- #3159, do not force a bbox cache on ST_Affine
- #3018, GROUP BY geography sometimes returns duplicate rows
- #3048, shp2pgsql - illegal number format when specific system locale set
- #3094, Malformed GeoJSON inputs crash backend
- #3104, st_asgml introduces random characters in ID field
- #3155, Remove liblwgeom.h on make uninstall
- #3177, gserialized_is_empty cannot handle nested empty cases
- Fix crash in ST_LineLocatePoint
elasticsearch 1.6.0
Breaking changes
- Benchmark api: removed leftovers
- Wildcard field names in highlighting should only return fields that
can be highlighted
- Remove unsafe options
- Fix FSRepository location configuration
Deprecations
- Deprecate async replication
- Query DSL: deprecate BytesFilterBuilder in favour of WrapperFilterBuilder
- Deprecate async replication
- Deprecate the More-Like-This API in favour of the MLT query
- Deprecate rivers
- Warning in documentation for deprecation of rivers
- Deprecated the thrift and memcached transports
- Deprecate the top_children query
- Plugins: deprecate addQuery methods that are going to be removed in 2.0
- Deprecate Groovy sandbox and related settings
- Deprecate delete-by-query in client/transport/action APIs too
- Deprecate filter option in PhraseSuggester collate
New features
- Add ability to specify a SizeBasedTriggeringPolicy for log configuration
- Bring back numeric_resolution
- API: Add response filtering with filter_path parameter
- Synced flush backport
- Move index sealing terminology to synced flush
- Seal indices for faster recovery
- Add support for fine-grained settings
- Validate API: provide more verbose explanation
- Add ability to prompt for selected settings on startup
- bootstrap.mlockall for Windows (VirtualLock)
- Allow shards on shared filesystems to be recovered on any node
- Add field stats api
For a full changelog see here:
https://www.elastic.co/downloads/past-releases/elasticsearch-1-6-0
elasticsearch 1.5.2
Security
- Ensure URL expansion only works within the plugins directory
Enhancements
- Only flush for checkindex if we have uncommitted changes
- Update tree_level and precision parameter priorities
- Add merge conflicts to GeoShapeFieldMapper
- pom.xml updates to allow m2e integration to work correctly
- Fix to pom.xml to allow eclipse maven integration using m2e
- Eclipse fixes
- Implement retries for ShadowEngine creation
- Allow rebalancing primary shards on shared filesystems
- Update forbiddenapis to version 1.8
Bug fixes
- Fix _as_string output to only show when format specified
- _default_ mapping should be picked up from index template during
auto create index
- Correct ShapeBuilder coordinate parser to ignore values in 3rd+ dimension
- Fix hole intersection at tangential coordinate
- Search: FielddataTermsFilter equality is based on hash codes
- Fix possible NPE in InternalClusterService$NotifyTimeout, the future
field is set from a different thread
- Add missing hashCode method to RecoveryState#File
- Make GeoContext mapping idempotent
- Fixed an equality check in StringFieldMapper.
- Score value is 0 in _explanation with random_score query
- Fix updating templates.
- Analysis: fix ignoring tokenizer settings in SynonymTokenFilterFactory
- ShardTermVectorsService calls docFreq() on unpositioned TermsEnum
- FSTranslog#snapshot() can enter infinite loop
elasticsearch 1.5.1
Deprecations
- Warning in documentation for deprecation of rivers
Enhancements
- Core: also refresh if many deletes in a row use up too much
version map RAM
- Use static logger name in Engine.java
- service.bat file should explicitly use the Windows find command.
- AbstractBlobContainer.deleteByPrefix() should not list all blobs
Bug fixes
- Core: Lucene merges should run on the target shard during recovery
- Sync translog before closing engine
- Fix validate_* merge policy for GeoPointFieldMapper
- Make sure size=0 works on the inner_hits level.
- Make sure inner hits also work for nested fields defined in object field
- Fix bug where parse error is thrown if a inner filter is used in
a nested filter/query.
- Fix nested stored field support.
