(Upstream also has news for beta and rc, and it's not clear if those
are additional or folded into the following.)
PostGIS 2.5.0
2018/09/23
WARNING: If compiling with PostgreSQL+JIT, LLVM >= 6 is required
Supported PostgreSQL versions for this release are:
PostgreSQL 9.4 - PostgreSQL 12 (in development)
GEOS >= 3.5
* New Features *
- #1847, spgist 2d and 3d support for PG 11+
(Esteban Zimányi and Arthur Lesuisse from Université Libre de Bruxelles (ULB),
Darafei Praliaskouski)
- #4056, ST_FilterByM (Nicklas Avén)
- #4050, ST_ChaikinSmoothing (Nicklas Avén)
- #3989, ST_Buffer single sided option (Stephen Knox)
- #3876, ST_Angle function (Rémi Cura)
- #3564, ST_LineInterpolatePoints (Dan Baston)
- #3896, PostGIS_Extensions_Upgrade() (Regina Obe)
- #3913, Upgrade when creating extension from unpackaged (Sandro Santilli)
- #2256, _postgis_index_extent() for extent from index (Paul Ramsey)
- #3176, Add ST_OrientedEnvelope (Dan Baston)
- #4029, Add ST_QuantizeCoordinates (Dan Baston)
- #4063, Optional false origin point for ST_Scale (Paul Ramsey)
- #4082, Add ST_BandFileSize and ST_BandFileTimestamp,
extend ST_BandMetadata (Even Rouault)
- #2597, Add ST_Grayscale (Bborie Park)
- #4007, Add ST_SetBandPath (Bborie Park)
- #4008, Add ST_SetBandIndex (Bborie Park)
* Breaking Changes *
- #4054, ST_SimplifyVW changed from > tolerance to >= tolerance
- #3885, version number removed from address_standardize lib file
(Regina Obe)
- #3893, raster support functions can only be loaded in the same schema
with core PostGIS functions. (Sandro Santilli)
- #4035, remove dummy pgis_abs type from aggregate/collect routines.
(Paul Ramsey)
- #4069, drop support for GEOS < 3.5 and PostgreSQL < 9.4 (Regina Obe)
- #4082, ST_BandMetaData extended to include filesize, timestamp (Even Rouault)
* Enhancements and Fixes*
- Upgrade scripts from multiple old versions are now all symlinks
to a single upgrade script (Sandro Santilli)
- #3944, Update to EPSG register v9.2 (Even Rouault)
- #3927, Parallel implementation of ST_AsMVT
- #3925, Simplify geometry using map grid cell size before generating MVT
- #3899, BTree sort order is now defined on collections of EMPTY and
same-prefix geometries (Darafei Praliaskouski)
- #3864, Performance improvement for sorting POINT geometries
(Darafei Praliaskouski)
- #3900, GCC warnings fixed, make -j is now working (Darafei Praliaskouski)
- TopoGeo_addLinestring robustness improvements (Sandro Santilli)
#1855, #1946, #3718, #3838
- #3234, Do not accept EMPTY points as topology nodes (Sandro Santilli)
- #1014, Hashable geometry, allowing direct use in CTE signatures (Paul Ramsey)
- #3097, Really allow MULTILINESTRING blades in ST_Split() (Paul Ramsey)
- #3942, geojson: Do not include private header for json-c >= 0.13 (Björn Esser)
- #3954, ST_GeometricMedian now supports point weights (Darafei Praliaskouski)
- #3965, #3971, #3977, #4071 ST_ClusterKMeans rewritten: better initialization,
faster convergence, K=2 even faster (Darafei Praliaskouski)
- #3982, ST_AsEncodedPolyline supports LINESTRING EMPTY and MULTIPOINT EMPTY
(Darafei Praliaskouski)
- #3986, ST_AsText now has second argument to limit decimal digits
(Marc Ducobu, Darafei Praliaskouski)
- #4020, Casting from box3d to geometry now returns correctly connected
PolyhedralSurface (Matthias Bay)
- #2508, ST_OffsetCurve now works with collections (Darafei Praliaskouski)
- #4006, ST_GeomFromGeoJSON support for json and jsonb as input
(Paul Ramsey, Regina Obe)
- #4038, ST_Subdivide now selects pivot for geometry split that reuses input
vertices. (Darafei Praliaskouski)
- #4025, #4032 Fixed precision issue in ST_ClosestPointOfApproach,
ST_DistanceCPA, and ST_CPAWithin (Paul Ramsey, Darafei Praliaskouski)
- #4076, Reduce use of GEOS in topology implementation (Björn Harrtell)
- #4080, Add external raster band index to ST_BandMetaData
- Add Raster Tips section to Documentation for information about
Raster behavior (e.g. Out-DB performance, maximum open files)
- #4084: Fixed wrong code-comment regarding front/back of BOX3D (Matthias Bay)
- #4060, #4094, PostgreSQL JIT support (Raúl Marín, Laurenz Albe)
- #3960, ST_Centroid now uses lwgeom_centroid (Darafei Praliaskouski)
- #4027, Remove duplicated code in lwgeom_geos (Darafei Praliaskouski,
Daniel Baston)
- #4115, Fix a bug that created MVTs with incorrect property values under
parallel plans (Raúl Marín).
- #4120, ST_AsMVTGeom: Clip using tile coordinates (Raúl Marín).
