(bucket_width)::text
ELSE
bucket_width::text
END,
NULL, -- bucket_origin
NULL, -- bucket_offset
NULL, -- bucket_timezone
true -- bucket_fixed_width
FROM _timescaledb_catalog.continuous_agg WHERE bucket_width != -1;
ALTER EXTENSION timescaledb DROP FUNCTION _timescaledb_functions.cagg_get_bucket_function(INTEGER);
DROP FUNCTION IF EXISTS _timescaledb_functions.cagg_get_bucket_function(INTEGER);
-- (2) Rebuild catalog table
DROP VIEW IF EXISTS timescaledb_experimental.policies;
DROP VIEW IF EXISTS timescaledb_information.hypertables;
DROP VIEW IF EXISTS timescaledb_information.continuous_aggregates;
DROP PROCEDURE IF EXISTS @extschema@.cagg_migrate (REGCLASS, BOOLEAN, BOOLEAN);
DROP FUNCTION IF EXISTS _timescaledb_internal.cagg_migrate_pre_validation (TEXT, TEXT, TEXT);
DROP FUNCTION IF EXISTS _timescaledb_functions.cagg_migrate_pre_validation (TEXT, TEXT, TEXT);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_create_plan (_timescaledb_catalog.continuous_agg, TEXT, BOOLEAN, BOOLEAN);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_create_plan (_timescaledb_catalog.continuous_agg, TEXT, BOOLEAN, BOOLEAN);
DROP FUNCTION IF EXISTS _timescaledb_functions.cagg_migrate_plan_exists (INTEGER);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_plan (_timescaledb_catalog.continuous_agg);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_plan (_timescaledb_catalog.continuous_agg);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_create_new_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_create_new_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_disable_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_disable_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_enable_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_enable_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_copy_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_copy_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_refresh_new_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_refresh_new_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_copy_data (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_copy_data (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_override_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_override_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_drop_old_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_drop_old_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
ALTER TABLE _timescaledb_catalog.continuous_aggs_materialization_invalidation_log
DROP CONSTRAINT continuous_aggs_materialization_invalid_materialization_id_fkey;
ALTER TABLE _timescaledb_catalog.continuous_aggs_watermark
DROP CONSTRAINT continuous_aggs_watermark_mat_hypertable_id_fkey;
ALTER EXTENSION timescaledb
DROP TABLE _timescaledb_catalog.continuous_agg;
CREATE TABLE _timescaledb_catalog._tmp_continuous_agg AS
SELECT
mat_hypertable_id,
raw_hypertable_id,
parent_mat_hypertable_id,
user_view_schema,
user_view_name,
partial_view_schema,
partial_view_name,
direct_view_schema,
direct_view_name,
materialized_only,
finalized
FROM
_timescaledb_catalog.continuous_agg
ORDER BY
mat_hypertable_id;
DROP TABLE _timescaledb_catalog.continuous_agg;
CREATE TABLE _timescaledb_catalog.continuous_agg (
mat_hypertable_id integer NOT NULL,
