Referential integrity constraints sometimes used the wrong index, such that updating a row in the referenced table incorrectly failed with a constraint violation.
H2 Console: auto-complete is now triggered by the tab key.
@changelog_1003_li
The Polish translation was completed and corrected by Wojtek Jurczyk. Thanks a lot!
H2 Console: auto-complete did not work with multi-line statements.
@changelog_1004_li
Issue 545: Unnecessary duplicate code was removed.
CLOB and BLOB data was not immediately removed after a rollback.
@changelog_1005_li
The profiler tool can now process files with full thread dumps.
There is a new Aggregate API that supports the internal H2 data types (GEOMETRY for example). Thanks a lot to Nicolas Fortin for the patch!
@changelog_1006_li
MVStore: the file format was changed slightly.
Referential integrity constraints sometimes used the wrong index, such that updating a row in the referenced table incorrectly failed with a constraint violation.
@changelog_1007_li
MVStore mode: the CLOB and BLOB storage was re-implemented and is now much faster than with the PageStore (which is still the default storage).
The Polish translation was completed and corrected by Wojtek Jurczyk. Thanks a lot!
@changelog_1008_li
MVStore mode: creating indexes is now much faster (in many cases faster than with the default PageStore).
Issue 545: Unnecessary duplicate code was removed.
@changelog_1009_li
Various bugs in the MVStore storage and have been fixed, including a bug in the R-tree implementation. The database could get corrupt if there were transient IO exceptions while storing.
The profiler tool can now process files with full thread dumps.
@changelog_1010_li
The method org.h2.expression.Function.getCost could throw a NullPointException.
MVStore: the file format was changed slightly.
@changelog_1011_li
Storing LOBs in separate files (outside of the main database file) is no longer supported for new databases.
MVStore mode: the CLOB and BLOB storage was re-implemented and is now much faster than with the PageStore (which is still the default storage).
@changelog_1012_li
Lucene 2 is no longer supported.
MVStore mode: creating indexes is now much faster (in many cases faster than with the default PageStore).
@changelog_1013_li
Fix bug in calculating default MIN and MAX values for SEQUENCE.
Various bugs in the MVStore storage and have been fixed, including a bug in the R-tree implementation. The database could get corrupt if there were transient IO exceptions while storing.
@changelog_1014_li
The method org.h2.expression.Function.getCost could throw a NullPointException.
@changelog_1015_li
Storing LOBs in separate files (outside of the main database file) is no longer supported for new databases.
@changelog_1016_li
Lucene 2 is no longer supported.
@changelog_1017_li
Fix bug in calculating default MIN and MAX values for SEQUENCE.
@changelog_1018_li
Fix bug in performing IN queries with multiple values when IGNORECASE=TRUE
@changelog_1015_h2
@changelog_1019_li
Add entry-point to org.h2.tools.Shell so it can be called from inside an application. patch by Thomas Gillet.
@changelog_1020_li
Fix bug that prevented the PgServer from being stopped and started multiple times.
@changelog_1021_h2
Version 1.3.175 (2014-01-18)
@changelog_1016_li
@changelog_1022_li
EXPLAIN was incorrect for queries with "ANY" or "SOME" conditions.
@changelog_1017_li
@changelog_1023_li
CallableStatement with "out" parameters: running the same statement twice could result in an exception ("parameter not set").
@changelog_1018_li
@changelog_1024_li
Union queries: duplicate rows could be returned if the sub-queries contained "order by".
@changelog_1019_li
@changelog_1025_li
The GEOMETRY data type now works for user defined functions that return a result set.
@changelog_1020_li
@changelog_1026_li
PostgreSQL compatibility: the PgServer was not working properly when the setting database_to_upper was set to false.
@changelog_1021_li
@changelog_1027_li
JdbcDataSource: the methods setUrl and getUrl where added as aliases for setURL and getURL. This should solve problems with the HikariCP (Hikari connection pool).
@changelog_1022_li
@changelog_1028_li
Issue 537: H2 Console: when loading the schema, incorrect JDBC calls where issued, which caused the exception "Column PRECISION not found".
@changelog_1023_li
@changelog_1029_li
Improved Geometry processing (issue 535).
@changelog_1024_li
@changelog_1030_li
The collation can now be set in the database URL, even if there are data tables, if the collection is the same as the current collation.
@changelog_1025_li
@changelog_1031_li
Improved Oracle compatibility for CASE WHEN and DECODE.
@changelog_1026_li
@changelog_1032_li
The statement "drop all objects" did not work if a table depends on a view via a constraint.
@changelog_1027_li
@changelog_1033_li
Subqueries or views with "order by" an alias expression could not be executed due to a regression introduced in version 1.3.174.
@changelog_1028_li
@changelog_1034_li
Issue 73: MySQL compatibility: support REPLACE, patch by Cemo Koc.
@changelog_1029_li
@changelog_1035_li
The spatial index now works in MVCC mode when using the MVStore storage.
@changelog_1030_li
@changelog_1036_li
MVStore: concurrency problems have been fixed. The API has been simplified.
@changelog_1031_li
@changelog_1037_li
Improve error message when dropping an index that belongs to a constraint, specify constraint in error message.
@changelog_1032_li
@changelog_1038_li
Issue 518: java.sql.Connection.commit() freezes after LOB modification with EXCLUSIVE connection
@changelog_1033_li
@changelog_1039_li
Issue 517: Create or replace view statement has no effect on the others already existing JDBC connection
@changelog_1034_li
@changelog_1040_li
Support 123L syntax as in Java; example: SELECT (2000000000L*2).
@changelog_1035_li
@changelog_1041_li
Issue 520: Add support for sequence max value, min value and cycle, patch by Daniel Gredler.
@changelog_1036_li
@changelog_1042_li
Issue 521: ScriptReader should implement Closeable
@changelog_1037_li
@changelog_1043_li
Issue 524: RunScript.execute does not close its Statement, patch from Gaul.
@changelog_1038_li
@changelog_1044_li
Add support for DB2 "WITH UR" clause, patch from litailang
@changelog_1039_li
@changelog_1045_li
Added support for ON DUPLICATE KEY UPDATE like MySQL with the values() function to update with the value that was to be inserted. Patch from Jean-Francois Noel.
@changelog_1040_li
@changelog_1046_li
Issue 522: Treat empty strings like NULL in Oracle compatibility mode, patch by Daniel Gredler.
@changelog_1041_li
@changelog_1047_li
Issue 527: Oracle compatibility mode: incorrect scale behavior, patch by Daniel Gredler.
@changelog_1042_li
@changelog_1048_li
Slightly reduce the memory cost of View metadata.
@changelog_1043_li
@changelog_1049_li
Extend support of "GRANT ALTER ANY SCHEMA TO <user>" to allow grantee ability to manipulate tables
@changelog_1044_li
@changelog_1050_li
Issue 532: Javadoc for ErrorCode.ROLES_AND_RIGHT_CANNOT_BE_MIXED looks wrong
@changelog_1045_li
@changelog_1051_li
Issue 528: Add Oracle-compatible TO_CHAR function, patch by Daniel Gredler.
@changelog_1046_li
@changelog_1052_li
Issue 534: CREATE ALIAS NOCACHE, Patch by Nicolas Fortin
@changelog_1047_li
@changelog_1053_li
Fix an issue with storing Unicode surrogate pairs in CLOB columns.
Fix bug in DB2 syntax "select * from test with ur", patch from litailang
@changelog_1050_li
@changelog_1056_li
Fix bug in DROP ALL OBJECTS when dealing with inter-schema dependencies.
@changelog_1051_h2
@changelog_1057_h2
Version 1.3.174 (2013-10-19)
@changelog_1052_li
@changelog_1058_li
LIRS cache: bugfix for caches that only contain non-resident entries.
@changelog_1053_li
@changelog_1059_li
For in-memory databases, queries with a "group by" column that is also a hash index threw a RuntimeException.
@changelog_1054_li
@changelog_1060_li
Improved error message for some syntax errors.
@changelog_1055_li
@changelog_1061_li
File system abstraction: if used directly, some file systems did not work correctly with spliced byte buffers (the database engine doesn't use those).
@changelog_1056_li
@changelog_1062_li
To use the MVStore storage engine (which is still work in progress), append ";mv_store=true" to the database URL. Using the MVTableEngine when creating the table is no longer recommended.
@changelog_1057_li
@changelog_1063_li
To compile user defined functions, the javax.tools.JavaCompiler is now used if available, and no temporary files are created. This should solve problems when multiple H2 database concurrently compile the same user defined functions. To disable, system the system property "h2.javaSystemCompiler" to false.
@changelog_1058_li
@changelog_1064_li
Concurrently creating function aliases in different databases could result in the exception "javac: file not found".
@changelog_1059_li
@changelog_1065_li
The function "regexp_replace" threw the wrong kind of exception if the replacement string was invalid.
@changelog_1060_li
@changelog_1066_li
A checkpoint is now done every MAX_LOG_SIZE / 2 instead of every MAX_LOG_SIZE, so that the transaction log doesn't grow as large.
@changelog_1061_li
@changelog_1067_li
MVStore table engine: new setting "retention_time" to configure the time to retain old data. The default is 45 seconds.
@changelog_1062_li
@changelog_1068_li
The method TableEngine.createTable() now returns a Table object.
@changelog_1063_li
@changelog_1069_li
For read-only databases, for the trace level "debug", the trace info is written to the temp directory.
@changelog_1064_li
@changelog_1070_li
Closing the file lock will now wait until the background thread is stopped.
@changelog_1065_li
@changelog_1071_li
In version 1.3.172, a performance regression was introduced when fixing the issue 389 (when there is a multi-column primary key, H2 does not seem to always pick the right index). This was related to boosting an index that matches the "order by" column list (the wrong index was used in some cases).
@changelog_1066_li
@changelog_1072_li
Improved spatial index and data type.
@changelog_1067_li
@changelog_1073_li
Issue 467: OSGi Class Loader (ability to create reference to class in other ClassLoader, for example in another OSGi bundle).
@changelog_1068_li
@changelog_1074_li
Queries "group by" that contain a subquery with an aggregate function returned the wrong result in some cases.
@changelog_1069_li
@changelog_1075_li
Fix bug in unique and non-unique hash indexes which manifested as incorrect results when the search key was a different cardinal type from the table index key. e.g. where the one was INT and the other was LONG
@changelog_1070_li
@changelog_1076_li
Bug: Changes to the database structure did not result in the Session query cache being invalidated.
@changelog_1071_li
@changelog_1077_li
New feature from Davide Cavestro - allow using custom Java object serialization engines on a per-DB basis.
@changelog_1072_li
@changelog_1078_li
When running the Recover tool on very large (>6G) databases, some statistics were reported with negative numbers.
@changelog_1073_li
@changelog_1079_li
Add a CONTAINS_UNCOMMITTED column to the SESSIONS metadata table, to allow detecting when rogue sessions are creating large transactions.
@changelog_1074_li
@changelog_1080_li
Some small fixes to the GEOMETRY support, patches by Nicolas Fortin.
@changelog_1075_li
@changelog_1081_li
The BNF tool and the autocomplete feature are exported in OSGi, which makes it possible for smart editors to do autocomplete. Patch from Nicolas Fortin.
@changelog_1076_li
@changelog_1082_li
Fix DROP ALL OBJECTS and DROP SCHEMA in the presence of tables with computed column dependencies.
@changelog_1077_li
@changelog_1083_li
Session-temporary LOB's could sometimes accumulate, increasing the size of the DB file until shutdown. Now they are cleared out at every commit.
@changelog_1078_li
@changelog_1084_li
There was a bug where a hash index with more than one column would be silently converted to a regular index. It will now throw an exception.
@changelog_1079_li
@changelog_1085_li
Query Statistics: new feature which stores the newest 100 SQL queries executed and their performance data. Useful for tracking down badly performing queries.
@changelog_1080_li
@changelog_1086_li
Fix an LOB deadlock between reading and updating LOB columns.
@changelog_1081_li
@changelog_1087_li
Support the JDBC DatabaseMetaData#getClientInfoProperties() method, in the sense of returning an empty result, in order to make WebSphere happy.
@changelog_1082_li
@changelog_1088_li
Make Server#openBrowser respect the $BROWSER environment variable. Add "chromium" to the list of browsers tried. Patch from Felix Kaiser.
@changelog_1083_li
@changelog_1089_li
Fix to org.h2.util.ScriptReader when handling unclosed block comments.
@changelog_1084_li
@changelog_1090_li
Make org.h2.util.ScriptReader throw a better exception when handling broken scripts which generate extremely large statements.
@changelog_1085_li
@changelog_1091_li
Fix bug with ALLOW_LITERALS=NONE, where the periodic analyze table on insert would throw an exception. A similar problem was fixed in the Console tool.
@changelog_1086_li
@changelog_1092_li
Issue 510: Make org.h2.bnf public for consumption by external projects, patch by Nicolas Fortin
@changelog_1087_li
@changelog_1093_li
Issue 509: Important fix on ValueGeometry, patch by Nicolas Fortin (with some tweaking) Make ValueGeometry#getDimensionCount more reliable. Add unit test to check for illegal ValueGeometry comparison Add unit test for conversion of Geometry object into Object Add optional export to MANIFEST.MF for JTS Geometry classes Validate that geometry values can be represented in WKB.