- Bugfix+unittest for unneccesary mapping refreshes caused by unordered
fielddata settings
- Don't try to send a mapping refresh if there is no master
- Fix _field_names to be disabled on pre 1.3.0 indexes
- Transport: fix racing condition in timeout handling
- Children aggregation: Fix 2 bugs in children agg
- Fix wrong use of currentFieldName outside of a parsing loop
- Avoid NPE during query parsing
- Function score: apply min_score to sub query score if
no function provided
- Function_score: undo "Remove explanation of query score from functions"
- State: Refactor state format to use incremental state IDs
- Recovery: RecoveryState.File.toXContent reports file length
as recovered bytes
- Fail shard when index service/mappings fails to instantiate
- Don't reuse source index UUID on restore
- Snapshot/Restore: separate repository registration
- Automatically add "index." prefix to the settings are changed
on restore...
elasticsearch 1.5.0
Breaking changes
- Aliases: Throw exception if index is null or missing when creating
an alias
- Benchmark api: removed leftovers
- Resiliency: Throw exception if the JVM will corrupt data
- [ENGINE] Remove full flush / FlushType.NEW_WRITER
- Change behaviour of indices segments api to allow no indices
- Plugins: Don't overwrite plugin configuration when removing/upgrading
plugins
- [QUERY] Remove lowercase_expanded_terms and locale options
- Recovery: RecoveryState clean up
- Scripting: cleanup ScriptService & friends
- Disable dynamic Groovy scripting by marking Groovy as not sandboxed
- Scripting: Script with _score: remove dependency of DocLookup and scorer
Deprecations
- Deprecate async replication
- Core: deprecate index.fail_on_merge_failure
- Mappings: Deprecate _analyzer and per query analyzers
- Deprecation: MLT Field Query
- Deprecated the thrift and memcached transports
- Core: add deprecation messages for delete-by-query
New features
- New aggregations feature - "PercentageScore" heuristic
for significant_terms
- significant terms: add scriptable significance heuristic
- Cat API: show open and closed indices in _cat/indices
- Circuit Breakers: Add NoopCircuitBreaker used in NoneCircuitBreakerService
- Shadow replicas on shared filesystems
- MLT Query: Support for artificial documents
- Add time_zone setting for query_string
- Search: add format support for date range filter and queries
- Add min_score parameter to function score query to only match docs
above this threshold
- Add inner hits to nested and parent/child queries
- Add index.data_path setting
- Term Vectors/MLT Query: support for different analyzers than default
at the field
For a full changelog see here:
https://www.elastic.co/downloads/past-releases/elasticsearch-1-5-0
1.1.9
* This release fixes a common crash in 1.1.8, which itself was introduced
while fixing a rare crash in 1.1.7
1.1.8
* Crash freeing client after a replica set auth error.
* Compile error strict C89 mode.
0.9.6
-----
* Fix ibmdb2 index name handling
0.9.5
-----
* Don't run the test if _setup() fails
* Correcting minor typo
* Fix .gitignore for .tox and .testrepository
* allow dropping fkeys with sqlite
* Add pretty_tox setup
* script: strip comments in SQL statements
0.9.4
-----
* Remove svn version tag setting
0.9.3
-----
* Ignore transaction management statements in SQL scripts
* Use native sqlalchemy 0.9 quote attribute with ibmdb2
* Don't add warnings filter on import
* Replace assertNotEquals with assertNotEqual
* Update requirements file matching global requ
* Work toward Python 3.4 support and testing
* pep8: mark all pep8 checks that currently fail as ignored
0.9.2
-----
* SqlScript: execute multiple statements one by one
* Make sure we don't throw away exception on SQL script failure
* Pin testtools to < 0.9.36
* Fix ibmdb2 unique constraint handling for sqlalchemy 0.9
* Fixes the auto-generated manage.py
What's new in psycopg 2.6.1
^^^^^^^^^^^^^^^^^^^^^^^^^^^
- Lists consisting of only `None` are escaped correctly (🎫`#285`).