- #4132, ST_Intersection on Raster now works without throwing TopologyException
(Vinícius A.B. Schmidt, Darafei Praliaskouski)
- #4177, #4180 Support for PostgreSQL 12 dev branch (Laurenz Albe, Raúl Marín)
- #4156, ST_ChaikinSmoothing: also smooth start/end point of
polygon by default (Darafei Praliaskouski)
Currently, this package builds with all versions of postgreqsl in
pkgsrc, and it is likely to do so in the future, so there is no need
for an explicit list.
== v1.1.3 [2018-09-06] Michael Granger <ged@FaerieMUD.org>
- Revert opimization that was sometimes causing EBADF in rb_wait_for_single_fd().
== v1.1.2 [2018-08-28] Michael Granger <ged@FaerieMUD.org>
- Don't generate aliases for JOHAB encoding.
This avoids linking to deprecated/private function rb_enc(db)_alias().
== v1.1.1 [2018-08-27] Michael Granger <ged@FaerieMUD.org>
- Reduce deprecation warnings to only one message per deprecation.
== v1.1.0 [2018-08-24] Michael Granger <ged@FaerieMUD.org>
Deprecated (disable warnings per PG_SKIP_DEPRECATION_WARNING=1):
- Forwarding conn.exec to conn.exec_params is deprecated.
- Forwarding conn.exec_params to conn.exec is deprecated.
- Forwarding conn.async_exec to conn.async_exec_params.
- Forwarding conn.send_query to conn.send_query_params is deprecated.
- Forwarding conn.async_exec_params to conn.async_exec is deprecated.
PG::Connection enhancements:
- Provide PG::Connection#sync_* and PG::Connection#async_* query methods for explicit calling syncronous or asynchronous libpq API.
- Make PG::Connection#exec and siblings switchable between sync and async API per PG::Connection.async_api= and change the default to async flavors.
- Add async flavors of exec_params, prepare, exec_prepared, describe_prepared and describe_portal.
They are identical to their syncronous counterpart, but make use of PostgreSQL's async API.
- Replace `rb_thread_fd_select()` by faster `rb_wait_for_single_fd()` in `conn.block` and `conn.async_exec` .
- Add PG::Connection#discard_results .
- Raise an ArgumentError for strings containing zero bytes by #escape, #escape_literal, #escape_identifier, #quote_ident and PG::TextEncoder::Identifier. These methods previously truncated strings.
Result retrieval enhancements:
- Add PG::Result#tuple_values to retrieve all field values of a row as array.
- Add PG::Tuple, PG::Result#tuple and PG::Result#stream_each_tuple .
PG::Tuple offers a way to lazy cast result values.
- Estimate PG::Result size allocated by libpq and notify the garbage collector about it when running on Ruby-2.4 or newer.
- Make the estimated PG::Result size available to ObjectSpace.memsize_of(result) .
Type cast enhancements:
- Replace Ruby code by a faster C implementation of the SimpleDecoder's timestamp decode functions. github #20
- Interpret years with up to 7 digists and BC dates by timestamp decoder.
- Add text timestamp decoders for UTC vs. local timezone variations.
- Add text timestamp encoders for UTC timezone.
- Add decoders for binary timestamps: PG::BinaryDecoder::Timestamp and variations.
- Add PG::Coder#flags accessor to allow modifications of de- respectively encoder behaviour.
- Add a flag to raise TypeError for invalid input values to PG::TextDecoder::Array .
- Add a text decoder for inet/cidr written in C.
- Add a numeric decoder written in C.
- Ensure input text is zero terminated for text format in PG::Coder#decode .
Source code enhancements:
- Fix headers and permission bits of various repository files.
Bugfixes:
- Properly decode array with prepended dimensions. #272
For now dimension decorations are ignored, but a correct Array is returned.
- Array-Decoder: Avoid leaking memory when an Exception is raised while parsing. Fixes#279
0.5.2 (2018/07/06)
New Features
None
Bug Fixes
* Bounds-check the encoding translation array from MySQL to Ruby encodings.
Changes
* Add more encoding translations for MySQL 8 compatibility.
Upstream changes:
2018-09-15 Daniël van Eeden, Patrick Galbraith, Michiel Beijen,
DBI/DBD community (4.048)
* Fix corrupted META.json so cpan installations work as expected.
https://github.com/perl5-dbi/DBD-mysql/issues/263
0.33.5:
Added optional attr parameter for locale calleble in TranslationHybrid
Fixed an issue with PasswordType so that it is compatible with older versions of passlib
0.33.4:
Made PasswordType use hash function instead of deprecated encrypt function
Release 1.2.12 includes a series of mostly minor bugfixes, as well as one critical issue allowing the recently released version 7.0 of the cx_Oracle DBAPI to work correctly.
SQLite Release 3.25.1:
Extra sanity checking added to ALTER TABLE in the 3.25.0 release sometimes raises a false-positive when the table being modified has a trigger that updates a virtual table. The false-positive caused the ALTER TABLE to rollback, thus leaving the schema unchanged.
The fix in the 3.25.0 release for the endless-loop in the byte-code associated with the ORDER BY LIMIT optimization did not work for some queries involving window functions. An additional correction is required.
SQLite Release 3.25.0:
Add support for window functions
Enhancements the ALTER TABLE command:
- Add support for renaming columns within a table using ALTER TABLE table RENAME COLUMN oldname TO newname.