@changelog_1088_li
@changelog_1094_li
Issue 506: RFE: Include Thread.getName() in case of a deadlock
@changelog_1089_li
@changelog_1095_li
Adding support for "GRANT ALTER ANY SCHEMA TO <user>", patch by John Yates
@changelog_1090_h2
@changelog_1096_h2
Version 1.3.173 (2013-07-28)
@changelog_1091_li
@changelog_1097_li
Support empty statements that just contains a comment.
@changelog_1092_li
@changelog_1098_li
Server mode: if there was an error while reading from a LOB, the session was closed in some cases.
@changelog_1093_li
@changelog_1099_li
Issue 463: Driver name and version are now the same in OsgiDataSourceFactory and JdbcDatabaseMetaData.
@changelog_1094_li
@changelog_1100_li
JaQu: The data type VARCHAR is now (again) used for Strings (no longer TEXT, except when explicitly set).
@changelog_1095_li
@changelog_1101_li
For in-memory databases, creating an index on a CLOB or BLOB column is no longer supported. This is to simplify the MVTableEngine.
@changelog_1096_li
@changelog_1102_li
New column "information_schema.tables.row_count_estimate".
@changelog_1097_li
@changelog_1103_li
Issue 468: trunc(timestamp) could return the wrong value (+12 hours), and trunc(number) throw a NullPointerException.
@changelog_1098_li
@changelog_1104_li
The expression trunc(number) threw a NullPointerException.
@changelog_1099_li
@changelog_1105_li
Fixed a deadlock when updating LOB's concurrently. See TestLob.testDeadlock2().
@changelog_1100_li
@changelog_1106_li
Fixed a deadlock related to very large temporary result sets.
@changelog_1101_li
@changelog_1107_li
Add "-list" command line option to Shell tool so that result-list-mode can be triggered when reading from a file.
@changelog_1102_li
@changelog_1108_li
Issue 474: H2 MySQL Compatibility code fails to ignore "COMMENT" in CREATE TABLE, patch from Aaron Azeckoski.
@changelog_1103_li
@changelog_1109_li
Issue 476: Broken link in jaqu.html
@changelog_1104_li
@changelog_1110_li
Fix potential UTF8 encoding issue in org.h2.store.FileStore, reported by Juerg Spiess.
@changelog_1105_li
@changelog_1111_li
Improve error message when check constraint is broken, test case from Gili (cowwoc).
@changelog_1106_li
@changelog_1112_li
Improve error message when we have a unique constraint violation, displays the offending key in the error message.
@changelog_1107_li
@changelog_1113_li
Issue 478: Support for "SHOW TRANSACTION ISOLATION LEVEL", patch from Andrew Franklin.
@changelog_1108_li
@changelog_1114_li
Issue 475: PgServer: add support for CancelRequest, patch from Andrew Franklin.
@changelog_1109_li
@changelog_1115_li
Issue 473: PgServer missing -key option, patch from Andrew Franklin.
@changelog_1110_li
@changelog_1116_li
Issue 471: CREATE VIEW does not check user rights, patch from Andrew Franklin.
@changelog_1111_li
@changelog_1117_li
Issue 477: PgServer binary transmission of query params is unimplemented, patch from Andrew Franklin.
@changelog_1112_li
@changelog_1118_li
Issue 479: Support for SUBSTRING without a FROM condition, patch from Andrew Franklin.
@changelog_1113_li
@changelog_1119_li
Issue 472: PgServer does not work with any recent Postgres JDBC driver, patch from Andrew Franklin.
@changelog_1114_li
@changelog_1120_li
Add syntax for passing additional parameters into custom TableEngine implementations.
@changelog_1115_li
@changelog_1121_li
Issue 480: Bugfix post issue 475, 477, patch from Andrew Franklin.
@changelog_1116_li
@changelog_1122_li
Issue 481: Further extensions to PgServer to support better support PG JDBC, patch from Andrew Franklin.
@changelog_1117_li
@changelog_1123_li
Add support for spatial datatype GEOMETRY.
@changelog_1118_li
@changelog_1124_li
Add support for in-memory spatial index.
@changelog_1119_li
@changelog_1125_li
change the PageStore#changeCount field from an int to a long, to cope with databases with very high transaction rates.
@changelog_1120_li
@changelog_1126_li
Fix a NullPointerException when attempting to add foreign key reference to a view.
@changelog_1121_li
@changelog_1127_li
Add sufficient ClientInfo support to our javax.sql.Connection implementation to make WebSphere happy.
@changelog_1122_li
@changelog_1128_li
Issue 482: class LobStorageBackend$LobInputStream does not override the method InputStream.available().
@changelog_1123_li
@changelog_1129_li
Fix corruption resulting from a mix of the "WRITE_DELAY=0" option and "SELECT DISTINCT" queries that don't fit in memory.
@changelog_1124_li
@changelog_1130_li
Fix the combination of updating a table which contains an LOB, and reading from the LOB at the same time. Previously it would throw an exception, now it works.
@changelog_1125_li
@changelog_1131_li
Issue 484: In the H2 Console tool, all schemas starting with "INFO" where hidden. Now they are hidden only if the database is not H2. Patch from "mgcodeact"/"cumer d"
@changelog_1126_li
@changelog_1132_li
MySQL compatibility, support the "AUTO_INCREMENT=3" part of the CREATE TABLE statement.
@changelog_1127_li
@changelog_1133_li
Issue 486: MySQL compatibility, support the "DEFAULT CHARSET" part of the CREATE TABLE statement.
@changelog_1128_li
@changelog_1134_li
Issue 487: support the MySQL "SET foreign_key_checks = 0" command
@changelog_1129_li
@changelog_1135_li
Issue 490: support MySQL "USING BTREE" index declaration
@changelog_1130_li
@changelog_1136_li
Issue 485: Database get corrupted when column is renamed for which check constraint was defined inside create table statement.
@changelog_1131_li
@changelog_1137_li
Issue 499: support MySQL "UNIQUE KEY (ID) USING BTREE" constraint syntax
@changelog_1132_li
@changelog_1138_li
Issue 501: "CREATE TABLE .. WITH" not serialized, patch from nico.devel
@changelog_1133_li
@changelog_1139_li
Avoid problems with runtime-compiled ALIAS methods when people have set the JAVA_TOOL_OPTIONS environment variable.
@changelog_1134_h2
@changelog_1140_h2
Version 1.3.172 (2013-05-25)
@changelog_1135_li
@changelog_1141_li
Referential integrity: when adding a referential integrity constraint failed, and if creating the constraint automatically created an index, this index was not removed.
@changelog_1136_li
@changelog_1142_li
The auto-analyze feature now only reads 1000 rows per table instead of 10000.
@changelog_1137_li
@changelog_1143_li
The optimization for IN(...) queries combined with OR could result in a strange exception of the type "column x must be included in the group by list".
@changelog_1138_li
@changelog_1144_li
Issue 454: Use Charset for type-safety.
@changelog_1139_li
@changelog_1145_li
Queries with both LIMIT and OFFSET could throw an IllegalArgumentException.
@changelog_1140_li
@changelog_1146_li
MVStore: multiple issues were fixed: 460, 461, 462, 464, 466.
@changelog_1141_li
@changelog_1147_li
MVStore: larger stores (multiple GB) are now much faster.
@changelog_1142_li
@changelog_1148_li
When using local temporary tables and not dropping them manually before closing the session, and then killing the process could result in a database that couldn't be opened (except when using the recover tool).
@changelog_1143_li
@changelog_1149_li
Support TRUNC(timestamp) for improved Oracle compatibility.
@changelog_1144_li
@changelog_1150_li
Add support for CREATE TABLE TEST (ID BIGSERIAL) for PostgreSQL compatibility. Patch from Jesse Long.
@changelog_1145_li
@changelog_1151_li
Add new collation command SET BINARY_COLLATION UNSIGNED, helps with people testing BINARY columns in MySQL mode.
@changelog_1146_li
@changelog_1152_li
Issue 453: ABBA race conditions in TABLE LINK connection sharing.
@changelog_1147_li
@changelog_1153_li
Issue 449: Postgres Serial data type should not automatically be marked as primary key
@changelog_1148_li
@changelog_1154_li
Issue 406: Support "select h2version()"
@changelog_1149_li
@changelog_1155_li
Issue 389: When there is a multi-column primary key, H2 does not seem to always pick the right index
@changelog_1150_li
@changelog_1156_li
Issue 305: Implement SELECT ... FOR FETCH ONLY
@changelog_1151_li
@changelog_1157_li
Issue 274: Sybase/MSSQLServer compatibility - Add GETDATE and CHARINDEX system functions
@changelog_1152_li
@changelog_1158_li
Issue 274: Sybase/MSSQLServer compatibility - swap parameters of CONVERT function.
@changelog_1153_li
@changelog_1159_li
Issue 274: Sybase/MSSQLServer compatibility - support index clause e.g. "select * from test (index table1_index)"
@changelog_1154_li
@changelog_1160_li
Fix bug in Optimizing SELECT * FROM A WHERE X=1 OR X=2 OR X=3 into SELECT * FROM A WHERE X IN (1,2,3)
@changelog_1155_li
@changelog_1161_li
Issue 442: Groovy patch for SourceCompiler (function ALIAS)
@changelog_1156_li
@changelog_1162_li
Issue 459: Improve LOB documentation
@changelog_1157_h2
@changelog_1163_h2
Version 1.3.171 (2013-03-17)
@changelog_1158_li
@changelog_1164_li
Security: the TCP server did not correctly restrict access rights of clients in some cases. This was specially a problem when using the flag "tcpAllowOthers".
@changelog_1159_li
@changelog_1165_li
H2 Console: the session timeout can now be configured using the system property "h2.consoleTimeout".
@changelog_1160_li
@changelog_1166_li
Issue 431: Improved compatibility with MySQL: support for "ENGINE=InnoDB charset=UTF8" when creating a table.
@changelog_1161_li
@changelog_1167_li
Issue 249: Improved compatibility with MySQL in the MySQL mode: now the methods DatabaseMetaData methods stores*Case*Identifiers return the same as MySQL when using the MySQL mode.
@changelog_1162_li
@changelog_1168_li
Issue 434: H2 Console didn't work in the Chrome browser due to a wrong viewport argument.
@changelog_1163_li
@changelog_1169_li
There was a possibility that the .lock.db file was not deleted when the database was closed, which could slow down opening the database.
@changelog_1164_li
@changelog_1170_li
The SQL script generated by the "script" command contained inconsistent newlines on Windows.
@changelog_1165_li
@changelog_1171_li
When using trace level 4 (SLF4J) in the server mode, a directory "trace.db" and an empty file was created on the client side. This is no longer made.
@changelog_1166_li
@changelog_1172_li
Optimize IN(...) queries: there was a bug in version 1.3.170 if the type of the left hand side didn't match the type of the right hand side. Fixed.
@changelog_1167_li
@changelog_1173_li
Optimize IN(...) queries: there was a bug in version 1.3.170 for comparison of the type "X IN(NULL, NULL)". Fixed.
@changelog_1168_li
@changelog_1174_li
Timestamps with timezone that were passed as a string were not always converted correctly. For example "2012-11-06T23:00:00.000Z" was converted to "2012-11-06" instead of to "2012-11-07" in the timezone CET. Thanks a lot to Steve Hruda for reporting the problem!
@changelog_1169_li
@changelog_1175_li
New table engine "org.h2.mvstore.db.MVTableEngine" that internally uses the MVStore to persist data. To try it out, append ";DEFAULT_TABLE_ENGINE=org.h2.mvstore.db.MVTableEngine" to the database URL. This is still very experimental, and many features are not supported yet. The data is stored in a file with the suffix ".mv.db".
@changelog_1170_li
@changelog_1176_li
New connection setting "DEFAULT_TABLE_ENGINE" to use a specific table engine if none is set explicitly. This is to simplify testing the MVStore table engine.
@changelog_1171_li
@changelog_1177_li
MVStore: encrypted stores are now supported. Only standardized algorithms are used: PBKDF2, SHA-256, XTS-AES, AES-128.
@changelog_1172_li
@changelog_1178_li
MVStore: improved API thanks to Simo Tripodi.
@changelog_1173_li
@changelog_1179_li
MVStore: maps can now be renamed.
@changelog_1174_li
@changelog_1180_li
MVStore: store the file header also at the end of each chunk, which results in a further reduced number of write operations.
@changelog_1175_li
@changelog_1181_li
MVStore: a map implementation that supports concurrent operations.
@changelog_1176_li
@changelog_1182_li
MVStore: unified exception handling; the version is included in the messages.
@changelog_1177_li
@changelog_1183_li
MVStore: old data is now retained for 45 seconds by default.
@changelog_1178_li
@changelog_1184_li
MVStore: compress is now disabled by default, and can be enabled on request.
@changelog_1179_li
@changelog_1185_li
Support ALTER TABLE ADD ... AFTER. Patch from Andrew Gaul (argaul at gmail.com). Fixes issue 401.
@changelog_1180_li
@changelog_1186_li
Improved OSGi support. H2 now registers itself as a DataSourceFactory service. Fixes issue 365.