raw_hypertable_id integer NOT NULL,
parent_mat_hypertable_id integer,
user_view_schema name NOT NULL,
user_view_name name NOT NULL,
partial_view_schema name NOT NULL,
partial_view_name name NOT NULL,
direct_view_schema name NOT NULL,
direct_view_name name NOT NULL,
materialized_only bool NOT NULL DEFAULT FALSE,
finalized bool NOT NULL DEFAULT TRUE,
-- table constraints
CONSTRAINT continuous_agg_pkey PRIMARY KEY (mat_hypertable_id),
CONSTRAINT continuous_agg_partial_view_schema_partial_view_name_key UNIQUE (partial_view_schema, partial_view_name),
CONSTRAINT continuous_agg_user_view_schema_user_view_name_key UNIQUE (user_view_schema, user_view_name),
CONSTRAINT continuous_agg_mat_hypertable_id_fkey
FOREIGN KEY (mat_hypertable_id) REFERENCES _timescaledb_catalog.hypertable (id) ON DELETE CASCADE,
CONSTRAINT continuous_agg_raw_hypertable_id_fkey
FOREIGN KEY (raw_hypertable_id) REFERENCES _timescaledb_catalog.hypertable (id) ON DELETE CASCADE,
CONSTRAINT continuous_agg_parent_mat_hypertable_id_fkey
FOREIGN KEY (parent_mat_hypertable_id)
REFERENCES _timescaledb_catalog.continuous_agg (mat_hypertable_id) ON DELETE CASCADE
);
INSERT INTO _timescaledb_catalog.continuous_agg
SELECT * FROM _timescaledb_catalog._tmp_continuous_agg;
DROP TABLE _timescaledb_catalog._tmp_continuous_agg;
CREATE INDEX continuous_agg_raw_hypertable_id_idx ON _timescaledb_catalog.continuous_agg (raw_hypertable_id);
SELECT pg_catalog.pg_extension_config_dump('_timescaledb_catalog.continuous_agg', '');
GRANT SELECT ON TABLE _timescaledb_catalog.continuous_agg TO PUBLIC;
ALTER TABLE _timescaledb_catalog.continuous_aggs_materialization_invalidation_log
ADD CONSTRAINT continuous_aggs_materialization_invalid_materialization_id_fkey
FOREIGN KEY (materialization_id)
REFERENCES _timescaledb_catalog.continuous_agg(mat_hypertable_id) ON DELETE CASCADE;
ALTER TABLE _timescaledb_catalog.continuous_aggs_watermark
ADD CONSTRAINT continuous_aggs_watermark_mat_hypertable_id_fkey
FOREIGN KEY (mat_hypertable_id)
REFERENCES _timescaledb_catalog.continuous_agg (mat_hypertable_id) ON DELETE CASCADE;
ANALYZE _timescaledb_catalog.continuous_agg;
--
-- END Rebuild the catalog table `_timescaledb_catalog.continuous_agg`
--
--
-- START bgw_job_stat_history
--
DROP VIEW IF EXISTS timescaledb_information.job_errors;
CREATE SEQUENCE _timescaledb_internal.bgw_job_stat_history_id_seq MINVALUE 1;
CREATE TABLE _timescaledb_internal.bgw_job_stat_history (
id INTEGER NOT NULL DEFAULT nextval('_timescaledb_internal.bgw_job_stat_history_id_seq'),
job_id INTEGER NOT NULL,
pid INTEGER,
execution_start TIMESTAMPTZ NOT NULL DEFAULT NOW(),
execution_finish TIMESTAMPTZ,
succeeded boolean NOT NULL DEFAULT FALSE,
data jsonb,
-- table constraints
CONSTRAINT bgw_job_stat_history_pkey PRIMARY KEY (id)
);
ALTER SEQUENCE _timescaledb_internal.bgw_job_stat_history_id_seq OWNED BY _timescaledb_internal.bgw_job_stat_history.id;
CREATE INDEX bgw_job_stat_history_job_id_idx ON _timescaledb_internal.bgw_job_stat_history (job_id);
REVOKE ALL ON _timescaledb_internal.bgw_job_stat_history FROM PUBLIC;
INSERT INTO _timescaledb_internal.bgw_job_stat_history (job_id, pid, execution_start, execution_finish, data)
SELECT
job_errors.job_id,
job_errors.pid,
job_errors.start_time,
job_errors.finish_time,