- Fix table rename feature so that it also updates references to the renamed table in triggers and views.
Query optimizer improvements:
- Avoid unnecessary loads of columns in an aggregate query that are not within an aggregate function and that are not part of the GROUP BY clause.
- The IN-early-out optimization: When doing a look-up on a multi-column index and an IN operator is used on a column other than the left-most column, then if no rows match against the first IN value, check to make sure there exist rows that match the columns to the right before continuing with the next IN value.
- Use the transitive property to try to propagate constant values within the WHERE clause. For example, convert "a=99 AND b=a" into "a=99 AND b=99".
Use a separate mutex on every inode in the unix VFS, rather than a single mutex shared among them all, for slightly better concurrency in multi-threaded environments.
Enhance the PRAGMA integrity_check command for improved detection of problems on the page freelist.
Output infinity as 1e999 in the ".dump" command of the command-line shell.
Added the SQLITE_FCNTL_DATA_VERSION file-control.
Added the Geopoly module
Bug fixes:
The ORDER BY LIMIT optimization might have caused an infinite loop in the byte code of the prepared statement under very obscure circumstances, due to a confluence of minor defects in the query optimizer.
On an UPSERT when the order of constraint checks is rearranged, ensure that the affinity transformations on the inserted content occur before any of the constraint checks.
Avoid using a prepared statement for ".stats on" command of the CLI after it has been closed by the ".eqp full" logicc.
The LIKE optimization was generating incorrect byte-code and hence getting the wrong answer if the left-hand operand has numeric affinity and the right-hand-side pattern is '/%' or if the pattern begins with the ESCAPE character.
v3.0.12 released 2018-09-13
- Fixed bug PT-1611: pt-archiver fails with UTF-8 chars
- Fixed bug PT-1574: pt-online-schema-change fails on UK and NULLs
- Fixed bug PT-1572: Better usage of ENUM fields in keys in
NibbleIterator
- Fixed bug PT-1422: pt-mysql-summary may get stuck when Time: NULL
in processlist
- Improvement PT-1321: Add required MySQL privileges to
pt-online-schema-change documentation
v3.0.11 released 2018-07-06
- Improvement PT-1571 : Improved hostname recognition in
pt-secure-collect
- Fixed bug PT-1570 : pt-archiver fails to detect columns with the
word GENERATED as part of the comment
- Improvement PT-1569 : Disabled --alter-foreign-keys-method=drop_swap
in pt-osc
- Fixed bug PT-1563 : Fixed pt-show-grants for MySQL 5.6
- Improvement PT-1562 : pt-mysql-summary: Fix mysqld command for
Travis
- Fixed bug PT-1551 : pt-table-checksum fails on MySQL 8.0.11
- Improvement PT-242 : (pt-stalk) Include SHOW SLAVE STATUS on 5.7
- Fixed bug PT-241 : (pt-stalk) Slave queries doesn't run on 5.7
v3.0.10 released 2018-05-21
- Fixed bug PT-1556 : pt-table-checksum 3.0.9 doesn't change
binlog_format to statement anymore
- Improvement PT-1546 : Improved support of MySQL 8 roles
- Improvement PT-1543 : Encrypted table status query causes high load
over multiple minutes
- Improvement PT-1536 : Add info about encrypted tablespaces in
pt-mysql-summary
- Feature PT-131 : Make pt-table checksum to disable QRT plugin
- Feature PT-118 : pt-table-checksum report the number of rows of
difference between master and slave
v3.0.9 released 2018-04-17
- Feature PT-1530 : Add support for encryption status to
mysql-summary
- Feature PT-1526 : Add ndb status to pt-mysql-summary
- Feature PT-1525 : Added support for MySQL 8 roles into
pt-mysql-summary
- Feature PT-1509 : Only set binlog_format when necessary
- Feature PT-1508 : Adding --read-only-interval flag, and
read-only check on wake-up
- Improvement PT-1507 : pt-summary does not reliably read in the
transparent huge pages setting
- New tool PT-1501 : pt-secure-collect - New tool to collect and
sanitize pt-tools outputs
- Feature PT-243 : Adding --max-hostname-length and
--max-line-length to pt-query-digest
v3.0.8 released 2018-03-13
- Feature PT-1500 : add --secure-slowlog option to pt-query digest
v3.0.7 released 2018-03-01
- Fixed Bug PT-244 : pt-online-schema-change --data-dir option
broken for partitioned table
- Feature PT-633 : Added --mysql-only option to pt-stalk for RDS
- Fixed bug PT-1256: pt-table-sync does not use the character set
for the table it is synchronizing
- Fixed bug PT-1455: pt-osc is stuck when the table that is being
altered is filtered out in the slave
- Fixed bug PT-1485: pt-mysql-summary has broken Security section
in versions bigger then 5.6
- Fixed bug PMM-1905: Explain fails if encounters negative
"ntoreturn"
=== 5.12.0 (2018-08-31)
* Make constraint_validations extension respect Database#constraint_validations_table setting (jeremyevans)
* Make Sequel.extension load files from gems (jeremyevans)
* Map clob prepared statement argument type to OCI8::CLOB in the oracle adapter (pipistrellka) (#1534)
* Make Model.load_cache public in the static_cache plugin (AlexWayfer) (#1533)
* Enable support for NOWAIT on MariaDB 10.3+ (jeremyevans)
* Enable support for INTERSECT and EXCEPT on MariaDB 10.3+ (jeremyevans)
* Make tactical_eager_loading plugin handle automatic eager loading for associated objects created by eager_graph (jeremyevans)
* Cache eager_graph loader to speed up subsequent loads from the same dataset (jeremyevans)