@changelog_1181_li
@changelog_1187_li
Add a DISK_SPACE_USED system function. Fixes issue 270.
@changelog_1182_li
@changelog_1188_li
Fix a compile-time ambiguity when compiling with JDK7, thanks to a patch from Lukas Eder.
@changelog_1183_li
@changelog_1189_li
Supporting dropping an index for Lucene full-text indexes.
@changelog_1184_li
@changelog_1190_li
Optimized performance for SELECT ... ORDER BY X LIMIT Y OFFSET Z queries for in-memory databases using partial sort (by Sergi Vladykin).
@changelog_1185_li
@changelog_1191_li
Experimental off-heap memory storage engine "nioMemFS:" and "nioMemLZF:", suggestion from Mark Addleman.
@changelog_1186_li
@changelog_1192_li
Issue 438: JdbcDatabaseMetaData.getSchemas() is no longer supported as of 1.3.169.
@changelog_1187_li
@changelog_1193_li
MySQL compatibility: support for ALTER TABLE tableName MODIFY [COLUMN] columnName columnDef. Patch from Ville Koskela.
@changelog_1188_li
@changelog_1194_li
Issue 404: SHOW COLUMNS FROM tableName does not work with ALLOW_LITERALS=NUMBERS.
@changelog_1189_li
@changelog_1195_li
Throw an explicit error to make it clear we don't support the TRIGGER combination of SELECT and FOR EACH ROW.
@changelog_1190_li
@changelog_1196_li
Issue 439: Utils.sortTopN does not handle single-element arrays.
@changelog_1191_h2
@changelog_1197_h2
Version 1.3.170 (2012-11-30)
@changelog_1192_li
@changelog_1198_li
Issue 407: The TriggerAdapter didn't work with CLOB and BLOB columns.
@changelog_1193_li
@changelog_1199_li
PostgreSQL compatibility: support for data types BIGSERIAL and SERIAL as an alias for AUTO_INCREMENT.
@changelog_1194_li
@changelog_1200_li
Issue 417: H2 Console: the web session timeout didn't work, resulting in a memory leak. This was only a problem if the H2 Console was run for a long time and many sessions were opened.
@changelog_1195_li
@changelog_1201_li
Issue 412: Running the Server tool with just the option "-browser" will now log a warning.
@changelog_1196_li
@changelog_1202_li
Issue 411: CloseWatcher registration was not concurrency-safe.
@changelog_1197_li
@changelog_1203_li
MySQL compatibility: support for CONCAT_WS. Thanks a lot to litailang for the patch!
@changelog_1198_li
@changelog_1204_li
PostgreSQL compatibility: support for EXTRACT(WEEK FROM dateColumn). Thanks to Prashant Bhat for the patch!
@changelog_1199_li
@changelog_1205_li
Fix for a bug where we would sometimes use the wrong unique constraint to validate foreign key constraints.
@changelog_1200_li
@changelog_1206_li
Support BOM at the beginning of files for the RUNSCRIPT command
@changelog_1201_li
@changelog_1207_li
Fix in calling SET @X = IDENTITY() where it would return NULL incorrectly
@changelog_1202_li
@changelog_1208_li
Fix ABBA deadlock between adding a constraint and the H2-Log-Writer thread.
@changelog_1203_li
@changelog_1209_li
Optimize IN(...) queries where the values are constant and of the same type.
@changelog_1204_li
@changelog_1210_li
Restore tool: the parameter "quiet" was not used and is now removed.
@changelog_1205_li
@changelog_1211_li
Fix ConcurrentModificationException when creating tables and executing SHOW TABLES in parallel. Reported by Viktor Voytovych.
@changelog_1206_li
@changelog_1212_li
Serialization is now pluggable using the system property "h2.javaObjectSerializer". Thanks to Sergi Vladykin for the patch!
@changelog_1207_h2
@changelog_1213_h2
Version 1.3.169 (2012-09-09)
@changelog_1208_li
@changelog_1214_li
The default jar file is now compiled for Java 6.
@changelog_1209_li
@changelog_1215_li
The new jar file will probably not end up in the central Maven repository in the next few weeks because Sonatype has disabled automatic synchronization from SourceForge (which they call 'legacy sync' now). It will probably take some time until this is sorted out. The H2 jar files are deployed to http://h2database.com/m2-repo/com/h2database/h2/maven-metadata.xml and http://hsql.sourceforge.net/m2-repo/com/h2database/h2/maven-metadata.xml as usual.
@changelog_1210_li
@changelog_1216_li
A part of the documentation and the H2 Console has been changed to support the Apple retina display.
@changelog_1211_li
@changelog_1217_li
The CreateCluster tool could not be used if the source database contained a CLOB or BLOB. The root cause was that the TCP server did not synchronize on the session, which caused a problem when using the exclusive mode.
@changelog_1212_li
@changelog_1218_li
Statement.getQueryTimeout(): only the first call to this method will query the database. If the query timeout was changed in another way than calling setQueryTimeout, this method will always return the last value. This was changed because Hibernate calls getQueryTimeout() a lot.
@changelog_1213_li
@changelog_1219_li
Issue 416: PreparedStatement.setNString throws AbstractMethodError. All implemented JDBC 4 methods that don't break compatibility with Java 5 are now included in the default jar file.
@changelog_1214_li
@changelog_1220_li
Issue 414: for some functions, the parameters were evaluated twice (for example "char(nextval(..))" ran "nextval(..)" twice).
@changelog_1215_li
@changelog_1221_li
The ResultSetMetaData methods getSchemaName and getTableName could return null instead of "" (an empty string) as specified in the JDBC API.
@changelog_1216_li
@changelog_1222_li
Added compatibility for "SET NAMES" query in MySQL compatibility mode.
@changelog_1217_h2
@changelog_1223_h2
Version 1.3.168 (2012-07-13)
@changelog_1218_li
@changelog_1224_li
The message "Transaction log could not be truncated" was sometimes written to the .trace.db file even if there was no problem truncating the transaction log.
@changelog_1219_li
@changelog_1225_li
New system property "h2.serializeJavaObject" (default: true) that allows to disable serializing Java objects, so that the objects compareTo and toString methods can be used.
@changelog_1220_li
@changelog_1226_li
Dylan has translated the H2 Console tool to Korean. Thanks a lot!
@changelog_1221_li
@changelog_1227_li
Executing the statement CREATE INDEX IF ALREADY EXISTS if the index already exists no longer fails for a read only database.
@changelog_1222_li
@changelog_1228_li
MVCC: concurrently updating a row could result in the row to appear deleted in the second connection, if there are multiple unique indexes (or a primary key and at least one unique index). Thanks a lot to Teruo for the patch!
@changelog_1223_li
@changelog_1229_li
Fulltext search: in-memory Lucene indexes are now supported.
@changelog_1224_li
@changelog_1230_li
Fulltext search: UUID primary keys are now supported.
@changelog_1225_li
@changelog_1231_li
Apache Tomcat 7.x will now longer log a warning when unloading the web application, if using a connection pool.
@changelog_1226_li
@changelog_1232_li
H2 Console: support the Midori browser (for Debian / Raspberry Pi)
@changelog_1227_li
@changelog_1233_li
When opening a remote session, don't open a temporary file if the trace level is set to zero
@changelog_1228_li
@changelog_1234_li
Use HMAC for authenticating remote LOB id's, removing the need for maintaining a cache, and removing the limit on the number of LOBs per result set.
@changelog_1229_li
@changelog_1235_li
H2 Console: HTML and XML documents can now be edited in an updatable result set. There is (limited) support for editing multi-line documents.
@cheatSheet_1000_h1
...
...
@@ -7529,195 +7547,225 @@ Storage Engine for H2
File Format
@mvstore_1118_p
The data is stored in one file. The file contains two file headers (to be safe), and a number of chunks. The file headers are one block each; a block is 4096 bytes. Each chunk is at least one block, but typically 200 blocks or more. There might be a number of free blocks in front of every chunk. There is one chunk for every version.
The data is stored in one file. The file contains two file headers (for safety), and a number of chunks. The file headers are one block each; a block is 4096 bytes. Each chunk is at least one block, but typically 200 blocks or more. Data is stored in the chunks in the form of a <a href="https://en.wikipedia.org/wiki/Log-structured_file_system">log structured storage</a>. There is one chunk for every version.
@mvstore_1119_h3
File Header
@mvstore_1119_p
Each chunk contains a number of B-tree pages. As an example, the following code:
@mvstore_1120_p
will result in the following two chunks (excluding metadata):
@mvstore_1121_b
Chunk 1:
@mvstore_1122_p
- Page 1: leaf with 140 entries (keys 0 - 139)
@mvstore_1123_p
- Page 2: leaf with 260 entries (keys 140 - 399)
@mvstore_1124_p
- Page 3: (root) node with 2 entries pointing to page 1 and 2
@mvstore_1125_b
Chunk 2:
@mvstore_1126_p
- Page 4: leaf with 140 entries (keys 0 - 139)
@mvstore_1127_p
- Page 5: (root) node with 2 entries pointing to page 4 and 2
@mvstore_1128_p
That means each chunk contains the changes of one version: the new version of the changed pages and the parent pages, recursively, up to the root page. Pages in subsequent chunks refer to pages in earlier chunks.
@mvstore_1129_h3
File Header
@mvstore_1130_p
There are two file headers, which normally contain the exact same data. But once in a while, the file headers are updated, and writing could partially fail, which could corrupt a header. That's why there is a second header. Only the file headers are updated in this way (called "in-place update"). The headers contain the following data:
@mvstore_1121_p
@mvstore_1131_p
The data is stored in the form of a key-value pair. Each value is stored as a hexadecimal number. The entries are:
@mvstore_1122_li
@mvstore_1132_li
H: The entry "H:2" stands for the the H2 database.
@mvstore_1123_li
@mvstore_1133_li
block: The block number where one of the newest chunks starts (but not necessarily the newest).
@mvstore_1124_li
@mvstore_1134_li
blockSize: The block size of the file; currently always hex 1000, which is decimal 4096, to match the <a href="https://en.wikipedia.org/wiki/Disk_sector">disk sector</a> length of modern hard disks.
@mvstore_1125_li
@mvstore_1135_li
chunk: The chunk id, which is normally the same value as the version; however, the chunk id might roll over to 0, while the version doesn't.
@mvstore_1126_li
@mvstore_1136_li
created: The number of milliseconds since 1970 when the file was created.
@mvstore_1127_li
@mvstore_1137_li
format: The file format number. Currently 1.
@mvstore_1128_li
@mvstore_1138_li
version: The version number of the chunk.
@mvstore_1129_li
@mvstore_1139_li
fletcher: The <a href="https://en.wikipedia.org/wiki/Fletcher's_checksum"> Fletcher-32 checksum</a> of the header.
@mvstore_1130_p
@mvstore_1140_p
When opening the file, both headers are read and the checksum is verified. If both headers are valid, the one with the newer version is used. The chunk with the latest version is then detected (details about this see below), and the rest of the metadata is read from there. If the chunk id, block and version are not stored in the file header, then the latest chunk lookup starts with the last chunk in the file.
@mvstore_1131_h3
@mvstore_1141_h3
Chunk Format
@mvstore_1132_p
@mvstore_1142_p
There is one chunk per version. Each chunk consists of a header, the pages that were modified in this version, and a footer. The pages contain the actual data of the maps. The pages inside a chunk are stored right after the header, next to each other (unaligned). The size of a chunk is a multiple of the block size. The footer is stored in the last 128 bytes of the chunk.
@mvstore_1133_p
@mvstore_1143_p
The footer allows to verify that the chunk is completely written (a chunk is written as one write operation), and allows to find the start position of the very last chunk in the file. The chunk header and footer contain the following data:
@mvstore_1134_p
@mvstore_1144_p
The fields of the chunk header and footer are:
@mvstore_1135_li
@mvstore_1145_li
chunk: The chunk id.
@mvstore_1136_li
@mvstore_1146_li
block: The first block of the chunk (multiply by the block size to get the position in the file).
@mvstore_1137_li
@mvstore_1147_li
len: The size of the chunk in number of blocks.
@mvstore_1138_li
@mvstore_1148_li
map: The id of the newest map; incremented when a new map is created.
@mvstore_1139_li
@mvstore_1149_li
max: The sum of all maximum page sizes (see page format).
@mvstore_1140_li
@mvstore_1150_li
next: The predicted start block of the next chunk.
@mvstore_1141_li
@mvstore_1151_li
pages: The number of pages in the chunk.
@mvstore_1142_li
@mvstore_1152_li
root: The position of the metadata root page (see page format).
@mvstore_1143_li
@mvstore_1153_li
time: The time the chunk was written, in milliseconds after the file was created.
@mvstore_1144_li
@mvstore_1154_li
version: The version this chunk represents.
@mvstore_1145_li
@mvstore_1155_li
fletcher: The checksum of the footer.