- Fixed deadlock in multithread programs using OpenSSL (🎫`#290`).
- Correctly unlock the connection after error in flush (🎫`#294`).
- Fixed ``MinTimeLoggingCursor.callproc()`` (🎫`#309`).
* Fix WINDOW and HAVING params order in Select
* Add window functions
* Add filter and within group to aggregate
* Add limitstyle with 'offset' and 'limit'
* Add Lateral
Changelog:
MongoDB 3.0.4 is released
June 6, 2015
MongoDB 3.0.4 is out and is ready for production deployment. This release contains only fixes since 3.0.3, and is a recommended upgrade for all 3.0 users.
Fixed in this release:
SERVER-17923 Creating/dropping multiple background indexes on the same collection can cause fatal error on secondaries
SERVER-18079 Large performance drop with documents > 16k on Windows
SERVER-18190 Secondary reads block replication
SERVER-18213 Lots of WriteConflict during multi-upsert with WiredTiger storage engine
SERVER-18316 Database with WT engine fails to recover after system crash
SERVER-18475 authSchemaUpgrade fails when the system.users contains non MONGODB-CR users
SERVER-18629 WiredTiger journal system syncs wrong directory
SERVER-18822 Sharded clusters with WiredTiger primaries may lose writes during chunk migration
Announcing MongoDB 3.0 and Bug Hunt Winners
March 3, 2015
Today MongoDB 3.0 is generally available; you can download now.
Our community was critical to ensuring the quality of the release. Thank you to everyone who participated in our 3.0 Bug Hunt. From the submissions, we've selected winners based on the user impact and severity of the bugs found.
First Prize
Mark Callaghan, Member of Technical Staff, Facebook
During the 3.0 release cycle, Mark submitted 10 bug reports and collaborated closely with the MongoDB engineering team to debug the issues he uncovered. As a first place winner, Mark will receive a free pass to MongoDB World in New York City on June 1-2, including a front row seat to the keynote sessions. Mark was also eligible to receive a $1,000 Amazon gift card but opted to donate the award to a charity. We are donating $1,000 to CodeStarters.org in his name.
Honorable Mentions
Nick Judson, Conevity
Nick submitted SERVER-17299, uncovering excessive memory allocation on Windows when using "snappy" compression in WiredTiger.
Koshelyaev Konstantin, RTEC
Koshelyaev submitted SERVER-16664, which uncovered a memory overflow in WiredTiger when using "zlib" compression.
Tim Callaghan, Crunchtime!
In submitting SERVER-16867, Tim found an uncaught WriteConflict exception affecting replicated writes during insert-heavy workloads.
Nathan Arthur, PreEmptive Solutions
Nathan submitted SERVER-16724, which found an issue with how collection metadata is persisted.
Thijs Cadier, AppSignal
Thijs submitted SERVER-16197, which revealed a bug in the build system interaction with the new MongoDB tools.
Nick, Koshelyaev, Tim, Nathan, and Thijs will also receive tickets to MongoDB World in New York City on June 1-2 (with reserved front-row seat for keynote sessions), $250 Amazon Gift Cards, and MongoDB t-shirts.
Congratulations to the winners and thanks to everyone who downloaded, tested and gave feedback on the release candidates.
This release primarily fixes issues not successfully fixed in prior releases. It should be applied as soon as possible all users of major versions 9.3 and 9.4. Other users should apply at the next available downtime.
Crash Recovery Fixes:
Earlier update releases attempted to fix an issue in PostgreSQL 9.3 and 9.4 with "multixact wraparound", but failed to account for issues doing multixact cleanup during crash recovery. This could cause servers to be unable to restart after a crash. As such, all users of 9.3 and 9.4 should apply this update as soon as possible.
(Because of the partitioning into client and server packages, the man
pages have to be partitioned to match; this interferes with the
configure script's handling of them so the list of pages ends up
hardcoded in these patches. And it seems the lists haven't been
updated since the first mysql 5.x package.)