* Add caller_logging database extension to log callers before queries, useful during development (jeremyevans)
* Add Database#call_procedure in the postgres adapter for calling PostgreSQL 11+ procedures (jeremyevans)
* Add eager_graph_eager plugin for chaining eager association loads after eager_graph association loads (jeremyevans)
* Support using Dataset#eager_graph in eager load callback for associations using join tables (jeremyevans)
* Make Dataset#graph handle existing selections without determinable aliases by forcing a subselect (jeremyevans)
* Freeze prepared statement arguments before returning the prepared statement (jeremyevans)
* Refactor emulated prepared statement internals to use a placeholder literalizer (jeremyevans)
=== 5.11.0 (2018-08-01)
* Fix using the jdbc/sqlserver adapter on JRuby 9.2+ (jeremyevans)
* Fix dumping schema for numeric/decimal columns with default values, broken starting in 5.9.0 (jeremyevans)
* Recognize additional check constraint violations on certain versions of SQLite (jeremyevans)
* Use cached model instances for Model.first calls without an argument or with a single integer argument in the static_cache plugin (AlexWayfer) (#1529)
* Support ON CONFLICT clause for INSERT on SQLite 3.24+ (jeremyevans)
* Support Dataset#window for WINDOW clause on MySQL 8 and SQLAnywhere (jeremyevans)
* Enable window function support on SQLAnywhere (jeremyevans)
* Support using a hash as a window function :frame option value, with support for ROWS/RANGE/GROUPS, numeric offsets, and EXCLUDE (jeremyevans)
* Allow using set_column_default with a nil value to remove the default value for a column on MySQL when the column is NOT NULL (jeremyevans)
=== 5.10.0 (2018-07-01)
* Use input type casts when using the postgres adapter with pg 0.18+ to reduce string allocations for some primitive types used as prepared statement arguments (jeremyevans)
* Assume local time if database timezone not specified when handling BC timestamps on JRuby 9.2.0.0 in the pg_extended_date_support extension (jeremyevans)
* Fix parsing of timetz types in the jdbc/postgresql adapter (jeremyevans)
* Make SQLTime.parse respect SQLTime.date and Sequel.application_timezone (jeremyevans)
* Add :top as an option in the list plugin (celsworth) (#1526)
* Fix Model#{ancestors,descendants,self_and_siblings} in the tree plugin when custom parent/children association names are used (jeremyevans) (#1525)
* Treat read-only mode error as disconnect error on mysql and mysql2 adapters, for better behavior on AWS Aurora cluster (jeremyevans)
* Don't use cached placeholder literalizers for in Dataset#{first,where_all,where_each,where_single_value} if argument is empty array or hash (jeremyevans)
* Support :tablespace option when adding tables, indexes, and materialized views on PostgreSQL (jeremyevans)
* Support :include option for indexes on PostgreSQL 11+ (jeremyevans)
* Allow the use of IN/NOT IN operators with set returning functions for Sequel::Model datasets (jeremyevans)
* Make many_to_pg_array associations in the pg_array_associations plugin work on PostgreSQL 11 (jeremyevans)
* Only load strscan library in pg_array extension if it is needed (jeremyevans)
* Don't remove related many_to_one associations from cache when setting column value to existing value for model instances that have not been persisted (jeremyevans) (#1521)
* Support ruby 2.6+ endless ranges in the pg_range extension (jeremyevans)
* Support ruby 2.6+ endless ranges in filters, using just a >= operator for them (jeremyevans)
2018-09-08 Daniël van Eeden, Patrick Galbraith, DBI/DBD community (4.047)
* Add options needed for public key based security.
* Allow several spaces after LIMIT clause.
* Basic GTID tracking.
* Fix GCC 8.1 warnings
* Fix warning: unused variable ‘buffer_len’
* Fix warning about bind->is_null assignment with incompatible pointer type
* Get rid of warning about unused variable ssl_verify_set
* Fix space/tab issue which leads to misleading indentation
* Change spaces/tabs to get rid of GCC8 warning
2018-03-09 Patrick Galbraith, Michiel Beijen, DBI/DBD community (4.046_01)
* Updated documentation re. multithreading (Daniël van Eeden)
* Remove #IFDEFs for code detecting ancient DBI versions. The minimum
DBI version we require is version 1.609 from 2009!
* ChopBlanks should not trim binary fields (urcheon)
* Skipped test which failed on OpenBSD because Proc::ProcessTable does not
come with a 'size' attribute on this platform (glasswalk3r, pali)
* Improved test suite with regards to MySQL 8.0
* Improved macOS installation notes.
* Use Devel::CheckLib 1.09 or newer, fixes
https://github.com/perl5-dbi/DBD-mysql/issues/109
* Specify bigint as test dependency
* Add recommends and suggests tests dependencies (pali)
* Check only for libs via Devel::CheckLib in Makefile.PL (pali)
* Remove param --with-mysql from Makefile.PL documentation (pali)
* Check if specified Makefile.PL settings are working (pali)
Fixes https://rt.cpan.org/Ticket/Display.html?id=119902
* Remember mysql_config value for Makefile.PL (pali)
* Locate MySQL libs using Devel::CheckLib (pali)
* Added a fix for 30insertfetch failure when using server-side prepared
statements. The stmt was being freed for DML statements before
the handle attribute was read, mysql_info was reading back nulls
3.7.0:
Backwards-incompatible changes
* Pool database close_all() method renamed to close_idle() to better
reflect the actual behavior.