jsonb_build_object('job', to_jsonb(bgw_job.*)) || jsonb_build_object('error_data', job_errors.error_data)
FROM
_timescaledb_internal.job_errors
LEFT JOIN _timescaledb_config.bgw_job ON bgw_job.id = job_errors.job_id
ORDER BY
job_errors.job_id, job_errors.start_time;
ALTER EXTENSION timescaledb
DROP TABLE _timescaledb_internal.job_errors;
DROP TABLE _timescaledb_internal.job_errors;
UPDATE _timescaledb_config.bgw_job SET scheduled = false WHERE id = 2;
INSERT INTO _timescaledb_config.bgw_job (
id,
application_name,
schedule_interval,
max_runtime,
max_retries,
retry_period,
proc_schema,
proc_name,
owner,
scheduled,
config,
check_schema,
check_name,
fixed_schedule,
initial_start
)
VALUES
(
3,
'Job History Log Retention Policy [3]',
INTERVAL '1 month',
INTERVAL '1 hour',
-1,
INTERVAL '1h',
'_timescaledb_functions',
'policy_job_stat_history_retention',
pg_catalog.quote_ident(current_role)::regrole,
true,
'{"drop_after":"1 month"}',
'_timescaledb_functions',
'policy_job_stat_history_retention_check',
true,
'2000-01-01 00:00:00+00'::timestamptz
) ON CONFLICT (id) DO NOTHING;
DROP FUNCTION IF EXISTS _timescaledb_internal.policy_job_error_retention(job_id integer,config jsonb);
DROP FUNCTION IF EXISTS _timescaledb_internal.policy_job_error_retention_check(config jsonb);
DROP FUNCTION IF EXISTS _timescaledb_functions.policy_job_error_retention(job_id integer,config jsonb);
DROP FUNCTION IF EXISTS _timescaledb_functions.policy_job_error_retention_check(config jsonb);
--
-- END bgw_job_stat_history
--
-- Migrate existing CAggs using time_bucket_ng to time_bucket
CREATE PROCEDURE _timescaledb_functions.cagg_migrate_to_time_bucket(cagg REGCLASS)
AS '@MODULE_PATHNAME@', 'ts_continuous_agg_migrate_to_time_bucket' LANGUAGE C;
DO $$
DECLARE
cagg_name regclass;
BEGIN
FOR cagg_name IN
SELECT pg_catalog.format('%I.%I', user_view_schema, user_view_name)::regclass
FROM _timescaledb_catalog.continuous_agg cagg
JOIN _timescaledb_catalog.continuous_aggs_bucket_function AS bf ON (cagg.mat_hypertable_id = bf.mat_hypertable_id)
WHERE
bf.bucket_func::text LIKE '%time_bucket_ng%'
LOOP
CALL _timescaledb_functions.cagg_migrate_to_time_bucket(cagg_name);
END LOOP;
END
$$;

440
sql/updates/latest-dev.sql
View File

@ -1,440 +0,0 @@
-- Remove multi-node CAGG support
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_cagg_log_add_entry(integer,bigint,bigint);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_hyper_log_add_entry(integer,bigint,bigint);
DROP FUNCTION IF EXISTS _timescaledb_internal.materialization_invalidation_log_delete(integer);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_process_cagg_log(integer,integer,regtype,bigint,bigint,integer[],bigint[],bigint[]);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_process_cagg_log(integer,integer,regtype,bigint,bigint,integer[],bigint[],bigint[],text[]);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_process_hypertable_log(integer,integer,regtype,integer[],bigint[],bigint[]);
DROP FUNCTION IF EXISTS _timescaledb_internal.invalidation_process_hypertable_log(integer,integer,regtype,integer[],bigint[],bigint[],text[]);
DROP FUNCTION IF EXISTS _timescaledb_internal.hypertable_invalidation_log_delete(integer);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_cagg_log_add_entry(integer,bigint,bigint);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_hyper_log_add_entry(integer,bigint,bigint);