@mvstore_1146_p
Chunks are never updated in-place. Each chunk contains the pages that were changed in that version (there is one chunk per version, see above), plus all the parent nodes of those pages, recursively, up to the root page. If an entry in a map is changed, removed, or added, then the respective page is copied to be stored in the next chunk, and the number of live pages in the old chunk is decremented. This mechanism is called copy-on-write, and is similar to how the <a href="https://en.wikipedia.org/wiki/Btrfs">Btrfs</a> file system works. Chunks without live pages are marked as free, so the space can be re-used by more recent chunks. Because not all chunks are of the same size, there can be some unused space in front of a chunk for some time (until a small chunk is written or the chunks are compacted). There is a <a href="http://stackoverflow.com/questions/13650134/after-how-many-seconds-are-file-system-write-buffers-typically-flushed"> delay of 45 seconds</a> (by default) before a free chunk is overwritten, to ensure new versions are persisted first, as hard disks sometimes re-order write operations.
@mvstore_1156_p
Chunks are never updated in-place. Each chunk contains the pages that were changed in that version (there is one chunk per version, see above), plus all the parent nodes of those pages, recursively, up to the root page. If an entry in a map is changed, removed, or added, then the respective page is copied to be stored in the next chunk, and the number of live pages in the old chunk is decremented. This mechanism is called copy-on-write, and is similar to how the <a href="https://en.wikipedia.org/wiki/Btrfs">Btrfs</a> file system works. Chunks without live pages are marked as free, so the space can be re-used by more recent chunks. Because not all chunks are of the same size, there can be a number of free blocks in front of a chunk for some time (until a small chunk is written or the chunks are compacted). There is a <a href="http://stackoverflow.com/questions/13650134/after-how-many-seconds-are-file-system-write-buffers-typically-flushed"> delay of 45 seconds</a> (by default) before a free chunk is overwritten, to ensure new versions are persisted first.
@mvstore_1147_p
@mvstore_1157_p
How the newest chunk is located when opening a store: The file header contains the position of a recent chunk, but not always the newest one. This is to reduce the number of file header updates. After opening the file, the file headers, and the chunk footer of the very last chunk (at the end of the file) are read. From those candidates, the header of the most recent chunk is read. If it contains a "next" pointer (see above), those chunk's header and footer are read as well. If it turned out to be a newer valid chunk, this is repeated, until the newest chunk was found. Before writing a chunk, the position of the next chunk is predicted based on the assumption that the next chunk will be of the same size as the current one. When the next chunk is written, and the previous prediction turned out to be incorrect, the file header is updated as well. In any case, the file header is updated if the next chain gets longer than 20 hops.
@mvstore_1148_h3
@mvstore_1158_h3
Page Format
@mvstore_1149_p
@mvstore_1159_p
Each map is a <a href="https://en.wikipedia.org/wiki/B-tree">B-tree</a>, and the map data is stored in (B-tree-) pages. There are leaf pages that contain the key-value pairs of the map, and internal nodes, which only contain keys and pointers to leaf pages. The root of a tree is either a leaf or an internal node. Unlike file header and chunk header and footer, the page data is not human readable. Instead, it is stored as byte arrays, with long (8 bytes), int (4 bytes), short (2 bytes), and <a href="https://en.wikipedia.org/wiki/Variable-length_quantity">variable size int and long</a> (1 to 5 / 10 bytes). The page format is:
@mvstore_1150_li
@mvstore_1160_li
length (int): Length of the page in bytes.
@mvstore_1151_li
@mvstore_1161_li
checksum (short): Checksum (chunk id xor offset within the chunk xor page length).
@mvstore_1152_li
@mvstore_1162_li
mapId (variable size int): The id of the map this page belongs to.
@mvstore_1153_li
@mvstore_1163_li
len (variable size int): The number of keys in the page.
@mvstore_1154_li
@mvstore_1164_li
type (byte): The page type (0 for leaf page, 1 for internal node; plus 2 if the page data is compressed).
@mvstore_1155_li
keys (byte array): All keys, stored depending on the data type.
@mvstore_1156_li
@mvstore_1165_li
children (array of long; internal nodes only): The position of the children.
@mvstore_1157_li
@mvstore_1166_li
childCounts (array of variable size long; internal nodes only): The total number of entries for the given child page.
@mvstore_1158_li
@mvstore_1167_li
keys (byte array): All keys, stored depending on the data type.
@mvstore_1168_li
values (byte array; leaf pages only): All values, stored depending on the data type.
@mvstore_1159_p
@mvstore_1169_p
Even though this is not required by the file format, each B-tree is stored "upside down", that means the leaf pages first, then the internal nodes, and lastly the root page.
@mvstore_1160_p
Pointers to pages are stored as a long, using a special format: 26 bits for the chunk id, 32 bits for the offset within the chunk, 5 bits for the length code, 1 bit for the page type (leaf or internal node). The page type is encoded so that when clearing or removing a map, leaf pages don't have to be read (internal nodes do have to be read in order to know where all the pages are; but in a typical B-tree the vast majority of the pages are leaf pages). The absolute file position is not included so that chunks can be moved within the file without having to change page pointers; only the chunk metadata needs to be changed. The length code is a number from 0 to 31, where 0 means the maximum length of the page is 32 bytes, 1 means 48 bytes, 2: 64, 3: 96, 4: 128, 5: 192, and so on until 31 which means longer than 1 MB. That way, reading a page only requires one read operation (except for very large pages, in which case two read operations might be needed). The sum of the maximum length of all pages is stored in the chunk metadata (field "max"), and when a page is marked as removed, the live maximum length is adjusted. This allows to estimate the amount of free space within a block, in addition to the number of free pages.
@mvstore_1170_p
Pointers to pages are stored as a long, using a special format: 26 bits for the chunk id, 32 bits for the offset within the chunk, 5 bits for the length code, 1 bit for the page type (leaf or internal node). The page type is encoded so that when clearing or removing a map, leaf pages don't have to be read (internal nodes do have to be read in order to know where all the pages are; but in a typical B-tree the vast majority of the pages are leaf pages). The absolute file position is not included so that chunks can be moved within the file without having to change page pointers; only the chunk metadata needs to be changed. The length code is a number from 0 to 31, where 0 means the maximum length of the page is 32 bytes, 1 means 48 bytes, 2: 64, 3: 96, 4: 128, 5: 192, and so on until 31 which means longer than 1 MB. That way, reading a page only requires one read operation (except for very large pages). The sum of the maximum length of all pages is stored in the chunk metadata (field "max"), and when a page is marked as removed, the live maximum length is adjusted. This allows to estimate the amount of free space within a block, in addition to the number of free pages.
@mvstore_1161_p
@mvstore_1171_p
The total number of entries in a child nodes is kept to allow efficient range counting, lookup by index, and skip operations. The pages form a <a href="http://www.chiark.greenend.org.uk/~sgtatham/algorithms/cbtree.html">counted B-tree</a>.
@mvstore_1162_p
@mvstore_1172_p
Data compression: The data after the page type are optionally compressed using the LZF algorithm.
@mvstore_1163_h3
@mvstore_1173_h3
Metadata Map
@mvstore_1164_p
@mvstore_1174_p
In addition to the user maps, there is one metadata map that contains names and positions of user maps, and chunk metadata. The very last page of a chunk contains the root page of that metadata map. The exact position of this root page is stored in the chunk header. This page (directly or indirectly) points to the root pages of all other maps. The metadata map of a store with a map named "data", and one chunk, contains the following entries:
@mvstore_1165_li
@mvstore_1175_li
chunk.1: The metadata of chunk 1. This is the same data as the chunk header, plus the number of live pages, and the maximum live length.
@mvstore_1166_li
setting.storeVersion: The store version (a user defined value).
@mvstore_1167_li
@mvstore_1176_li
map.1: The metadata of map 1. The entries are: name, createVersion, and type.
@mvstore_1168_li
@mvstore_1177_li
name.data: The map id of the map named "data". The value is "1".
@mvstore_1169_li
@mvstore_1178_li
root.1: The root position of map 1.
@mvstore_1170_h2
@mvstore_1179_li
setting.storeVersion: The store version (a user defined value).
@mvstore_1180_h2
Similar Projects and Differences to Other Storage Engines
@mvstore_1171_p
@mvstore_1181_p
Unlike similar storage engines like LevelDB and Kyoto Cabinet, the MVStore is written in Java and can easily be embedded in a Java and Android application.
@mvstore_1172_p
@mvstore_1182_p
The MVStore is somewhat similar to the Berkeley DB Java Edition because it is also written in Java, and is also a log structured storage, but the H2 license is more liberal.
@mvstore_1173_p
@mvstore_1183_p
Like SQLite 3, the MVStore keeps all data in one file. Unlike SQLite 3, the MVStore uses is a log structured storage. The plan is to make the MVStore both easier to use as well as faster than SQLite 3. In a recent (very simple) test, the MVStore was about twice as fast as SQLite 3 on Android.
@mvstore_1174_p
@mvstore_1184_p
The API of the MVStore is similar to MapDB (previously known as JDBM) from Jan Kotek, and some code is shared between MVStore and MapDB. However, unlike MapDB, the MVStore uses is a log structured storage. The MVStore does not have a record size limit.
@mvstore_1175_h2
@mvstore_1185_h2
Current State
@mvstore_1176_p
@mvstore_1186_p
The code is still experimental at this stage. The API as well as the behavior may partially change. Features may be added and removed (even though the main features will stay).
@mvstore_1177_h2
@mvstore_1187_h2
Requirements
@mvstore_1178_p
@mvstore_1188_p
The MVStore is included in the latest H2 jar file.
@mvstore_1179_p
@mvstore_1189_p
There are no special requirements to use it. The MVStore should run on any JVM as well as on Android.
@mvstore_1180_p
@mvstore_1190_p
To build just the MVStore (without the database engine), run:
@mvstore_1181_p
@mvstore_1191_p
This will create the file <code>bin/h2mvstore-1.3.175.jar</code> (about 130 KB).
#Referential integrity constraints sometimes used the wrong index, such that updating a row in the referenced table incorrectly failed with a constraint violation.
#H2 Console: auto-complete is now triggered by the tab key.
@changelog_1003_li
#The Polish translation was completed and corrected by Wojtek Jurczyk. Thanks a lot!
#H2 Console: auto-complete did not work with multi-line statements.
@changelog_1004_li
#Issue 545: Unnecessary duplicate code was removed.
#CLOB and BLOB data was not immediately removed after a rollback.
@changelog_1005_li
#The profiler tool can now process files with full thread dumps.
#There is a new Aggregate API that supports the internal H2 data types (GEOMETRY for example). Thanks a lot to Nicolas Fortin for the patch!
@changelog_1006_li
#MVStore: the file format was changed slightly.
#Referential integrity constraints sometimes used the wrong index, such that updating a row in the referenced table incorrectly failed with a constraint violation.
@changelog_1007_li
#MVStore mode: the CLOB and BLOB storage was re-implemented and is now much faster than with the PageStore (which is still the default storage).
#The Polish translation was completed and corrected by Wojtek Jurczyk. Thanks a lot!
@changelog_1008_li
#MVStore mode: creating indexes is now much faster (in many cases faster than with the default PageStore).
#Issue 545: Unnecessary duplicate code was removed.
@changelog_1009_li
#Various bugs in the MVStore storage and have been fixed, including a bug in the R-tree implementation. The database could get corrupt if there were transient IO exceptions while storing.
#The profiler tool can now process files with full thread dumps.
@changelog_1010_li
#The method org.h2.expression.Function.getCost could throw a NullPointException.
#MVStore: the file format was changed slightly.
@changelog_1011_li
#Storing LOBs in separate files (outside of the main database file) is no longer supported for new databases.
#MVStore mode: the CLOB and BLOB storage was re-implemented and is now much faster than with the PageStore (which is still the default storage).
@changelog_1012_li
#Lucene 2 is no longer supported.
#MVStore mode: creating indexes is now much faster (in many cases faster than with the default PageStore).
@changelog_1013_li
#Fix bug in calculating default MIN and MAX values for SEQUENCE.
#Various bugs in the MVStore storage and have been fixed, including a bug in the R-tree implementation. The database could get corrupt if there were transient IO exceptions while storing.
@changelog_1014_li
#The method org.h2.expression.Function.getCost could throw a NullPointException.
@changelog_1015_li
#Storing LOBs in separate files (outside of the main database file) is no longer supported for new databases.
@changelog_1016_li
#Lucene 2 is no longer supported.
@changelog_1017_li
#Fix bug in calculating default MIN and MAX values for SEQUENCE.
@changelog_1018_li
#Fix bug in performing IN queries with multiple values when IGNORECASE=TRUE
@changelog_1015_h2
@changelog_1019_li
#Add entry-point to org.h2.tools.Shell so it can be called from inside an application. patch by Thomas Gillet.
@changelog_1020_li
#Fix bug that prevented the PgServer from being stopped and started multiple times.
@changelog_1021_h2
#Version 1.3.175 (2014-01-18)
@changelog_1016_li
@changelog_1022_li
#EXPLAIN was incorrect for queries with "ANY" or "SOME" conditions.
@changelog_1017_li
@changelog_1023_li
#CallableStatement with "out" parameters: running the same statement twice could result in an exception ("parameter not set").
@changelog_1018_li
@changelog_1024_li
#Union queries: duplicate rows could be returned if the sub-queries contained "order by".
@changelog_1019_li
@changelog_1025_li
#The GEOMETRY data type now works for user defined functions that return a result set.