* Databases will now raise InterfaceError when connect() or close() are
called on an uninitialized, deferred database object.
New features
* Add methods to the migrations extension to support adding and dropping table
constraints.
* Add Model.bulk_create() method for bulk-inserting unsaved model instances.
* Add close_stale() method to the connection pool to support closing stale
connections.
* The FlaskDB class in playhouse.flask_utils now accepts a model_class
parameter, which can be used to specify a custom base-class for models.
Bugfixes
* Parentheses were not added to subqueries used in function calls with more
than one argument.
* Fixed bug when attempting to serialize many-to-many fields which were created
initially with a DeferredThroughModel.
* Fixed bug when using the Postgres ArrayField with an array of BlobField.
* Allow Proxy databases to be used as a context-manager.
* Fixed bug where the APSW driver was referring to the SQLite version from the
standard library sqlite3 driver, rather than from apsw.
* Reflection library attempts to wrap server-side column defaults in quotation
marks if the column data-type is text/varchar.
* Missing import in migrations library, which would cause errors when
attempting to add indexes whose name exceeded 64 chars.
* When using the Postgres connection pool, ensure any open/pending transactions
are rolled-back when the connection is recycled.
* Even *more* changes to the setup.py script. In this case I've added a
helper function which will reliably determine if the SQLite3 extensions can
be built. This follows the approach taken by the Python YAML package.
6.4.0 release highlights
Analysis
- Option to index phrases on text fields
- Korean analysis tools
- Add multiplexing token filter
Machine learning
- Improve your machine learning results with custom rules
- The {ml} analytics can now detect specific change points in a time
series
Mappings
- alias field type
- _ignored meta field
Rank Eval API
- Expected Reciprocal Rank metric for Rank Eval API
Search
- Cross Cluster Search will no longer use dedicated master nodes as
gateway nodes
- Format option for doc_value fields
- Support second level of field collapse
Security
- Kerberos authentication support is now available
- {es} now offers a FIPS 140-2 compliant mode
6.3.0 release highlights
- SQL: This experimental feature enables users who are familiar with
SQL to use SQL statements to query {es} indices.
- Rollups: This experimental feature enables you to summarize and
store historical data so that is still available for analysis, but
consumes significantly less storage space.
- Java 10 Support
For full releases since 6.2.4 please see
https://www.elastic.co/guide/en/elasticsearch/reference/6.4/release-notes-6.3.1.htmlhttps://www.elastic.co/guide/en/elasticsearch/reference/6.4/release-notes-6.3.2.htmlhttps://www.elastic.co/guide/en/elasticsearch/reference/6.4/release-notes-6.3.0.htmlhttps://www.elastic.co/guide/en/elasticsearch/reference/6.4/release-notes-6.4.0.html
Features
--------
- Parse Virtual Keyspace Metadata
Bug Fixes
---------
- Tokenmap.get_replicas returns the wrong value if token coincides
with the end of the range
- Python Driver fails with "more than 255 arguments" python exception
when > 255 columns specified in query response
- Hang in
integration.standard.test_cluster.ClusterTests.test_set_keyspace_twice
- Asyncore reactors should use a global variable instead of a class
variable for the event loop
Other
-----
- Use global variable for libev loops so it can be subclassed
- Update SchemaParser for V4
- Bump Cython dependency version to 0.28
From Michael Moll via GH #29 ( https://github.com/NetBSD/pkgsrc/pull/29 )
v3.5.5
[CORE-2851] - includeAll tag with a relative path duplicates the database changes with an absolute and with a relative changelog
v3.5.4
[CORE-2851] - includeAll tag with a relative path duplicates the database changes with an absolute and with a relative changelog
[CORE-2863] - Issue with Spring boot 1.4.0 - 1.4.3
[CORE-2898] - includeAll broken in 3.5.1
[CORE-2948] - Changelog with includeAll will not find child changelogs in Spring Boot's executable JAR
[CORE-2978] - AddAutoIncrement on Postgres does not work when no schema is specified
[CORE-3123] - ResourceComparator is not applied for includeAll
[CORE-3139] - ClassLoaderResourceAccessor cannot read jar path resources from SpringLiquibase
[CORE-3015] - Oracle: diffChangeLog TIMESTAMP WITH LOCAL TIME ZONE correctly
1.842 14 July 2018
* Document the MaocOS build failures & how MacPort can work around th eissue.
# 125238
# 125585
# RT 133280
* Write all test db files into a directory created by File::Temp
Protects against any possible races if running a parallel build.
* #96357: DB_File destructor is not thread-safe
Added CLONE_SKIP to DB_File.pm and included the test db-threads.t from the patch.
Note - This only works with perl >= 5.8.7
2.2.7
- Fix warning with flex 2.6+
- Fix compile errors with PG11
- Fix bug in with regex in 'set add sequence' when
specifying the sequence as a regular expression.