DROP FUNCTION IF EXISTS _timescaledb_functions.materialization_invalidation_log_delete(integer);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_process_cagg_log(integer,integer,regtype,bigint,bigint,integer[],bigint[],bigint[]);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_process_cagg_log(integer,integer,regtype,bigint,bigint,integer[],bigint[],bigint[],text[]);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_process_hypertable_log(integer,integer,regtype,integer[],bigint[],bigint[]);
DROP FUNCTION IF EXISTS _timescaledb_functions.invalidation_process_hypertable_log(integer,integer,regtype,integer[],bigint[],bigint[],text[]);
DROP FUNCTION IF EXISTS _timescaledb_functions.hypertable_invalidation_log_delete(integer);
-- Remove chunk metadata when marked as dropped
CREATE FUNCTION _timescaledb_functions.remove_dropped_chunk_metadata(_hypertable_id INTEGER)
RETURNS INTEGER LANGUAGE plpgsql AS $$
DECLARE
_chunk_id INTEGER;
_removed INTEGER := 0;
BEGIN
FOR _chunk_id IN
SELECT id FROM _timescaledb_catalog.chunk
WHERE hypertable_id = _hypertable_id
AND dropped IS TRUE
AND NOT EXISTS (
SELECT FROM information_schema.tables
WHERE tables.table_schema = chunk.schema_name
AND tables.table_name = chunk.table_name
)
AND NOT EXISTS (
SELECT FROM _timescaledb_catalog.hypertable
JOIN _timescaledb_catalog.continuous_agg ON continuous_agg.raw_hypertable_id = hypertable.id
WHERE hypertable.id = chunk.hypertable_id
-- for the old caggs format we need to keep chunk metadata for dropped chunks
AND continuous_agg.finalized IS FALSE
)
LOOP
_removed := _removed + 1;
RAISE INFO 'Removing metadata of chunk % from hypertable %', _chunk_id, _hypertable_id;
WITH _dimension_slice_remove AS (
DELETE FROM _timescaledb_catalog.dimension_slice
USING _timescaledb_catalog.chunk_constraint
WHERE dimension_slice.id = chunk_constraint.dimension_slice_id
AND chunk_constraint.chunk_id = _chunk_id
RETURNING _timescaledb_catalog.dimension_slice.id
)
DELETE FROM _timescaledb_catalog.chunk_constraint
USING _dimension_slice_remove
WHERE chunk_constraint.dimension_slice_id = _dimension_slice_remove.id;
DELETE FROM _timescaledb_internal.bgw_policy_chunk_stats
WHERE bgw_policy_chunk_stats.chunk_id = _chunk_id;
DELETE FROM _timescaledb_catalog.chunk_index
WHERE chunk_index.chunk_id = _chunk_id;
DELETE FROM _timescaledb_catalog.compression_chunk_size
WHERE compression_chunk_size.chunk_id = _chunk_id
OR compression_chunk_size.compressed_chunk_id = _chunk_id;
DELETE FROM _timescaledb_catalog.chunk
WHERE chunk.id = _chunk_id
OR chunk.compressed_chunk_id = _chunk_id;
END LOOP;
RETURN _removed;
END;
$$ SET search_path TO pg_catalog, pg_temp;
SELECT _timescaledb_functions.remove_dropped_chunk_metadata(id) AS chunks_metadata_removed
FROM _timescaledb_catalog.hypertable;
--
-- Rebuild the catalog table `_timescaledb_catalog.continuous_aggs_bucket_function`
--
CREATE OR REPLACE FUNCTION _timescaledb_functions.cagg_get_bucket_function(
mat_hypertable_id INTEGER
) RETURNS regprocedure AS '@MODULE_PATHNAME@', 'ts_continuous_agg_get_bucket_function' LANGUAGE C STRICT VOLATILE;
-- Since we need now the regclass of the used bucket function, we have to recover it
-- by parsing the view query by calling 'cagg_get_bucket_function'.