1.48 2015-06-12
- Switched to a production version. (ISHIGAKI)
1.47_05 2015-05-08
- Updated to SQLite 3.8.10
1.47_04 2015-05-02
- Used MY_CXT instead of a global variable
1.47_03 2015-04-16
- Added :all to EXPORT_TAGS in ::Constants
1.47_02 2015-04-16
- Updated to SQLite 3.8.9
- Added DBD::SQLite::Constants, from which you can import any
"useful" constants into your applications.
- Removed previous Cygwin hack as SQLite 3.8.9 compiles well again
- Now create_function/aggregate accepts an extra bit
(SQLITE_DETERMINISTIC) for better performance.
1.47_01 2015-02-17
*** (EXPERIMENTAL) CHANGES THAT MAY POSSIBLY BREAK YOUR OLD APPLICATIONS ***
- Commented OPTIMIZE out of WriteMakefile (RT #94207).
If your perl is not compiled with -O2, your DBD::SQLite may
possibly behave differently under some circumstances.
(This release is to find notable examples from CPAN Testers).
- Set THREADSAFE to 0 under Cygwin to cope with an upstream
regression since 3.8.7 (GH #7).
- Updated to SQLite 3.8.8.2
- Resolved#35449: Fast DBH->do (ptushnik, ISHIGAKI)
2.6.1
- #606, support self-referential joins with prefetch and aggregate_rows()
methods.
- #588, accomodate changes in SQLite's PRAGMA index_list() return value.
- #607, fixed bug where pwiz was not passing table names to introspector.
- #591, fixed bug with handling of named cursors in older psycopg2 version.
- Removed some cruft from the APSWDatabase implementation.
- Added CompressedField and AESEncryptedField
- #609, #610, added Django-style foreign key ID lookup.
- Added support for Hybrid Attributes (cool idea courtesy of SQLAlchemy).
- Added upsert keyword argument to the Model.save() function (SQLite only).
- #587, added support for ON CONFLICT SQLite clause for INSERT and UPDATE
queries.
- #601, added hook for programmatically defining table names.
- #581, #611, support connection pools with playhouse.db_url.connect().
- Added Contributing section section to docs.
2.6.0
- get_or_create() now returns a 2-tuple consisting of the model instance
and a boolean indicating whether the instance was created. The function
now behaves just like the Django equivalent.
- #574, better support for setting the character encoding on Postgresql
database connections. Thanks @klen!
- Improved implementation of get_or_create().
mongo-c-driver 1.1.7
- Thread-safe use of Cyrus SASL library.
- Experimental support for building with CMake and SASL.
- Faster reconnection to replica set with some hosts down.
- Crash iterating a cursor after reconnecting to a replica set.
- Unchecked errors decoding invalid UTF-8 in MongoDB URIs.
- Fix error reporting from mongoc_client_get_database_names.
mongo-c-driver 1.1.6
- mongoc_bulk_operation_execute now coalesces consecutive update operations
into a single message to a MongoDB 2.6+ server, yielding huge performance
gains. Same for remove operations. (Inserts were always coalesced.)
- Large numbers of insert operations are now properly batched according
to number of documents and total data size.
- GSSAPI / Kerberos auth now works.
- The driver no longer tries three times in vain to reconnect to a primary,
so socketTimeoutMS and connectTimeoutMS now behave closer to what you
expect for replica sets with down members. A full fix awaits 1.2.0.
- mongoc_matcher_t now supports basic subdocument and array matching
mongo-c-driver 1.1.5
- The fsync and j write concern flags now imply acknowledged writes
- Prevent using fsync or j with conflicting w=0 write concern
- Obey socket timeout consistently in TLS/SSL mode
- Return an error promptly after a network hangup in TLS mode
- Prevent crash using SSL in FIPS mode
- Always return NULL from mongoc_database_get_collection_names on error
- Fix version check for GCC 5 and future versions of Clang
- Fix warnings and errors building on various platforms
- Add configure flag to enable/disable shared memory performance counters
- Minor docs improvements and fix links from C Driver docs to Libbson docs
Version 1.4.1 - 05 May 2015
* Fix for WAL archival stop working if first backup is EMPTY
(Closes: #64)
* Fix exception during error handling in Barman recovery (Closes:
#65)
* After a backup, limit cron activity to WAL archiving only
(Closes: #62)
* Improved robustness and error reporting of the backup delete
command (Closes: #63)
* Fix computation of WAL production ratio as reported in the
show-backup command
* Improved management of xlogb file, which is now correctly fsynced
when updated. Also, the rebuild-xlogdb command now operates on a
temporary new file, which overwrites the main one when finished.