It was not being escaped properly
- Add support to Slonik to specify the share direcotry
using the environment variable SLONY_SHARE_DIR
- Add slon config setting remote_listen_serializable_transactions
to use read committed instead of read-only-serializable deferable
transactions(default true)
- Add slon config setting enable_version_check to disable
the slony version check that ensures all nodes run the
same slony version (default true, version check enabled)
Redis 4.0.11:
Upgrade urgency HIGH: not critical but very important bugs fixed.
Dear users, this is just a bugfix release of Redis 4. All new work
is now focused on Redis 5, however we backported a number of bug fixes here:
* The disconnection time between the master and slave was reset in an
incorrect place, sometimes a good slave will not be able to failover
because it claims it was disconnected for too much time from the master.
* A replication bug, rare to trigger but non impossible, is in Redis for
years. It was lately discovered at Redis Labs and fixed by Oran Agra.
It may cause disconnections, desynchronizations and other issues.
* RANDOMKEY may go in infinite loop on rare situations. Now fixed.
* EXISTS now works in a more consistent way on slaves.
* Sentinel: backport of an option to deny a potential security problem
when the SENTINEL command is used to configure an arbitrary script
to execute.
5.7.23:
Functionality Added or Changed
Previously, for the --ssl-mode=VERIFY_IDENTITY or --ssl-verify-server-cert option, the client checked whether the host name that it used for connecting matched the Common Name value in the certificate but not the Subject Alternative Name value. Now, if the client uses OpenSSL 1.0.2 or higher, the client checks whether the host name matches either the Subject Alternative Name value or the Common Name value in the server certificate. Thanks to Daniël van Eeden for a patch on which this change was based.
Bugs Fixed
Important Change; Partitioning: After creating partitioned InnoDB tables with very long names, the table_name columns in the corresponding entries in the mysql.innodb_index_stats and mysql.innodb_table_stats system tables were truncated. To fix this issue, the length of the table_name column in each of these tables has been increased from 64 to 199 characters. In both cases, this is now the same as the lengths of these columns in MySQL 8.0.
InnoDB: The ngram full-text search parser permitted comma and period characters to be tokenized as words, which caused an inconsistency between boolean and natural language mode search results. Comma and period characters are no longer tokenized.
InnoDB: An I/O error returned by an fsync() operation is now treated as a hard error.
InnoDB: A schema mismatch error reported during an import tablespace operation failed to print mismatched table flags in a readable format.
InnoDB: A DDL operation failed to wait for a FULLTEXT index optimization operation to finish.
InnoDB: An unnecessary check for read-only transactions was removed from the trx_set_rw_mode() function. Thanks to Sandeep Sethia for the patch.
InnoDB: A DDL operation that added a foreign key constraint raised an assertion when it accessed a stale memory object that belonged to the parent table.
InnoDB: A DDL operation on a table with a FULLTEXT index during full-text index cache synchronization caused an assertion failure.
InnoDB: A failing assertion occurred after initiating a memcached get operation.
InnoDB: A corrupt index ID encountered during a foreign key check raised an assertion.
InnoDB: An internal deadlock during a DDL operation resulted in a long semaphore wait followed by a server exit.
InnoDB: A DDL operation encountered a serious error due to an invalid lock upgrade.
InnoDB: On a Windows 64-bit system, invalid buffer pool configuration values caused the server to exit on startup.
Partitioning: For a partitioned table, partition update time could be incorrect after rebuilding the table or restarting the server.
Partitioning: An extraneous row lock was imposed by an update to a partitioned InnoDB table.
Replication: The log messages generated when a member fails to join the group have been improved, for example when group_replication_group_name on the member joining the group does not match the seed's group_replication_group_name this is now described in the log message.
Replication: The ER_GRP_RPL_SQL_SERVICE_FAILED_TO_RUN_SQL_QUERY error was being logged incorrectly.
Replication: The use of replication filters or binary log filters can cause issues when they are applied to tables that are updated with XA transactions. Filtering of tables could cause an XA transaction to be empty on a replication slave, and empty XA transactions are not supported. Also, with the settings master_info_repository=TABLE and relay_log_info_repository=TABLE on a replication slave, which became the defaults in MySQL 8.0, the internal state of the data engine transaction is changed following a filtered XA transaction, and can become inconsistent with the replication transaction context state.
Due to these issues, the use of replication filters or binary log filters in combination with XA transactions is not supported. This fix adds the new error ER_XA_REPLICATION_FILTERS, which is logged whenever an XA transaction is impacted by a replication filter, whether or not the transaction was empty as a result. If the transaction is not empty, the replication slave is able to continue running, but you should take steps to discontinue the use of replication filters with XA transactions in order to avoid potential issues. If the transaction is empty, the replication slave stops. In that event, the replication slave might be in an undetermined state in which the consistency of the replication process might be compromised. In particular, the gtid_executed set on a slave of the slave might be inconsistent with that on the master. To resolve this situation, isolate the master and stop all replication, then check GTID consistency across the replication topology. Undo the XA transaction that generated the error message, then restart replication.
Replication: When a transaction larger than the binary log transaction cache size (binlog_cache_size) was flushed to a temporary file during processing, and the flush failed due to a lack of space in the temporary directory, the flush error was not handled correctly. No message was written to the error log, and the binary log cache was not cleared after the transaction was rolled back. Now, in this situation, the server takes an appropriate action based on the binlog_error_action setting (shut down the server or halt logging), and writes a message to the error log. When the transaction is rolled back, the server checks for flush errors and clears the binary log cache if any occurred.