@changelog_1020_li
@changelog_1026_li
#PostgreSQL compatibility: the PgServer was not working properly when the setting database_to_upper was set to false.
@changelog_1021_li
@changelog_1027_li
#JdbcDataSource: the methods setUrl and getUrl where added as aliases for setURL and getURL. This should solve problems with the HikariCP (Hikari connection pool).
@changelog_1022_li
@changelog_1028_li
#Issue 537: H2 Console: when loading the schema, incorrect JDBC calls where issued, which caused the exception "Column PRECISION not found".
@changelog_1023_li
@changelog_1029_li
#Improved Geometry processing (issue 535).
@changelog_1024_li
@changelog_1030_li
#The collation can now be set in the database URL, even if there are data tables, if the collection is the same as the current collation.
@changelog_1025_li
@changelog_1031_li
#Improved Oracle compatibility for CASE WHEN and DECODE.
@changelog_1026_li
@changelog_1032_li
#The statement "drop all objects" did not work if a table depends on a view via a constraint.
@changelog_1027_li
@changelog_1033_li
#Subqueries or views with "order by" an alias expression could not be executed due to a regression introduced in version 1.3.174.
@changelog_1028_li
@changelog_1034_li
#Issue 73: MySQL compatibility: support REPLACE, patch by Cemo Koc.
@changelog_1029_li
@changelog_1035_li
#The spatial index now works in MVCC mode when using the MVStore storage.
@changelog_1030_li
@changelog_1036_li
#MVStore: concurrency problems have been fixed. The API has been simplified.
@changelog_1031_li
@changelog_1037_li
#Improve error message when dropping an index that belongs to a constraint, specify constraint in error message.
@changelog_1032_li
@changelog_1038_li
#Issue 518: java.sql.Connection.commit() freezes after LOB modification with EXCLUSIVE connection
@changelog_1033_li
@changelog_1039_li
#Issue 517: Create or replace view statement has no effect on the others already existing JDBC connection
@changelog_1034_li
@changelog_1040_li
#Support 123L syntax as in Java; example: SELECT (2000000000L*2).
@changelog_1035_li
@changelog_1041_li
#Issue 520: Add support for sequence max value, min value and cycle, patch by Daniel Gredler.
@changelog_1036_li
@changelog_1042_li
#Issue 521: ScriptReader should implement Closeable
@changelog_1037_li
@changelog_1043_li
#Issue 524: RunScript.execute does not close its Statement, patch from Gaul.
@changelog_1038_li
@changelog_1044_li
#Add support for DB2 "WITH UR" clause, patch from litailang
@changelog_1039_li
@changelog_1045_li
#Added support for ON DUPLICATE KEY UPDATE like MySQL with the values() function to update with the value that was to be inserted. Patch from Jean-Francois Noel.
@changelog_1040_li
@changelog_1046_li
#Issue 522: Treat empty strings like NULL in Oracle compatibility mode, patch by Daniel Gredler.
@changelog_1041_li
@changelog_1047_li
#Issue 527: Oracle compatibility mode: incorrect scale behavior, patch by Daniel Gredler.
@changelog_1042_li
@changelog_1048_li
#Slightly reduce the memory cost of View metadata.
@changelog_1043_li
@changelog_1049_li
#Extend support of "GRANT ALTER ANY SCHEMA TO <user>" to allow grantee ability to manipulate tables
@changelog_1044_li
@changelog_1050_li
#Issue 532: Javadoc for ErrorCode.ROLES_AND_RIGHT_CANNOT_BE_MIXED looks wrong
@changelog_1045_li
@changelog_1051_li
#Issue 528: Add Oracle-compatible TO_CHAR function, patch by Daniel Gredler.
@changelog_1046_li
@changelog_1052_li
#Issue 534: CREATE ALIAS NOCACHE, Patch by Nicolas Fortin
@changelog_1047_li
@changelog_1053_li
#Fix an issue with storing Unicode surrogate pairs in CLOB columns.
#Fix bug in DB2 syntax "select * from test with ur", patch from litailang
@changelog_1050_li
@changelog_1056_li
#Fix bug in DROP ALL OBJECTS when dealing with inter-schema dependencies.
@changelog_1051_h2
@changelog_1057_h2
#Version 1.3.174 (2013-10-19)
@changelog_1052_li
@changelog_1058_li
#LIRS cache: bugfix for caches that only contain non-resident entries.
@changelog_1053_li
@changelog_1059_li
#For in-memory databases, queries with a "group by" column that is also a hash index threw a RuntimeException.
@changelog_1054_li
@changelog_1060_li
#Improved error message for some syntax errors.
@changelog_1055_li
@changelog_1061_li
#File system abstraction: if used directly, some file systems did not work correctly with spliced byte buffers (the database engine doesn't use those).
@changelog_1056_li
@changelog_1062_li
#To use the MVStore storage engine (which is still work in progress), append ";mv_store=true" to the database URL. Using the MVTableEngine when creating the table is no longer recommended.
@changelog_1057_li
@changelog_1063_li
#To compile user defined functions, the javax.tools.JavaCompiler is now used if available, and no temporary files are created. This should solve problems when multiple H2 database concurrently compile the same user defined functions. To disable, system the system property "h2.javaSystemCompiler" to false.
@changelog_1058_li
@changelog_1064_li
#Concurrently creating function aliases in different databases could result in the exception "javac: file not found".
@changelog_1059_li
@changelog_1065_li
#The function "regexp_replace" threw the wrong kind of exception if the replacement string was invalid.
@changelog_1060_li
@changelog_1066_li
#A checkpoint is now done every MAX_LOG_SIZE / 2 instead of every MAX_LOG_SIZE, so that the transaction log doesn't grow as large.
@changelog_1061_li
@changelog_1067_li
#MVStore table engine: new setting "retention_time" to configure the time to retain old data. The default is 45 seconds.
@changelog_1062_li
@changelog_1068_li
#The method TableEngine.createTable() now returns a Table object.
@changelog_1063_li
@changelog_1069_li
#For read-only databases, for the trace level "debug", the trace info is written to the temp directory.
@changelog_1064_li
@changelog_1070_li
#Closing the file lock will now wait until the background thread is stopped.
@changelog_1065_li
@changelog_1071_li
#In version 1.3.172, a performance regression was introduced when fixing the issue 389 (when there is a multi-column primary key, H2 does not seem to always pick the right index). This was related to boosting an index that matches the "order by" column list (the wrong index was used in some cases).
@changelog_1066_li
@changelog_1072_li
#Improved spatial index and data type.
@changelog_1067_li
@changelog_1073_li
#Issue 467: OSGi Class Loader (ability to create reference to class in other ClassLoader, for example in another OSGi bundle).
@changelog_1068_li
@changelog_1074_li
#Queries "group by" that contain a subquery with an aggregate function returned the wrong result in some cases.
@changelog_1069_li
@changelog_1075_li
#Fix bug in unique and non-unique hash indexes which manifested as incorrect results when the search key was a different cardinal type from the table index key. e.g. where the one was INT and the other was LONG
@changelog_1070_li
@changelog_1076_li
#Bug: Changes to the database structure did not result in the Session query cache being invalidated.
@changelog_1071_li
@changelog_1077_li
#New feature from Davide Cavestro - allow using custom Java object serialization engines on a per-DB basis.
@changelog_1072_li
@changelog_1078_li
#When running the Recover tool on very large (>6G) databases, some statistics were reported with negative numbers.
@changelog_1073_li
@changelog_1079_li
#Add a CONTAINS_UNCOMMITTED column to the SESSIONS metadata table, to allow detecting when rogue sessions are creating large transactions.
@changelog_1074_li
@changelog_1080_li
#Some small fixes to the GEOMETRY support, patches by Nicolas Fortin.
@changelog_1075_li
@changelog_1081_li
#The BNF tool and the autocomplete feature are exported in OSGi, which makes it possible for smart editors to do autocomplete. Patch from Nicolas Fortin.
@changelog_1076_li
@changelog_1082_li
#Fix DROP ALL OBJECTS and DROP SCHEMA in the presence of tables with computed column dependencies.
@changelog_1077_li
@changelog_1083_li
#Session-temporary LOB's could sometimes accumulate, increasing the size of the DB file until shutdown. Now they are cleared out at every commit.
@changelog_1078_li
@changelog_1084_li
#There was a bug where a hash index with more than one column would be silently converted to a regular index. It will now throw an exception.
@changelog_1079_li
@changelog_1085_li
#Query Statistics: new feature which stores the newest 100 SQL queries executed and their performance data. Useful for tracking down badly performing queries.
@changelog_1080_li
@changelog_1086_li
#Fix an LOB deadlock between reading and updating LOB columns.
@changelog_1081_li
@changelog_1087_li
#Support the JDBC DatabaseMetaData#getClientInfoProperties() method, in the sense of returning an empty result, in order to make WebSphere happy.
@changelog_1082_li
@changelog_1088_li
#Make Server#openBrowser respect the $BROWSER environment variable. Add "chromium" to the list of browsers tried. Patch from Felix Kaiser.
@changelog_1083_li
@changelog_1089_li
#Fix to org.h2.util.ScriptReader when handling unclosed block comments.
@changelog_1084_li
@changelog_1090_li
#Make org.h2.util.ScriptReader throw a better exception when handling broken scripts which generate extremely large statements.
@changelog_1085_li
@changelog_1091_li
#Fix bug with ALLOW_LITERALS=NONE, where the periodic analyze table on insert would throw an exception. A similar problem was fixed in the Console tool.
@changelog_1086_li
@changelog_1092_li
#Issue 510: Make org.h2.bnf public for consumption by external projects, patch by Nicolas Fortin
@changelog_1087_li
@changelog_1093_li
#Issue 509: Important fix on ValueGeometry, patch by Nicolas Fortin (with some tweaking) Make ValueGeometry#getDimensionCount more reliable. Add unit test to check for illegal ValueGeometry comparison Add unit test for conversion of Geometry object into Object Add optional export to MANIFEST.MF for JTS Geometry classes Validate that geometry values can be represented in WKB.
@changelog_1088_li
@changelog_1094_li
#Issue 506: RFE: Include Thread.getName() in case of a deadlock
@changelog_1089_li
@changelog_1095_li
#Adding support for "GRANT ALTER ANY SCHEMA TO <user>", patch by John Yates
@changelog_1090_h2
@changelog_1096_h2
#Version 1.3.173 (2013-07-28)
@changelog_1091_li
@changelog_1097_li
#Support empty statements that just contains a comment.
@changelog_1092_li
@changelog_1098_li
#Server mode: if there was an error while reading from a LOB, the session was closed in some cases.
@changelog_1093_li
@changelog_1099_li
#Issue 463: Driver name and version are now the same in OsgiDataSourceFactory and JdbcDatabaseMetaData.
@changelog_1094_li
@changelog_1100_li
#JaQu: The data type VARCHAR is now (again) used for Strings (no longer TEXT, except when explicitly set).
@changelog_1095_li
@changelog_1101_li
#For in-memory databases, creating an index on a CLOB or BLOB column is no longer supported. This is to simplify the MVTableEngine.
#Issue 468: trunc(timestamp) could return the wrong value (+12 hours), and trunc(number) throw a NullPointerException.
@changelog_1098_li
@changelog_1104_li
#The expression trunc(number) threw a NullPointerException.
@changelog_1099_li
@changelog_1105_li
#Fixed a deadlock when updating LOB's concurrently. See TestLob.testDeadlock2().
@changelog_1100_li
@changelog_1106_li
#Fixed a deadlock related to very large temporary result sets.
@changelog_1101_li
@changelog_1107_li
#Add "-list" command line option to Shell tool so that result-list-mode can be triggered when reading from a file.
@changelog_1102_li
@changelog_1108_li
#Issue 474: H2 MySQL Compatibility code fails to ignore "COMMENT" in CREATE TABLE, patch from Aaron Azeckoski.
@changelog_1103_li
@changelog_1109_li
#Issue 476: Broken link in jaqu.html
@changelog_1104_li
@changelog_1110_li
#Fix potential UTF8 encoding issue in org.h2.store.FileStore, reported by Juerg Spiess.
@changelog_1105_li
@changelog_1111_li
#Improve error message when check constraint is broken, test case from Gili (cowwoc).
@changelog_1106_li
@changelog_1112_li
#Improve error message when we have a unique constraint violation, displays the offending key in the error message.
@changelog_1107_li
@changelog_1113_li
#Issue 478: Support for "SHOW TRANSACTION ISOLATION LEVEL", patch from Andrew Franklin.
@changelog_1108_li
@changelog_1114_li
#Issue 475: PgServer: add support for CancelRequest, patch from Andrew Franklin.
@changelog_1109_li
@changelog_1115_li
#Issue 473: PgServer missing -key option, patch from Andrew Franklin.
@changelog_1110_li
@changelog_1116_li
#Issue 471: CREATE VIEW does not check user rights, patch from Andrew Franklin.
@changelog_1111_li
@changelog_1117_li
#Issue 477: PgServer binary transmission of query params is unimplemented, patch from Andrew Franklin.
@changelog_1112_li
@changelog_1118_li
#Issue 479: Support for SUBSTRING without a FROM condition, patch from Andrew Franklin.