* Add unit tests for dateutil module compatibility
* Modified Barman version following PEP 440 rules and added support
of tests in Python 3.4
--[ Redis 3.0.2 ] Release date: 4 Jun 2015
Upgrade urgency: HIGH for Redis because of a security issue.
LOW for Sentinel.
* [FIX] Critical security issue fix by Ben Murphy: http://t.co/LpGTyZmfS7
* [FIX] SMOVE reply fixed when src and dst keys are the same. (Glenn Nethercutt)
* [FIX] Lua cmsgpack lib updated to support str8 type. (Sebastian Waisbrot)
* [NEW] ZADD support for options: NX, XX, CH. See new doc at redis.io.
(Salvatore Sanfilippo)
* [NEW] Senitnel: CKQUORUM and FLUSHCONFIG commands back ported.
(Salvatore Sanfilippo and Bill Anderson)
--[ Redis 3.0.1 ] Release date: 5 May 2015
Upgrade urgency: LOW for Redis and Cluster, MODERATE for Sentinel.
* [FIX] Sentinel memory leak due to hiredis fixed. (Salvatore Sanfilippo)
* [FIX] Sentinel memory leak on duplicated instance. (Charsyam)
* [FIX] Redis crash on Lua reaching output buffer limits. (Yossi Gottlieb)
* [FIX] Sentinel flushes config on +slave events. (Bill Anderson)
* File Permissions Fix
* Have pg_get_functiondef() show the LEAKPROOF property
* Make pushJsonbValue() function push jbvBinary type
* Allow building with threaded Python on OpenBSD
- Avoid failures while fsync'ing data directory during crash restart
- Fix pg_get_functiondef() to show functions' LEAKPROOF property, if set
- Remove configure's check prohibiting linking to a threaded libpython
on OpenBSD
- Allow libpq to use TLS protocol versions beyond v1
Upstream changes are mainly housekeeping and minor build system
changes not visible to pkgsrc users, plus the usual bugfixes. Some
procedures previously advertised for deprecation have been dropped,
and some new ones are added to the deprectation list, notably
dbcoltypes.
### 0.13.1 - May 03, 2015
This is a bug fix release.
The new `reconnect` method introduced new struct members, which clashed with pre-defined names in pre-C99 code.
Another commit forced C99 compilation just to make it work, but of course this is not desirable for outside projects.
Other non-C99 code can now use hiredis as usual again.
Sorry for the inconvenience.
* Fix memory leak in async reply handling (Salvatore Sanfilippo)
* Rename struct member to avoid name clash with pre-c99 code (Alex Balashov, ncopa)
### 0.13.0 - April 16, 2015
This release adds a minimal Windows compatibility layer.