Replication: Using an IP address or hostname in any Group Replication related configuration on macOS was failing.
Replication: When GTIDs are in use for replication, replicated transactions that are filtered out on the slave are persisted. If binary logging is enabled on the slave, the filtered-out transaction is written to the binary log as a Gtid_log_event followed by an empty transaction containing only BEGIN and COMMIT statements. If binary logging is disabled, the GTID of the filtered-out transaction is written to the mysql.gtid_executed table. This process ensures that there are no gaps in the set of executed GTIDs, and that the filtered-out transactions are not retrieved again if the slave reconnects to the master. Previously, this process was not done for CREATE DATABASE, ALTER DATABASE, and DROP DATABASE statements, but it is now carried out for those statements as well as for others.
Replication: On a multithreaded slave, when a STOP SLAVE statement is executed on the slave, followed by a START SLAVE statement, the error log can report a different position in the binary log for the slave SQL thread when exiting, compared to the position reported for the slave SQL thread at the subsequent initialization.
For a multithreaded slave, the position reported for the SQL thread on exit is a low water mark, up to which the replication stream is consistent and has no gaps. Transactions appearing before the position are guaranteed to have committed, but transactions after the position may have committed or not. However, this low water mark was being reported before the process to stop the worker threads was actually carried out, and the low water mark was subsequently updated by a checkpoint routine during that process. The timing of the log message has now been changed so that the final low water mark is reported as the position for the SQL thread on exit.
Replication: In certain situations, such as during distributed recovery procedure, the certification info garbage collection was purging more data than it should, resulting in conflicts not being detected. The garbage collection procedure has been improved to take this case in consideration.
Replication: When the group_replication_applier channel's applier thread encountered an error, the master_log_name and end_log_pos in the error message were incorrect. In Group Replication, the events of a transaction are replicated before they are written to the binary log of the member where the transaction originated. The result is that the final master_log_name and the end_log_pos of those events are unknown at the time they are applied on the replica by group_replication_applier channel's applier thread. To avoid confusion, now any such error messages encountered by a group_replication_applier channel do not contain the binary log name and the binary log position.
A heap overflow vulnerability in the MySQL client library was fixed.
For generated columns that used the INTERVAL() function, incorrect behavior could occur.
The exec_in_background command for mysqltest is now available in MySQL 5.7.
An unencrypted connection could result from a client connection attempt specifying that an encrypted connection was required, if the server was not configured to support SSL.
A BETWEEN clause comparing negative values could lead to erroneous results.
Audit log filter rules did not permit class names to be specified as an array of strings.
shared-mime-info 1.10 (2018-28-06)
* Add mime-type for MJPEG raw streams
* Add MPEG-4 Ringtone mime-type
* Add mime-type for HEIF images
* Add application/vnd.sqlite3, application/x-sqlite2 IANA types
* Add application/pkcs8-encrypted
* Add audio/usac
* Add application/vnd.youtube.yt
* Add mime-types for raw floppy disk images
* Add aliases for SAP and HFE floppy images mime-types
* Use new application/vnd.comicbook-rar IANA type
* Remove magic from STL binary
* Add *.adts and *.ass globs from IANA to audio/aac
* Use new IANA type for application/xliff+xml
* Use new IANA type for model/stl
* Add text/x-perl alias to application/x-perl
* Bump priority for our DjVu mime-types
* Differentiate Neo Geo Pocket Color ROMs from non-Color ones
* Remove *.bin from application/octet-stream
* Add an application/xps alias for application/oxps
* Change Qt Linguist file type to text/vnd.qt.linguist
* Fix the QML magic
Tools:
- Don't warn with unknown top level type names
- Look in obsolete KDE sub-dirs for clean up opportunities
- Add some debug when parsing XML package files
Spec:
- ASCII check should be done with 128 leading bytes, not 32 bytes
Version 2.4:
- Add standard and retry hook scripts for backup deletion (pre/post)
- Add standard and retry hook scripts for recovery (pre/post)
- Add standard and retry hook scripts for WAL deletion (pre/post)
- Add --standby-mode option to barman recover to add standby_mode = on
in pre-generated recovery.conf
- Add --target-action option to barman recover, allowing users to add
shutdown, pause or promote to the pre-generated recovery.conf file
- Improve usability of point-in-time recovery with consistency checks
(e.g. recovery time is after end time of backup)
- Minor documentation improvements
- Drop support for Python 3.3
Relevant bug fixes:
- Fix remote get_file_content method, preventing
incremental recovery from happening
- Unicode issues with command
- Add --wal-method=none when pg_basebackup >= 10
Minor bug fixes:
- Stop process manager module from ovewriting lock files content
- Relax the rules for rsync output parsing
- Ignore vanished files in streaming directory
- Case insensitive slot names
- Make DataTransferFailure.from_command_error() more resilient
- Rename command() to barman_command()
- Initialise synchronous standby names list if not set
- Correct placeholders ordering
- Force datestyle to iso for replication connections
- Returns error if delete command does not remove the backup
- Fix exception when calling is_power_of_two(None)
- Downgraded sync standby names messages to debug
Version 2.3:
- Add support to PostgreSQL 10
- Follow naming changes in PostgreSQL 10:
- The switch-xlog command has been renamed to switch-wal.
- In commands output, the xlog word has been changed to WAL and
location has been changed to LSN when appropriate.