CREATE TABLE _timescaledb_catalog._tmp_continuous_aggs_bucket_function AS
SELECT
mat_hypertable_id,
_timescaledb_functions.cagg_get_bucket_function(mat_hypertable_id),
bucket_width,
origin,
NULL::text AS bucket_offset,
timezone,
false AS bucket_fixed_width
FROM
_timescaledb_catalog.continuous_aggs_bucket_function
ORDER BY
mat_hypertable_id;
ALTER EXTENSION timescaledb
DROP TABLE _timescaledb_catalog.continuous_aggs_bucket_function;
DROP TABLE _timescaledb_catalog.continuous_aggs_bucket_function;
CREATE TABLE _timescaledb_catalog.continuous_aggs_bucket_function (
mat_hypertable_id integer NOT NULL,
-- The bucket function
bucket_func regprocedure NOT NULL,
-- `bucket_width` argument of the function, e.g. "1 month"
bucket_width text NOT NULL,
-- optional `origin` argument of the function provided by the user
bucket_origin text,
-- optional `offset` argument of the function provided by the user
bucket_offset text,
-- optional `timezone` argument of the function provided by the user
bucket_timezone text,
-- fixed or variable sized bucket
bucket_fixed_width bool NOT NULL,
-- table constraints
CONSTRAINT continuous_aggs_bucket_function_pkey PRIMARY KEY (mat_hypertable_id),
CONSTRAINT continuous_aggs_bucket_function_mat_hypertable_id_fkey FOREIGN KEY (mat_hypertable_id) REFERENCES _timescaledb_catalog.hypertable (id) ON DELETE CASCADE
);
INSERT INTO _timescaledb_catalog.continuous_aggs_bucket_function
SELECT * FROM _timescaledb_catalog._tmp_continuous_aggs_bucket_function;
DROP TABLE _timescaledb_catalog._tmp_continuous_aggs_bucket_function;
SELECT pg_catalog.pg_extension_config_dump('_timescaledb_catalog.continuous_aggs_bucket_function', '');
GRANT SELECT ON TABLE _timescaledb_catalog.continuous_aggs_bucket_function TO PUBLIC;
ANALYZE _timescaledb_catalog.continuous_aggs_bucket_function;
ALTER EXTENSION timescaledb DROP FUNCTION _timescaledb_functions.cagg_get_bucket_function(INTEGER);
DROP FUNCTION IF EXISTS _timescaledb_functions.cagg_get_bucket_function(INTEGER);
--
-- End rebuild the catalog table `_timescaledb_catalog.continuous_aggs_bucket_function`
--
-- Convert _timescaledb_catalog.continuous_aggs_bucket_function.bucket_origin to TimestampTZ
UPDATE _timescaledb_catalog.continuous_aggs_bucket_function
SET bucket_origin = bucket_origin::timestamp::timestamptz::text
WHERE length(bucket_origin) > 1;
-- Historically, we have used empty strings for undefined bucket_origin and timezone
-- attributes. This is now replaced by proper NULL values. We use TRIM() to ensure we handle empty string well.
UPDATE _timescaledb_catalog.continuous_aggs_bucket_function SET bucket_origin = NULL WHERE TRIM(bucket_origin) = '';
UPDATE _timescaledb_catalog.continuous_aggs_bucket_function SET bucket_timezone = NULL WHERE TRIM(bucket_timezone) = '';
-- So far, there were no difference between 0 and -1 retries. Since now on, 0 means no retries. Updating the retry
-- count of existing jobs to -1 to keep the current semantics.
UPDATE _timescaledb_config.bgw_job SET max_retries = -1 WHERE max_retries = 0;
DROP FUNCTION IF EXISTS _timescaledb_functions.get_chunk_relstats;
DROP FUNCTION IF EXISTS _timescaledb_functions.get_chunk_colstats;
DROP FUNCTION IF EXISTS _timescaledb_internal.get_chunk_relstats;
DROP FUNCTION IF EXISTS _timescaledb_internal.get_chunk_colstats;
-- In older TSDB versions, we disabled autovacuum for compressed chunks
-- to keep the statistics. However, this restriction was removed in
-- #5118 but no migration was performed to remove the custom
-- autovacuum setting for existing chunks.