@changelog_1113_li
@changelog_1119_li
#Issue 472: PgServer does not work with any recent Postgres JDBC driver, patch from Andrew Franklin.
@changelog_1114_li
@changelog_1120_li
#Add syntax for passing additional parameters into custom TableEngine implementations.
@changelog_1115_li
@changelog_1121_li
#Issue 480: Bugfix post issue 475, 477, patch from Andrew Franklin.
@changelog_1116_li
@changelog_1122_li
#Issue 481: Further extensions to PgServer to support better support PG JDBC, patch from Andrew Franklin.
@changelog_1117_li
@changelog_1123_li
#Add support for spatial datatype GEOMETRY.
@changelog_1118_li
@changelog_1124_li
#Add support for in-memory spatial index.
@changelog_1119_li
@changelog_1125_li
#change the PageStore#changeCount field from an int to a long, to cope with databases with very high transaction rates.
@changelog_1120_li
@changelog_1126_li
#Fix a NullPointerException when attempting to add foreign key reference to a view.
@changelog_1121_li
@changelog_1127_li
#Add sufficient ClientInfo support to our javax.sql.Connection implementation to make WebSphere happy.
@changelog_1122_li
@changelog_1128_li
#Issue 482: class LobStorageBackend$LobInputStream does not override the method InputStream.available().
@changelog_1123_li
@changelog_1129_li
#Fix corruption resulting from a mix of the "WRITE_DELAY=0" option and "SELECT DISTINCT" queries that don't fit in memory.
@changelog_1124_li
@changelog_1130_li
#Fix the combination of updating a table which contains an LOB, and reading from the LOB at the same time. Previously it would throw an exception, now it works.
@changelog_1125_li
@changelog_1131_li
#Issue 484: In the H2 Console tool, all schemas starting with "INFO" where hidden. Now they are hidden only if the database is not H2. Patch from "mgcodeact"/"cumer d"
@changelog_1126_li
@changelog_1132_li
#MySQL compatibility, support the "AUTO_INCREMENT=3" part of the CREATE TABLE statement.
@changelog_1127_li
@changelog_1133_li
#Issue 486: MySQL compatibility, support the "DEFAULT CHARSET" part of the CREATE TABLE statement.
@changelog_1128_li
@changelog_1134_li
#Issue 487: support the MySQL "SET foreign_key_checks = 0" command
@changelog_1129_li
@changelog_1135_li
#Issue 490: support MySQL "USING BTREE" index declaration
@changelog_1130_li
@changelog_1136_li
#Issue 485: Database get corrupted when column is renamed for which check constraint was defined inside create table statement.
@changelog_1131_li
@changelog_1137_li
#Issue 499: support MySQL "UNIQUE KEY (ID) USING BTREE" constraint syntax
@changelog_1132_li
@changelog_1138_li
#Issue 501: "CREATE TABLE .. WITH" not serialized, patch from nico.devel
@changelog_1133_li
@changelog_1139_li
#Avoid problems with runtime-compiled ALIAS methods when people have set the JAVA_TOOL_OPTIONS environment variable.
@changelog_1134_h2
@changelog_1140_h2
#Version 1.3.172 (2013-05-25)
@changelog_1135_li
@changelog_1141_li
#Referential integrity: when adding a referential integrity constraint failed, and if creating the constraint automatically created an index, this index was not removed.
@changelog_1136_li
@changelog_1142_li
#The auto-analyze feature now only reads 1000 rows per table instead of 10000.
@changelog_1137_li
@changelog_1143_li
#The optimization for IN(...) queries combined with OR could result in a strange exception of the type "column x must be included in the group by list".
@changelog_1138_li
@changelog_1144_li
#Issue 454: Use Charset for type-safety.
@changelog_1139_li
@changelog_1145_li
#Queries with both LIMIT and OFFSET could throw an IllegalArgumentException.
@changelog_1140_li
@changelog_1146_li
#MVStore: multiple issues were fixed: 460, 461, 462, 464, 466.
@changelog_1141_li
@changelog_1147_li
#MVStore: larger stores (multiple GB) are now much faster.
@changelog_1142_li
@changelog_1148_li
#When using local temporary tables and not dropping them manually before closing the session, and then killing the process could result in a database that couldn't be opened (except when using the recover tool).
@changelog_1143_li
@changelog_1149_li
#Support TRUNC(timestamp) for improved Oracle compatibility.
@changelog_1144_li
@changelog_1150_li
#Add support for CREATE TABLE TEST (ID BIGSERIAL) for PostgreSQL compatibility. Patch from Jesse Long.
@changelog_1145_li
@changelog_1151_li
#Add new collation command SET BINARY_COLLATION UNSIGNED, helps with people testing BINARY columns in MySQL mode.
@changelog_1146_li
@changelog_1152_li
#Issue 453: ABBA race conditions in TABLE LINK connection sharing.
@changelog_1147_li
@changelog_1153_li
#Issue 449: Postgres Serial data type should not automatically be marked as primary key
@changelog_1148_li
@changelog_1154_li
#Issue 406: Support "select h2version()"
@changelog_1149_li
@changelog_1155_li
#Issue 389: When there is a multi-column primary key, H2 does not seem to always pick the right index
@changelog_1150_li
@changelog_1156_li
#Issue 305: Implement SELECT ... FOR FETCH ONLY
@changelog_1151_li
@changelog_1157_li
#Issue 274: Sybase/MSSQLServer compatibility - Add GETDATE and CHARINDEX system functions
@changelog_1152_li
@changelog_1158_li
#Issue 274: Sybase/MSSQLServer compatibility - swap parameters of CONVERT function.
@changelog_1153_li
@changelog_1159_li
#Issue 274: Sybase/MSSQLServer compatibility - support index clause e.g. "select * from test (index table1_index)"
@changelog_1154_li
@changelog_1160_li
#Fix bug in Optimizing SELECT * FROM A WHERE X=1 OR X=2 OR X=3 into SELECT * FROM A WHERE X IN (1,2,3)
@changelog_1155_li
@changelog_1161_li
#Issue 442: Groovy patch for SourceCompiler (function ALIAS)
@changelog_1156_li
@changelog_1162_li
#Issue 459: Improve LOB documentation
@changelog_1157_h2
@changelog_1163_h2
#Version 1.3.171 (2013-03-17)
@changelog_1158_li
@changelog_1164_li
#Security: the TCP server did not correctly restrict access rights of clients in some cases. This was specially a problem when using the flag "tcpAllowOthers".
@changelog_1159_li
@changelog_1165_li
#H2 Console: the session timeout can now be configured using the system property "h2.consoleTimeout".
@changelog_1160_li
@changelog_1166_li
#Issue 431: Improved compatibility with MySQL: support for "ENGINE=InnoDB charset=UTF8" when creating a table.
@changelog_1161_li
@changelog_1167_li
#Issue 249: Improved compatibility with MySQL in the MySQL mode: now the methods DatabaseMetaData methods stores*Case*Identifiers return the same as MySQL when using the MySQL mode.
@changelog_1162_li
@changelog_1168_li
#Issue 434: H2 Console didn't work in the Chrome browser due to a wrong viewport argument.
@changelog_1163_li
@changelog_1169_li
#There was a possibility that the .lock.db file was not deleted when the database was closed, which could slow down opening the database.
@changelog_1164_li
@changelog_1170_li
#The SQL script generated by the "script" command contained inconsistent newlines on Windows.
@changelog_1165_li
@changelog_1171_li
#When using trace level 4 (SLF4J) in the server mode, a directory "trace.db" and an empty file was created on the client side. This is no longer made.
@changelog_1166_li
@changelog_1172_li
#Optimize IN(...) queries: there was a bug in version 1.3.170 if the type of the left hand side didn't match the type of the right hand side. Fixed.
@changelog_1167_li
@changelog_1173_li
#Optimize IN(...) queries: there was a bug in version 1.3.170 for comparison of the type "X IN(NULL, NULL)". Fixed.
@changelog_1168_li
@changelog_1174_li
#Timestamps with timezone that were passed as a string were not always converted correctly. For example "2012-11-06T23:00:00.000Z" was converted to "2012-11-06" instead of to "2012-11-07" in the timezone CET. Thanks a lot to Steve Hruda for reporting the problem!
@changelog_1169_li
@changelog_1175_li
#New table engine "org.h2.mvstore.db.MVTableEngine" that internally uses the MVStore to persist data. To try it out, append ";DEFAULT_TABLE_ENGINE=org.h2.mvstore.db.MVTableEngine" to the database URL. This is still very experimental, and many features are not supported yet. The data is stored in a file with the suffix ".mv.db".
@changelog_1170_li
@changelog_1176_li
#New connection setting "DEFAULT_TABLE_ENGINE" to use a specific table engine if none is set explicitly. This is to simplify testing the MVStore table engine.
@changelog_1171_li
@changelog_1177_li
#MVStore: encrypted stores are now supported. Only standardized algorithms are used: PBKDF2, SHA-256, XTS-AES, AES-128.
@changelog_1172_li
@changelog_1178_li
#MVStore: improved API thanks to Simo Tripodi.
@changelog_1173_li
@changelog_1179_li
#MVStore: maps can now be renamed.
@changelog_1174_li
@changelog_1180_li
#MVStore: store the file header also at the end of each chunk, which results in a further reduced number of write operations.
@changelog_1175_li
@changelog_1181_li
#MVStore: a map implementation that supports concurrent operations.
@changelog_1176_li
@changelog_1182_li
#MVStore: unified exception handling; the version is included in the messages.
@changelog_1177_li
@changelog_1183_li
#MVStore: old data is now retained for 45 seconds by default.
@changelog_1178_li
@changelog_1184_li
#MVStore: compress is now disabled by default, and can be enabled on request.
@changelog_1179_li
@changelog_1185_li
#Support ALTER TABLE ADD ... AFTER. Patch from Andrew Gaul (argaul at gmail.com). Fixes issue 401.
@changelog_1180_li
@changelog_1186_li
#Improved OSGi support. H2 now registers itself as a DataSourceFactory service. Fixes issue 365.
@changelog_1181_li
@changelog_1187_li
#Add a DISK_SPACE_USED system function. Fixes issue 270.
@changelog_1182_li
@changelog_1188_li
#Fix a compile-time ambiguity when compiling with JDK7, thanks to a patch from Lukas Eder.
@changelog_1183_li
@changelog_1189_li
#Supporting dropping an index for Lucene full-text indexes.
@changelog_1184_li
@changelog_1190_li
#Optimized performance for SELECT ... ORDER BY X LIMIT Y OFFSET Z queries for in-memory databases using partial sort (by Sergi Vladykin).
@changelog_1185_li
@changelog_1191_li
#Experimental off-heap memory storage engine "nioMemFS:" and "nioMemLZF:", suggestion from Mark Addleman.
@changelog_1186_li
@changelog_1192_li
#Issue 438: JdbcDatabaseMetaData.getSchemas() is no longer supported as of 1.3.169.
@changelog_1187_li
@changelog_1193_li
#MySQL compatibility: support for ALTER TABLE tableName MODIFY [COLUMN] columnName columnDef. Patch from Ville Koskela.
@changelog_1188_li
@changelog_1194_li
#Issue 404: SHOW COLUMNS FROM tableName does not work with ALLOW_LITERALS=NUMBERS.
@changelog_1189_li
@changelog_1195_li
#Throw an explicit error to make it clear we don't support the TRIGGER combination of SELECT and FOR EACH ROW.
@changelog_1190_li
@changelog_1196_li
#Issue 439: Utils.sortTopN does not handle single-element arrays.
@changelog_1191_h2
@changelog_1197_h2
#Version 1.3.170 (2012-11-30)
@changelog_1192_li
@changelog_1198_li
#Issue 407: The TriggerAdapter didn't work with CLOB and BLOB columns.
@changelog_1193_li
@changelog_1199_li
#PostgreSQL compatibility: support for data types BIGSERIAL and SERIAL as an alias for AUTO_INCREMENT.
@changelog_1194_li
@changelog_1200_li
#Issue 417: H2 Console: the web session timeout didn't work, resulting in a memory leak. This was only a problem if the H2 Console was run for a long time and many sessions were opened.
@changelog_1195_li
@changelog_1201_li
#Issue 412: Running the Server tool with just the option "-browser" will now log a warning.
@changelog_1196_li
@changelog_1202_li
#Issue 411: CloseWatcher registration was not concurrency-safe.
@changelog_1197_li
@changelog_1203_li
#MySQL compatibility: support for CONCAT_WS. Thanks a lot to litailang for the patch!
@changelog_1198_li
@changelog_1204_li
#PostgreSQL compatibility: support for EXTRACT(WEEK FROM dateColumn). Thanks to Prashant Bhat for the patch!
@changelog_1199_li
@changelog_1205_li
#Fix for a bug where we would sometimes use the wrong unique constraint to validate foreign key constraints.
@changelog_1200_li
@changelog_1206_li
#Support BOM at the beginning of files for the RUNSCRIPT command
@changelog_1201_li
@changelog_1207_li
#Fix in calling SET @X = IDENTITY() where it would return NULL incorrectly
@changelog_1202_li
@changelog_1208_li
#Fix ABBA deadlock between adding a constraint and the H2-Log-Writer thread.