The parser, standalone since v0.12.0, can now be compiled on Windows
(and thus used in other client libraries as well)
* Windows compatibility layer for parser code (tzickel)
* Properly escape data printed to PKGCONF file (Dan Skorupski)
* Fix tests when assert() undefined (Keith Bennett, Matt Stancliff)
* Implement a reconnect method for the client context, this changes the structure of `redisContext` (Aaron Bedra)
### 0.12.1 - January 26, 2015
* Fix `make install`: DESTDIR support, install all required files, install PKGCONF in proper location
* Fix `make test` as 32 bit build on 64 bit platform
### 0.12.0 - January 22, 2015
* Add optional KeepAlive support
* Try again on EINTR errors
* Add libuv adapter
* Add IPv6 support
* Remove possiblity of multiple close on same fd
* Add ability to bind source address on connect
* Add redisConnectFd() and redisFreeKeepFd()
* Fix getaddrinfo() memory leak
* Free string if it is unused (fixes memory leak)
* Improve redisAppendCommandArgv performance 2.5x
* Add support for SO_REUSEADDR
* Fix redisvFormatCommand format parsing
* Add GLib 2.0 adapter
* Refactor reading code into read.c
* Fix errno error buffers to not clobber errors
* Generate pkgconf during build
* Silence _BSD_SOURCE warnings
* Improve digit counting for multibulk creation
=== 4.23.0 (2015-06-01)
* Make dataset.call_sproc(:insert) work in the jdbc adapter (flash-gordon) (#1013)
* Add update_refresh plugin, for refreshing a model instance when updating (jeremyevans)
* Add delay_add_association plugin, for delaying add_* method calls on new objects until after saving the object (jeremyevans)
* Add validate_associated plugin, for validating associated objects when validating the current object (jeremyevans)
* Make Postgres::JSONBOp#[] and #get_text return JSONBOp instances (jeremyevans) (#1005)
* Remove the fdbsql, jdbc/fdbsql, and openbase adapters (jeremyevans)
* Database#transaction now returns block return value if :rollback=>:always is used (jeremyevans)
* Allow postgresql:// connection strings as aliases to postgres://, for compatibility with libpq (jeremyevans) (#1004)
* Make Model#move_to in the list plugin handle out-of-range targets without raising an exception (jeremyevans) (#1003)
* Make Database#add_named_conversion_proc on PostgreSQL handle conversion procs for enum types (celsworth) (#1002)
=== 4.22.0 (2015-05-01)
* Deprecate the db2, dbi, fdbsql, firebird, jdbc/fdbsql, informix, and openbase adapters (jeremyevans)
* Avoid hash allocations and rehashes (jeremyevans)
* Don't silently ignore :jdbc_properties Database option in jdbc adapter (jeremyevans)
* Make tree plugin set reciprocal association for children association correctly (lpil, jeremyevans) (#995)
* Add Sequel::MassAssignmentRestriction exception, raised for mass assignment errors in strict mode (jeremyevans) (#994)
* Handle ODBC::SQL_BIT type as boolean in the odbc adapter, fixing boolean handling on odbc/mssql (jrgns) (#993)
* Make :auto_validations plugin check :default entry instead of :ruby_default entry for checking existence of default value (jeremyevans) (#990)
* Adapters should now set :default schema option to nil when adapter can determine that the value is nil (jeremyevans)
* Do not add a schema :max_length entry for a varchar(max) column on MSSQL (jeremyevans)
* Allow :default value for PostgreSQL array columns to be a ruby array when using the pg_array extension (jeremyevans) (#989)
* Add csv_serializer plugin for serializing model objects to and from csv (bjmllr, jeremyevans) (#988)
* Make Dataset#to_hash and #to_hash_groups handle single array argument for model datasets (jeremyevans)
* Handle Model#cancel_action in association before hooks (jeremyevans)
* Use a condition variable instead of busy waiting in the threaded connection pools on ruby 1.9+ (jeremyevans)
* Use Symbol#to_proc instead of explicit blocks (jeremyevans)
=== 4.21.0 (2015-04-01)
* Support :tsquery and :tsvector options in Dataset#full_text_search on PostgreSQL, for using existing tsquery/tsvector expressions (jeremyevans)
* Fix TinyTds::Error being raised when trying to cancel a query on a closed connection in the tinytds adapter (jeremyevans)
* Add GenericExpression#!~ for inverting =~ on ruby 1.9 (similar to inverting a hash) (jeremyevans) (#979)
* Add GenericExpression#=~ for equality, inclusion, and pattern matching (similar to using a hash) (jeremyevans) (#979)
* Add Database#add_named_conversion_proc on PostgreSQL to make it easier to add conversion procs for types by name (jeremyevans)
* Make Sequel.pg_jsonb return JSONBOp instances instead of JSONOp instances when passed other than Array or Hash (jeremyevans) (#977)
* Demodulize default root name in json_serializer plugin (janko-m) (#968)
* Make Database#transaction work in after_commit/after_rollback blocks (jeremyevans)
== v0.18.2 [2015-05-14] Michael Granger <ged@FaerieMUD.org>
Enhancements:
- Allow URI connection string (thanks to Chris Bandy)
Bugfixes:
- Speedups and fixes for PG::TextDecoder::Identifier and quoting behavior
- Revert addition of PG::Connection#hostaddr [#202].