DO $$
DECLARE
chunk regclass;
BEGIN
FOR chunk IN
SELECT pg_catalog.format('%I.%I', schema_name, table_name)::regclass
FROM _timescaledb_catalog.chunk c
JOIN pg_catalog.pg_class AS pc ON (pc.oid=format('%I.%I', schema_name, table_name)::regclass)
CROSS JOIN unnest(reloptions) AS u(option)
WHERE
dropped = false
AND osm_chunk = false
AND option LIKE 'autovacuum_enabled%'
LOOP
EXECUTE pg_catalog.format('ALTER TABLE %s RESET (autovacuum_enabled);', chunk::text);
END LOOP;
END
$$;
--
-- Rebuild the catalog table `_timescaledb_catalog.continuous_agg`
--
-- (1) Create missing entries in _timescaledb_catalog.continuous_aggs_bucket_function
CREATE OR REPLACE FUNCTION _timescaledb_functions.cagg_get_bucket_function(
mat_hypertable_id INTEGER
) RETURNS regprocedure AS '@MODULE_PATHNAME@', 'ts_continuous_agg_get_bucket_function' LANGUAGE C STRICT VOLATILE;
-- Make sure function points to the new version of TSDB
CREATE OR REPLACE FUNCTION _timescaledb_functions.to_interval(unixtime_us BIGINT) RETURNS INTERVAL
AS '@MODULE_PATHNAME@', 'ts_pg_unix_microseconds_to_interval' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
-- We need to create entries in continuous_aggs_bucket_function for all CAggs that were treated so far
-- as fixed indicated by a bucket_width != -1
INSERT INTO _timescaledb_catalog.continuous_aggs_bucket_function
SELECT
mat_hypertable_id,
_timescaledb_functions.cagg_get_bucket_function(mat_hypertable_id),
-- Intervals needs to be converted into the proper interval format
-- Function name could be prefixed with 'public.'. Therefore LIKE instead of starts_with is used
CASE WHEN _timescaledb_functions.cagg_get_bucket_function(mat_hypertable_id)::text LIKE '%time_bucket(interval,%' THEN
_timescaledb_functions.to_interval(bucket_width)::text
ELSE
bucket_width::text
END,
NULL, -- bucket_origin
NULL, -- bucket_offset
NULL, -- bucket_timezone
true -- bucket_fixed_width
FROM _timescaledb_catalog.continuous_agg WHERE bucket_width != -1;
ALTER EXTENSION timescaledb DROP FUNCTION _timescaledb_functions.cagg_get_bucket_function(INTEGER);
DROP FUNCTION IF EXISTS _timescaledb_functions.cagg_get_bucket_function(INTEGER);
-- (2) Rebuild catalog table
DROP VIEW IF EXISTS timescaledb_experimental.policies;
DROP VIEW IF EXISTS timescaledb_information.hypertables;
DROP VIEW IF EXISTS timescaledb_information.continuous_aggregates;
DROP PROCEDURE IF EXISTS @extschema@.cagg_migrate (REGCLASS, BOOLEAN, BOOLEAN);
DROP FUNCTION IF EXISTS _timescaledb_internal.cagg_migrate_pre_validation (TEXT, TEXT, TEXT);
DROP FUNCTION IF EXISTS _timescaledb_functions.cagg_migrate_pre_validation (TEXT, TEXT, TEXT);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_create_plan (_timescaledb_catalog.continuous_agg, TEXT, BOOLEAN, BOOLEAN);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_create_plan (_timescaledb_catalog.continuous_agg, TEXT, BOOLEAN, BOOLEAN);
DROP FUNCTION IF EXISTS _timescaledb_functions.cagg_migrate_plan_exists (INTEGER);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_plan (_timescaledb_catalog.continuous_agg);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_plan (_timescaledb_catalog.continuous_agg);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_create_new_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_create_new_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_disable_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_disable_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_enable_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_enable_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_copy_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_copy_policies (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_refresh_new_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_refresh_new_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_copy_data (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_copy_data (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_override_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_override_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_internal.cagg_migrate_execute_drop_old_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
DROP PROCEDURE IF EXISTS _timescaledb_functions.cagg_migrate_execute_drop_old_cagg (_timescaledb_catalog.continuous_agg, _timescaledb_catalog.continuous_agg_migrate_plan_step);
ALTER TABLE _timescaledb_catalog.continuous_aggs_materialization_invalidation_log
DROP CONSTRAINT continuous_aggs_materialization_invalid_materialization_id_fkey;
ALTER TABLE _timescaledb_catalog.continuous_aggs_watermark
DROP CONSTRAINT continuous_aggs_watermark_mat_hypertable_id_fkey;
ALTER EXTENSION timescaledb
DROP TABLE _timescaledb_catalog.continuous_agg;
CREATE TABLE _timescaledb_catalog._tmp_continuous_agg AS
SELECT
mat_hypertable_id,
raw_hypertable_id,
parent_mat_hypertable_id,
user_view_schema,
user_view_name,
partial_view_schema,
partial_view_name,
direct_view_schema,
direct_view_name,
materialized_only,
finalized
FROM
_timescaledb_catalog.continuous_agg
ORDER BY
mat_hypertable_id;
DROP TABLE _timescaledb_catalog.continuous_agg;
CREATE TABLE _timescaledb_catalog.continuous_agg (
mat_hypertable_id integer NOT NULL,
raw_hypertable_id integer NOT NULL,
parent_mat_hypertable_id integer,
user_view_schema name NOT NULL,
user_view_name name NOT NULL,
partial_view_schema name NOT NULL,
partial_view_name name NOT NULL,
direct_view_schema name NOT NULL,
direct_view_name name NOT NULL,
materialized_only bool NOT NULL DEFAULT FALSE,
finalized bool NOT NULL DEFAULT TRUE,
-- table constraints
CONSTRAINT continuous_agg_pkey PRIMARY KEY (mat_hypertable_id),
CONSTRAINT continuous_agg_partial_view_schema_partial_view_name_key UNIQUE (partial_view_schema, partial_view_name),
CONSTRAINT continuous_agg_user_view_schema_user_view_name_key UNIQUE (user_view_schema, user_view_name),
CONSTRAINT continuous_agg_mat_hypertable_id_fkey
FOREIGN KEY (mat_hypertable_id) REFERENCES _timescaledb_catalog.hypertable (id) ON DELETE CASCADE,
CONSTRAINT continuous_agg_raw_hypertable_id_fkey
FOREIGN KEY (raw_hypertable_id) REFERENCES _timescaledb_catalog.hypertable (id) ON DELETE CASCADE,
CONSTRAINT continuous_agg_parent_mat_hypertable_id_fkey
FOREIGN KEY (parent_mat_hypertable_id)
REFERENCES _timescaledb_catalog.continuous_agg (mat_hypertable_id) ON DELETE CASCADE
);
INSERT INTO _timescaledb_catalog.continuous_agg
SELECT * FROM _timescaledb_catalog._tmp_continuous_agg;
DROP TABLE _timescaledb_catalog._tmp_continuous_agg;
CREATE INDEX continuous_agg_raw_hypertable_id_idx ON _timescaledb_catalog.continuous_agg (raw_hypertable_id);
SELECT pg_catalog.pg_extension_config_dump('_timescaledb_catalog.continuous_agg', '');
GRANT SELECT ON TABLE _timescaledb_catalog.continuous_agg TO PUBLIC;
ALTER TABLE _timescaledb_catalog.continuous_aggs_materialization_invalidation_log
ADD CONSTRAINT continuous_aggs_materialization_invalid_materialization_id_fkey
FOREIGN KEY (materialization_id)
REFERENCES _timescaledb_catalog.continuous_agg(mat_hypertable_id) ON DELETE CASCADE;
ALTER TABLE _timescaledb_catalog.continuous_aggs_watermark
ADD CONSTRAINT continuous_aggs_watermark_mat_hypertable_id_fkey
FOREIGN KEY (mat_hypertable_id)