@changelog_1203_li
@changelog_1209_li
#Optimize IN(...) queries where the values are constant and of the same type.
@changelog_1204_li
@changelog_1210_li
#Restore tool: the parameter "quiet" was not used and is now removed.
@changelog_1205_li
@changelog_1211_li
#Fix ConcurrentModificationException when creating tables and executing SHOW TABLES in parallel. Reported by Viktor Voytovych.
@changelog_1206_li
@changelog_1212_li
#Serialization is now pluggable using the system property "h2.javaObjectSerializer". Thanks to Sergi Vladykin for the patch!
@changelog_1207_h2
@changelog_1213_h2
#Version 1.3.169 (2012-09-09)
@changelog_1208_li
@changelog_1214_li
#The default jar file is now compiled for Java 6.
@changelog_1209_li
@changelog_1215_li
#The new jar file will probably not end up in the central Maven repository in the next few weeks because Sonatype has disabled automatic synchronization from SourceForge (which they call 'legacy sync' now). It will probably take some time until this is sorted out. The H2 jar files are deployed to http://h2database.com/m2-repo/com/h2database/h2/maven-metadata.xml and http://hsql.sourceforge.net/m2-repo/com/h2database/h2/maven-metadata.xml as usual.
@changelog_1210_li
@changelog_1216_li
#A part of the documentation and the H2 Console has been changed to support the Apple retina display.
@changelog_1211_li
@changelog_1217_li
#The CreateCluster tool could not be used if the source database contained a CLOB or BLOB. The root cause was that the TCP server did not synchronize on the session, which caused a problem when using the exclusive mode.
@changelog_1212_li
@changelog_1218_li
#Statement.getQueryTimeout(): only the first call to this method will query the database. If the query timeout was changed in another way than calling setQueryTimeout, this method will always return the last value. This was changed because Hibernate calls getQueryTimeout() a lot.
@changelog_1213_li
@changelog_1219_li
#Issue 416: PreparedStatement.setNString throws AbstractMethodError. All implemented JDBC 4 methods that don't break compatibility with Java 5 are now included in the default jar file.
@changelog_1214_li
@changelog_1220_li
#Issue 414: for some functions, the parameters were evaluated twice (for example "char(nextval(..))" ran "nextval(..)" twice).
@changelog_1215_li
@changelog_1221_li
#The ResultSetMetaData methods getSchemaName and getTableName could return null instead of "" (an empty string) as specified in the JDBC API.
@changelog_1216_li
@changelog_1222_li
#Added compatibility for "SET NAMES" query in MySQL compatibility mode.
@changelog_1217_h2
@changelog_1223_h2
#Version 1.3.168 (2012-07-13)
@changelog_1218_li
@changelog_1224_li
#The message "Transaction log could not be truncated" was sometimes written to the .trace.db file even if there was no problem truncating the transaction log.
@changelog_1219_li
@changelog_1225_li
#New system property "h2.serializeJavaObject" (default: true) that allows to disable serializing Java objects, so that the objects compareTo and toString methods can be used.
@changelog_1220_li
@changelog_1226_li
#Dylan has translated the H2 Console tool to Korean. Thanks a lot!
@changelog_1221_li
@changelog_1227_li
#Executing the statement CREATE INDEX IF ALREADY EXISTS if the index already exists no longer fails for a read only database.
@changelog_1222_li
@changelog_1228_li
#MVCC: concurrently updating a row could result in the row to appear deleted in the second connection, if there are multiple unique indexes (or a primary key and at least one unique index). Thanks a lot to Teruo for the patch!
@changelog_1223_li
@changelog_1229_li
#Fulltext search: in-memory Lucene indexes are now supported.
@changelog_1224_li
@changelog_1230_li
#Fulltext search: UUID primary keys are now supported.
@changelog_1225_li
@changelog_1231_li
#Apache Tomcat 7.x will now longer log a warning when unloading the web application, if using a connection pool.
@changelog_1226_li
@changelog_1232_li
#H2 Console: support the Midori browser (for Debian / Raspberry Pi)
@changelog_1227_li
@changelog_1233_li
#When opening a remote session, don't open a temporary file if the trace level is set to zero
@changelog_1228_li
@changelog_1234_li
#Use HMAC for authenticating remote LOB id's, removing the need for maintaining a cache, and removing the limit on the number of LOBs per result set.
@changelog_1229_li
@changelog_1235_li
#H2 Console: HTML and XML documents can now be edited in an updatable result set. There is (limited) support for editing multi-line documents.
@cheatSheet_1000_h1
...
...
@@ -7529,195 +7547,225 @@ H2 データベース エンジン
#File Format
@mvstore_1118_p
# The data is stored in one file. The file contains two file headers (to be safe), and a number of chunks. The file headers are one block each; a block is 4096 bytes. Each chunk is at least one block, but typically 200 blocks or more. There might be a number of free blocks in front of every chunk. There is one chunk for every version.
# The data is stored in one file. The file contains two file headers (for safety), and a number of chunks. The file headers are one block each; a block is 4096 bytes. Each chunk is at least one block, but typically 200 blocks or more. Data is stored in the chunks in the form of a <a href="https://en.wikipedia.org/wiki/Log-structured_file_system">log structured storage</a>. There is one chunk for every version.
@mvstore_1119_h3
#File Header
@mvstore_1119_p
# Each chunk contains a number of B-tree pages. As an example, the following code:
@mvstore_1120_p
# will result in the following two chunks (excluding metadata):
# - Page 3: (root) node with 2 entries pointing to page 1 and 2
@mvstore_1125_b
#Chunk 2:
@mvstore_1126_p
# - Page 4: leaf with 140 entries (keys 0 - 139)
@mvstore_1127_p
# - Page 5: (root) node with 2 entries pointing to page 4 and 2
@mvstore_1128_p
# That means each chunk contains the changes of one version: the new version of the changed pages and the parent pages, recursively, up to the root page. Pages in subsequent chunks refer to pages in earlier chunks.
@mvstore_1129_h3
#File Header
@mvstore_1130_p
# There are two file headers, which normally contain the exact same data. But once in a while, the file headers are updated, and writing could partially fail, which could corrupt a header. That's why there is a second header. Only the file headers are updated in this way (called "in-place update"). The headers contain the following data:
@mvstore_1121_p
@mvstore_1131_p
# The data is stored in the form of a key-value pair. Each value is stored as a hexadecimal number. The entries are:
@mvstore_1122_li
@mvstore_1132_li
#H: The entry "H:2" stands for the the H2 database.
@mvstore_1123_li
@mvstore_1133_li
#block: The block number where one of the newest chunks starts (but not necessarily the newest).
@mvstore_1124_li
@mvstore_1134_li
#blockSize: The block size of the file; currently always hex 1000, which is decimal 4096, to match the <a href="https://en.wikipedia.org/wiki/Disk_sector">disk sector</a> length of modern hard disks.
@mvstore_1125_li
@mvstore_1135_li
#chunk: The chunk id, which is normally the same value as the version; however, the chunk id might roll over to 0, while the version doesn't.
@mvstore_1126_li
@mvstore_1136_li
#created: The number of milliseconds since 1970 when the file was created.
@mvstore_1127_li
@mvstore_1137_li
#format: The file format number. Currently 1.
@mvstore_1128_li
@mvstore_1138_li
#version: The version number of the chunk.
@mvstore_1129_li
@mvstore_1139_li
#fletcher: The <a href="https://en.wikipedia.org/wiki/Fletcher's_checksum"> Fletcher-32 checksum</a> of the header.
@mvstore_1130_p
@mvstore_1140_p
# When opening the file, both headers are read and the checksum is verified. If both headers are valid, the one with the newer version is used. The chunk with the latest version is then detected (details about this see below), and the rest of the metadata is read from there. If the chunk id, block and version are not stored in the file header, then the latest chunk lookup starts with the last chunk in the file.
@mvstore_1131_h3
@mvstore_1141_h3
#Chunk Format
@mvstore_1132_p
@mvstore_1142_p
# There is one chunk per version. Each chunk consists of a header, the pages that were modified in this version, and a footer. The pages contain the actual data of the maps. The pages inside a chunk are stored right after the header, next to each other (unaligned). The size of a chunk is a multiple of the block size. The footer is stored in the last 128 bytes of the chunk.
@mvstore_1133_p
@mvstore_1143_p
# The footer allows to verify that the chunk is completely written (a chunk is written as one write operation), and allows to find the start position of the very last chunk in the file. The chunk header and footer contain the following data:
@mvstore_1134_p
@mvstore_1144_p
# The fields of the chunk header and footer are:
@mvstore_1135_li
@mvstore_1145_li
#chunk: The chunk id.
@mvstore_1136_li
@mvstore_1146_li
#block: The first block of the chunk (multiply by the block size to get the position in the file).
@mvstore_1137_li
@mvstore_1147_li
#len: The size of the chunk in number of blocks.
@mvstore_1138_li
@mvstore_1148_li
#map: The id of the newest map; incremented when a new map is created.
@mvstore_1139_li
@mvstore_1149_li
#max: The sum of all maximum page sizes (see page format).
@mvstore_1140_li
@mvstore_1150_li
#next: The predicted start block of the next chunk.
@mvstore_1141_li
@mvstore_1151_li
#pages: The number of pages in the chunk.
@mvstore_1142_li
@mvstore_1152_li
#root: The position of the metadata root page (see page format).
@mvstore_1143_li
@mvstore_1153_li
#time: The time the chunk was written, in milliseconds after the file was created.
@mvstore_1144_li
@mvstore_1154_li
#version: The version this chunk represents.
@mvstore_1145_li
@mvstore_1155_li
#fletcher: The checksum of the footer.
@mvstore_1146_p
# Chunks are never updated in-place. Each chunk contains the pages that were changed in that version (there is one chunk per version, see above), plus all the parent nodes of those pages, recursively, up to the root page. If an entry in a map is changed, removed, or added, then the respective page is copied to be stored in the next chunk, and the number of live pages in the old chunk is decremented. This mechanism is called copy-on-write, and is similar to how the <a href="https://en.wikipedia.org/wiki/Btrfs">Btrfs</a> file system works. Chunks without live pages are marked as free, so the space can be re-used by more recent chunks. Because not all chunks are of the same size, there can be some unused space in front of a chunk for some time (until a small chunk is written or the chunks are compacted). There is a <a href="http://stackoverflow.com/questions/13650134/after-how-many-seconds-are-file-system-write-buffers-typically-flushed"> delay of 45 seconds</a> (by default) before a free chunk is overwritten, to ensure new versions are persisted first, as hard disks sometimes re-order write operations.
@mvstore_1156_p
# Chunks are never updated in-place. Each chunk contains the pages that were changed in that version (there is one chunk per version, see above), plus all the parent nodes of those pages, recursively, up to the root page. If an entry in a map is changed, removed, or added, then the respective page is copied to be stored in the next chunk, and the number of live pages in the old chunk is decremented. This mechanism is called copy-on-write, and is similar to how the <a href="https://en.wikipedia.org/wiki/Btrfs">Btrfs</a> file system works. Chunks without live pages are marked as free, so the space can be re-used by more recent chunks. Because not all chunks are of the same size, there can be a number of free blocks in front of a chunk for some time (until a small chunk is written or the chunks are compacted). There is a <a href="http://stackoverflow.com/questions/13650134/after-how-many-seconds-are-file-system-write-buffers-typically-flushed"> delay of 45 seconds</a> (by default) before a free chunk is overwritten, to ensure new versions are persisted first.
@mvstore_1147_p
@mvstore_1157_p
# How the newest chunk is located when opening a store: The file header contains the position of a recent chunk, but not always the newest one. This is to reduce the number of file header updates. After opening the file, the file headers, and the chunk footer of the very last chunk (at the end of the file) are read. From those candidates, the header of the most recent chunk is read. If it contains a "next" pointer (see above), those chunk's header and footer are read as well. If it turned out to be a newer valid chunk, this is repeated, until the newest chunk was found. Before writing a chunk, the position of the next chunk is predicted based on the assumption that the next chunk will be of the same size as the current one. When the next chunk is written, and the previous prediction turned out to be incorrect, the file header is updated as well. In any case, the file header is updated if the next chain gets longer than 20 hops.
@mvstore_1148_h3
@mvstore_1158_h3
#Page Format
@mvstore_1149_p
@mvstore_1159_p
# Each map is a <a href="https://en.wikipedia.org/wiki/B-tree">B-tree</a>, and the map data is stored in (B-tree-) pages. There are leaf pages that contain the key-value pairs of the map, and internal nodes, which only contain keys and pointers to leaf pages. The root of a tree is either a leaf or an internal node. Unlike file header and chunk header and footer, the page data is not human readable. Instead, it is stored as byte arrays, with long (8 bytes), int (4 bytes), short (2 bytes), and <a href="https://en.wikipedia.org/wiki/Variable-length_quantity">variable size int and long</a> (1 to 5 / 10 bytes). The page format is:
@mvstore_1150_li
@mvstore_1160_li
#length (int): Length of the page in bytes.
@mvstore_1151_li
@mvstore_1161_li
#checksum (short): Checksum (chunk id xor offset within the chunk xor page length).