- Add the --network-compression/--no-network-compression options to
barman recover to enable or disable network compression at run-time
- Add --target-immediate option to recover command, in order to exit
recovery when a consistent state is reached (end of the backup,
available from PostgreSQL 9.4)
- Show cluster state (master or standby) with barman status command
- Documentation improvements
- Bug fixes:
- Fix high memory usage with parallel_jobs > 1
- Better handling of errors using parallel copy
- Make barman diagnose more robust with system exceptions
- Let archive-wal ignore files with .tmp extension
Version 2.2:
- Implement parallel copy for backup/recovery through the
parallel_jobs global/server option to be overridden by the --jobs or
-j runtime option for the backup and recover command. Parallel
backup is available only for the rsync copy method. By default, it
is set to 1 (for behaviour compatibility with previous versions).
- Support custom WAL size for PostgreSQL 8.4 and newer. At backup
time, Barman retrieves from PostgreSQL wal_segment_size and
wal_block_size values and computes the necessary calculations.
- Improve check command to ensure that incoming directory is empty
when archiver=off, and streaming directory is empty when
streaming_archiver=off
- Add external_configuration to backup_options so that users can
instruct Barman to ignore backup of configuration files when they
are not inside PGDATA (default for Debian/Ubuntu installations). In
this case, Barman does not display a warning anymore.
- Add --get-wal and --no-get-wal options to barman recover
- Add max_incoming_wals_queue global/server option for the check
command so that a non blocking error is returned in case incoming
WAL directories for both archiver and the streaming_archiver contain
more files than the specified value.
- Documentation improvements
- File format changes:
- The format of backup.info file has changed. For this reason a
backup taken with Barman 2.2 cannot be read by a previous
version of Barman. But, backups taken by previous versions can
be read by Barman 2.2.
- Minor bug fixes:
- Allow replication-status to work against a standby
- Close any PostgreSQL connection before starting pg_basebackup
- Safely handle paths containing special characters
- Archive .partial files after promotion of streaming source
- Recursively create directories during recovery
- Improve xlog.db locking
- Remove tablespace_map file during recover
- Reconnect to PostgreSQL if connection drops
Version 2.1:
- Add --archive and --archive-timeout options to switch-xlog command
- Preliminary support for PostgreSQL 10
- Minor additions:
- Add last archived WAL info to diagnose output
- Add start time and execution time to the output of delete
command
- Minor bug fixes:
- Return failure for get-wal command on inactive server
- Make streaming_archiver_names and streaming_backup_name options
global
- Fix rsync failures due to files truncated during transfer
- Correctly handle compressed history files
- Avoid de-referencing symlinks in pg_tblspc when preparing
recovery
- Fix comparison of last archiving failure
- Avoid failing recovery if postgresql.conf is not writable
- Fix output of replication-status command
- Exclude files from backups like pg_basebackup
- Exclude directories from other Postgres versions while copying
tablespaces
- Make retry hook script options global
Version 2.0 - 27 Sep 2016
- Support for pg_basebackup and base backups over the PostgreSQL
streaming replication protocol with backup_method=postgres
(PostgreSQL 9.1 or higher required)
- Support for physical replication slots through the slot_name
configuration option as well as the --create-slot and --drop-slot
options for the receive-wal command (PostgreSQL 9.4 or higher
required). When slot_name is specified and streaming_archiver is
enabled, receive-wal transparently integrates with pg_receivexlog,
and check makes sure that slots exist and are actively used
- Support for the new backup API introduced in PostgreSQL 9.6, which
transparently enables concurrent backups and backups from standby
servers using the standard rsync method of backup. Concurrent backup
was only possible for PostgreSQL 9.2 to 9.5 versions through the
pgespresso extension. The new backup API will make pgespresso
redundant in the future
- If properly configured, Barman can function as a synchronous standby
in terms of WAL streaming. By properly setting the
streaming_archiver_name in the synchronous_standby_names priority
list on the master, and enabling replication slot support, the
receive-wal command can now be part of a PostgreSQL synchronous
replication cluster, bringing RPO=0 (PostgreSQL 9.5.5 or
higher required)
- Introduce barman-wal-restore, a standard and robust script written
in Python that can be used as restore_command in recovery.conf files
of any standby server of a cluster. It supports remote parallel
fetching of WAL files by efficiently invoking get-wal through SSH.
Currently available as a separate project called barman-cli. The
barman-cli package is required for remote recovery when get-wal is
listed in recovery_options
- Control the maximum execution time of the check command through the
check_timeout global/server configuration option (30 seconds
by default)
- Limit the number of WAL segments that are processed by an
archive-wal run, through the archiver_batch_size and
streaming_archiver_batch_size global/server options which control
archiving of WAL segments coming from, respectively, the standard
archiver and receive-wal
- Removed locking of the XLOG database during check operations
- The show-backup command is now aware of timelines and properly
displays which timelines can be used as recovery targets for a given
base backup. Internally, Barman is now capable of parsing .history
files
- Improved the logic behind the retry mechanism when copy operations
experience problems. This involves backup (rsync and postgres) as
well as remote recovery (rsync)
- Code refactoring involving remote command and physical copy
interfaces
- Bug fixes:
- Correctly handle .history files from streaming
- Fix replication-status on PostgreSQL 9.1
- Fix replication-status when sent and write locations are not
available
- Fix misleading message on pg_receivexlog termination