REFERENCES _timescaledb_catalog.continuous_agg (mat_hypertable_id) ON DELETE CASCADE;
ANALYZE _timescaledb_catalog.continuous_agg;
--
-- END Rebuild the catalog table `_timescaledb_catalog.continuous_agg`
--
--
-- START bgw_job_stat_history
--
DROP VIEW IF EXISTS timescaledb_information.job_errors;
CREATE SEQUENCE _timescaledb_internal.bgw_job_stat_history_id_seq MINVALUE 1;
CREATE TABLE _timescaledb_internal.bgw_job_stat_history (
id INTEGER NOT NULL DEFAULT nextval('_timescaledb_internal.bgw_job_stat_history_id_seq'),
job_id INTEGER NOT NULL,
pid INTEGER,
execution_start TIMESTAMPTZ NOT NULL DEFAULT NOW(),
execution_finish TIMESTAMPTZ,
succeeded boolean NOT NULL DEFAULT FALSE,
data jsonb,
-- table constraints
CONSTRAINT bgw_job_stat_history_pkey PRIMARY KEY (id)
);
ALTER SEQUENCE _timescaledb_internal.bgw_job_stat_history_id_seq OWNED BY _timescaledb_internal.bgw_job_stat_history.id;
CREATE INDEX bgw_job_stat_history_job_id_idx ON _timescaledb_internal.bgw_job_stat_history (job_id);
REVOKE ALL ON _timescaledb_internal.bgw_job_stat_history FROM PUBLIC;
INSERT INTO _timescaledb_internal.bgw_job_stat_history (job_id, pid, execution_start, execution_finish, data)
SELECT
job_errors.job_id,
job_errors.pid,
job_errors.start_time,
job_errors.finish_time,
jsonb_build_object('job', to_jsonb(bgw_job.*)) || jsonb_build_object('error_data', job_errors.error_data)
FROM
_timescaledb_internal.job_errors
LEFT JOIN _timescaledb_config.bgw_job ON bgw_job.id = job_errors.job_id
ORDER BY
job_errors.job_id, job_errors.start_time;
ALTER EXTENSION timescaledb
DROP TABLE _timescaledb_internal.job_errors;
DROP TABLE _timescaledb_internal.job_errors;
UPDATE _timescaledb_config.bgw_job SET scheduled = false WHERE id = 2;
INSERT INTO _timescaledb_config.bgw_job (
id,
application_name,
schedule_interval,
max_runtime,
max_retries,
retry_period,
proc_schema,
proc_name,
owner,
scheduled,
config,
check_schema,
check_name,
fixed_schedule,
initial_start
)
VALUES
(
3,
'Job History Log Retention Policy [3]',
INTERVAL '1 month',
INTERVAL '1 hour',
-1,
INTERVAL '1h',
'_timescaledb_functions',
'policy_job_stat_history_retention',
pg_catalog.quote_ident(current_role)::regrole,
true,
'{"drop_after":"1 month"}',
'_timescaledb_functions',
'policy_job_stat_history_retention_check',
true,
'2000-01-01 00:00:00+00'::timestamptz
) ON CONFLICT (id) DO NOTHING;
DROP FUNCTION IF EXISTS _timescaledb_internal.policy_job_error_retention(job_id integer,config jsonb);
DROP FUNCTION IF EXISTS _timescaledb_internal.policy_job_error_retention_check(config jsonb);
DROP FUNCTION IF EXISTS _timescaledb_functions.policy_job_error_retention(job_id integer,config jsonb);
DROP FUNCTION IF EXISTS _timescaledb_functions.policy_job_error_retention_check(config jsonb);
--
-- END bgw_job_stat_history
--
-- Migrate existing CAggs using time_bucket_ng to time_bucket
CREATE PROCEDURE _timescaledb_functions.cagg_migrate_to_time_bucket(cagg REGCLASS)
AS '@MODULE_PATHNAME@', 'ts_continuous_agg_migrate_to_time_bucket' LANGUAGE C;
DO $$
DECLARE
cagg_name regclass;
BEGIN
FOR cagg_name IN
SELECT pg_catalog.format('%I.%I', user_view_schema, user_view_name)::regclass
FROM _timescaledb_catalog.continuous_agg cagg
JOIN _timescaledb_catalog.continuous_aggs_bucket_function AS bf ON (cagg.mat_hypertable_id = bf.mat_hypertable_id)
WHERE
bf.bucket_func::text LIKE '%time_bucket_ng%'
LOOP
CALL _timescaledb_functions.cagg_migrate_to_time_bucket(cagg_name);
END LOOP;
END
$$;

4
version.config
View File

@ -1,3 +1,3 @@
version = 2.15.0-dev
update_from_version = 2.14.2
version = 2.16.0-dev
update_from_version = 2.15.0
downgrade_to_version = 2.14.2