- Fix decoding of fractional timezones and timestamps [#203]
- Fixes for non-C99 compilers
- Avoid possible symbol name clash when linking againt static libpq.
Functionality Added or Changed
* MySQL Enterprise Firewall operates on parser states and does not work well together with the query cache, which circumvents the parser. MySQL Enterprise Firewall now checks whether the query cache is enabled. If so, it displays a message that the query cache must be disabled and does not load.
* my_print_defaults now masks passwords. To display passwords in cleartext, use the new --show option.
* MySQL distributions now include an innodb_stress suite of test cases. Thanks to Mark Callaghan for the contribution.
Bugs Fixed
* InnoDB; Partitioning: The CREATE_TIME column of the INFORMATION_SCHEMA.TABLES table now shows the correct table creation time for partitioned InnoDB tables. The CREATE_TIME column of the INFORMATION_SCHEMA.PARTITIONS table now shows the correct partition creation time for a partition of partitioned InnoDB tables.
The UPDATE_TIME column of the INFORMATION_SCHEMA.TABLES table now shows when a partitioned InnoDB table was last updated by an INSERT, DELETE, or UPDATE. The UPDATE_TIME column of the INFORMATION_SCHEMA.PARTITIONS table now shows when a partition of a partitioned InnoDB table was last updated.
* InnoDB: An assertion was raised on shutdown due to XA PREPARE transactions holding explicit locks.
* InnoDB: The strict_* forms of innodb_checksum_algorithm settings (strict_none, strict_innodb, and strict_crc32) caused the server to halt when a non-matching checksum was encountered, even though the non-matching checksum was valid. For example, with innodb_checksum_algorithm=strict_crc32, encountering a valid innodb checksum caused the server to halt. Instead of halting the server, a message is now printed to the error log and the page is accepted as valid if it matches an innodb, crc32 or none checksum.
* InnoDB: The memcached set command permitted a negative expire time value. Expire time is stored internally as an unsigned integer. A negative value would be converted to a large number and accepted. The maximum expire time value is now restricted to INT_MAX32 to prevent negative expire time values.
* InnoDB: Removal of a foreign key object from the data dictionary cache during error handling caused the server to exit.
* InnoDB: SHOW ENGINE INNODB STATUS output showed negative reservation and signal count values due to a counter overflow error.
* InnoDB: Failure to check the status of a cursor transaction read-only option before reusing the cursor transaction for a write operation resulted in a server exit during a memcached workload.
* InnoDB: MDL locks taken by memcached clients caused a MySQL Enterprise Backup FLUSH TABLES WITH READ LOCK operation to hang.
* InnoDB: Estimates that were too low for the size of merge chunks in the result sorting algorithm caused a server exit.
* InnoDB: For full-text searches, the optimizer could choose an index that does not produce correct relevancy rankings.
* Partitioning: When creating a partitioned table, partition-level DATA DIRECTORY or INDEX DIRECTORY option values that contained an excessive number of characters were handled incorrectly.
* Partitioning: Executing an ALTER TABLE on a partitioned table on which a write lock was in effect could cause subsequent SQL statements on this table to fail.
* Replication: When binary logging was enabled, using stored functions and triggers resulting in a long running procedure that inserted many records caused the memory use to increase rapidly. This was due to memory being allocated per variable. The fix ensures that in such a situation, memory is allocated once and the same memory is reused.
* Replication: If an error was encountered while adding a GTID to the received GTID set, the log lock was not being correctly released. This could cause a deadlock.
more...