@mvstore_1152_li
@mvstore_1162_li
#mapId (variable size int): The id of the map this page belongs to.
@mvstore_1153_li
@mvstore_1163_li
#len (variable size int): The number of keys in the page.
@mvstore_1154_li
@mvstore_1164_li
#type (byte): The page type (0 for leaf page, 1 for internal node; plus 2 if the page data is compressed).
@mvstore_1155_li
#keys (byte array): All keys, stored depending on the data type.
@mvstore_1156_li
@mvstore_1165_li
#children (array of long; internal nodes only): The position of the children.
@mvstore_1157_li
@mvstore_1166_li
#childCounts (array of variable size long; internal nodes only): The total number of entries for the given child page.
@mvstore_1158_li
@mvstore_1167_li
#keys (byte array): All keys, stored depending on the data type.
@mvstore_1168_li
#values (byte array; leaf pages only): All values, stored depending on the data type.
@mvstore_1159_p
@mvstore_1169_p
# Even though this is not required by the file format, each B-tree is stored "upside down", that means the leaf pages first, then the internal nodes, and lastly the root page.
@mvstore_1160_p
# Pointers to pages are stored as a long, using a special format: 26 bits for the chunk id, 32 bits for the offset within the chunk, 5 bits for the length code, 1 bit for the page type (leaf or internal node). The page type is encoded so that when clearing or removing a map, leaf pages don't have to be read (internal nodes do have to be read in order to know where all the pages are; but in a typical B-tree the vast majority of the pages are leaf pages). The absolute file position is not included so that chunks can be moved within the file without having to change page pointers; only the chunk metadata needs to be changed. The length code is a number from 0 to 31, where 0 means the maximum length of the page is 32 bytes, 1 means 48 bytes, 2: 64, 3: 96, 4: 128, 5: 192, and so on until 31 which means longer than 1 MB. That way, reading a page only requires one read operation (except for very large pages, in which case two read operations might be needed). The sum of the maximum length of all pages is stored in the chunk metadata (field "max"), and when a page is marked as removed, the live maximum length is adjusted. This allows to estimate the amount of free space within a block, in addition to the number of free pages.
@mvstore_1170_p
# Pointers to pages are stored as a long, using a special format: 26 bits for the chunk id, 32 bits for the offset within the chunk, 5 bits for the length code, 1 bit for the page type (leaf or internal node). The page type is encoded so that when clearing or removing a map, leaf pages don't have to be read (internal nodes do have to be read in order to know where all the pages are; but in a typical B-tree the vast majority of the pages are leaf pages). The absolute file position is not included so that chunks can be moved within the file without having to change page pointers; only the chunk metadata needs to be changed. The length code is a number from 0 to 31, where 0 means the maximum length of the page is 32 bytes, 1 means 48 bytes, 2: 64, 3: 96, 4: 128, 5: 192, and so on until 31 which means longer than 1 MB. That way, reading a page only requires one read operation (except for very large pages). The sum of the maximum length of all pages is stored in the chunk metadata (field "max"), and when a page is marked as removed, the live maximum length is adjusted. This allows to estimate the amount of free space within a block, in addition to the number of free pages.
@mvstore_1161_p
@mvstore_1171_p
# The total number of entries in a child nodes is kept to allow efficient range counting, lookup by index, and skip operations. The pages form a <a href="http://www.chiark.greenend.org.uk/~sgtatham/algorithms/cbtree.html">counted B-tree</a>.
@mvstore_1162_p
@mvstore_1172_p
# Data compression: The data after the page type are optionally compressed using the LZF algorithm.
@mvstore_1163_h3
@mvstore_1173_h3
#Metadata Map
@mvstore_1164_p
@mvstore_1174_p
# In addition to the user maps, there is one metadata map that contains names and positions of user maps, and chunk metadata. The very last page of a chunk contains the root page of that metadata map. The exact position of this root page is stored in the chunk header. This page (directly or indirectly) points to the root pages of all other maps. The metadata map of a store with a map named "data", and one chunk, contains the following entries:
@mvstore_1165_li
@mvstore_1175_li
#chunk.1: The metadata of chunk 1. This is the same data as the chunk header, plus the number of live pages, and the maximum live length.
@mvstore_1166_li
#setting.storeVersion: The store version (a user defined value).
@mvstore_1167_li
@mvstore_1176_li
#map.1: The metadata of map 1. The entries are: name, createVersion, and type.
@mvstore_1168_li
@mvstore_1177_li
#name.data: The map id of the map named "data". The value is "1".
@mvstore_1169_li
@mvstore_1178_li
#root.1: The root position of map 1.
@mvstore_1170_h2
@mvstore_1179_li
#setting.storeVersion: The store version (a user defined value).
@mvstore_1180_h2
#Similar Projects and Differences to Other Storage Engines
@mvstore_1171_p
@mvstore_1181_p
# Unlike similar storage engines like LevelDB and Kyoto Cabinet, the MVStore is written in Java and can easily be embedded in a Java and Android application.
@mvstore_1172_p
@mvstore_1182_p
# The MVStore is somewhat similar to the Berkeley DB Java Edition because it is also written in Java, and is also a log structured storage, but the H2 license is more liberal.
@mvstore_1173_p
@mvstore_1183_p
# Like SQLite 3, the MVStore keeps all data in one file. Unlike SQLite 3, the MVStore uses is a log structured storage. The plan is to make the MVStore both easier to use as well as faster than SQLite 3. In a recent (very simple) test, the MVStore was about twice as fast as SQLite 3 on Android.
@mvstore_1174_p
@mvstore_1184_p
# The API of the MVStore is similar to MapDB (previously known as JDBM) from Jan Kotek, and some code is shared between MVStore and MapDB. However, unlike MapDB, the MVStore uses is a log structured storage. The MVStore does not have a record size limit.
@mvstore_1175_h2
@mvstore_1185_h2
#Current State
@mvstore_1176_p
@mvstore_1186_p
# The code is still experimental at this stage. The API as well as the behavior may partially change. Features may be added and removed (even though the main features will stay).
@mvstore_1177_h2
@mvstore_1187_h2
必�?�?�件
@mvstore_1178_p
@mvstore_1188_p
# The MVStore is included in the latest H2 jar file.
@mvstore_1179_p
@mvstore_1189_p
# There are no special requirements to use it. The MVStore should run on any JVM as well as on Android.
@mvstore_1180_p
@mvstore_1190_p
# To build just the MVStore (without the database engine), run:
@mvstore_1181_p
@mvstore_1191_p
# This will create the file <code>bin/h2mvstore-1.3.175.jar</code> (about 130 KB).
@@ -621,234 +621,240 @@ build_1111_li=The rail images (one straight, four junctions, two turns) are gene
build_1112_p=\ To generate railroad diagrams for other grammars, see the package <code>org.h2.jcr</code>. This package is used to generate the SQL-2 railroad diagrams for the JCR 2.0 specification.
changelog_1000_h1=Change Log
changelog_1001_h2=Next Version (unreleased)
changelog_1002_li=Referential integrity constraints sometimes used the wrong index, such that updating a row in the referenced table incorrectly failed with a constraint violation.
changelog_1003_li=The Polish translation was completed and corrected by Wojtek Jurczyk. Thanks a lot\!
changelog_1004_li=Issue 545\:Unnecessary duplicate code was removed.
changelog_1005_li=The profiler tool can now process files with full thread dumps.
changelog_1006_li=MVStore\:the file format was changed slightly.
changelog_1007_li=MVStore mode\:the CLOB and BLOB storage was re-implemented and is now much faster than with the PageStore (which is still the default storage).
changelog_1008_li=MVStore mode\:creating indexes is now much faster (in many cases faster than with the default PageStore).
changelog_1009_li=Various bugs in the MVStore storage and have been fixed, including a bug in the R-tree implementation. The database could get corrupt if there were transient IO exceptions while storing.
changelog_1010_li=The method org.h2.expression.Function.getCost could throw a NullPointException.
changelog_1011_li=Storing LOBs in separate files (outside of the main database file) is no longer supported for new databases.
changelog_1012_li=Lucene 2 is no longer supported.
changelog_1013_li=Fix bug in calculating default MIN and MAX values for SEQUENCE.
changelog_1014_li=Fix bug in performing IN queries with multiple values when IGNORECASE\=TRUE
changelog_1015_h2=Version 1.3.175 (2014-01-18)
changelog_1016_li=EXPLAIN was incorrect for queries with "ANY" or "SOME" conditions.
changelog_1017_li=CallableStatement with "out" parameters\:running the same statement twice could result in an exception ("parameter not set").
changelog_1018_li=Union queries\:duplicate rows could be returned if the sub-queries contained "order by".
changelog_1019_li=The GEOMETRY data type now works for user defined functions that return a result set.
changelog_1020_li=PostgreSQL compatibility\:the PgServer was not working properly when the setting database_to_upper was set to false.
changelog_1021_li=JdbcDataSource\:the methods setUrl and getUrl where added as aliases for setURL and getURL. This should solve problems with the HikariCP (Hikari connection pool).
changelog_1022_li=Issue 537\:H2 Console\:when loading the schema, incorrect JDBC calls where issued, which caused the exception "Column PRECISION not found".
changelog_1024_li=The collation can now be set in the database URL, even if there are data tables, if the collection is the same as the current collation.
changelog_1025_li=Improved Oracle compatibility for CASE WHEN and DECODE.
changelog_1026_li=The statement "drop all objects" did not work if a table depends on a view via a constraint.
changelog_1027_li=Subqueries or views with "order by" an alias expression could not be executed due to a regression introduced in version 1.3.174.
changelog_1028_li=Issue 73\:MySQL compatibility\:support REPLACE, patch by Cemo Koc.
changelog_1029_li=The spatial index now works in MVCC mode when using the MVStore storage.
changelog_1030_li=MVStore\:concurrency problems have been fixed. The API has been simplified.
changelog_1031_li=Improve error message when dropping an index that belongs to a constraint, specify constraint in error message.
changelog_1032_li=Issue 518\:java.sql.Connection.commit() freezes after LOB modification with EXCLUSIVE connection
changelog_1033_li=Issue 517\:Create or replace view statement has no effect on the others already existing JDBC connection
changelog_1034_li=Support 123L syntax as in Java; example\:SELECT (2000000000L*2).
changelog_1035_li=Issue 520\:Add support for sequence max value, min value and cycle, patch by Daniel Gredler.
changelog_1036_li=Issue 521\:ScriptReader should implement Closeable
changelog_1037_li=Issue 524\:RunScript.execute does not close its Statement, patch from Gaul.
changelog_1038_li=Add support for DB2 "WITH UR" clause, patch from litailang
changelog_1039_li=Added support for ON DUPLICATE KEY UPDATE like MySQL with the values() function to update with the value that was to be inserted. Patch from Jean-Francois Noel.
changelog_1040_li=Issue 522\:Treat empty strings like NULL in Oracle compatibility mode, patch by Daniel Gredler.
changelog_1041_li=Issue 527\:Oracle compatibility mode\:incorrect scale behavior, patch by Daniel Gredler.
changelog_1042_li=Slightly reduce the memory cost of View metadata.
changelog_1043_li=Extend support of "GRANT ALTER ANY SCHEMA TO <user>" to allow grantee ability to manipulate tables
changelog_1044_li=Issue 532\:Javadoc for ErrorCode.ROLES_AND_RIGHT_CANNOT_BE_MIXED looks wrong
changelog_1045_li=Issue 528\:Add Oracle-compatible TO_CHAR function, patch by Daniel Gredler.
changelog_1046_li=Issue 534\:CREATE ALIAS NOCACHE, Patch by Nicolas Fortin
changelog_1047_li=Fix an issue with storing Unicode surrogate pairs in CLOB columns.
changelog_1049_li=Fix bug in DB2 syntax "select * from test with ur", patch from litailang
changelog_1050_li=Fix bug in DROP ALL OBJECTS when dealing with inter-schema dependencies.
changelog_1051_h2=Version 1.3.174 (2013-10-19)
changelog_1052_li=LIRS cache\:bugfix for caches that only contain non-resident entries.
changelog_1053_li=For in-memory databases, queries with a "group by" column that is also a hash index threw a RuntimeException.
changelog_1054_li=Improved error message for some syntax errors.
changelog_1055_li=File system abstraction\:if used directly, some file systems did not work correctly with spliced byte buffers (the database engine doesn't use those).
changelog_1056_li=To use the MVStore storage engine (which is still work in progress), append ";mv_store\=true" to the database URL. Using the MVTableEngine when creating the table is no longer recommended.
changelog_1057_li=To compile user defined functions, the javax.tools.JavaCompiler is now used if available, and no temporary files are created. This should solve problems when multiple H2 database concurrently compile the same user defined functions. To disable, system the system property "h2.javaSystemCompiler" to false.
changelog_1058_li=Concurrently creating function aliases in different databases could result in the exception "javac\:file not found".
changelog_1059_li=The function "regexp_replace" threw the wrong kind of exception if the replacement string was invalid.
changelog_1060_li=A checkpoint is now done every MAX_LOG_SIZE / 2 instead of every MAX_LOG_SIZE, so that the transaction log doesn't grow as large.
changelog_1061_li=MVStore table engine\:new setting "retention_time" to configure the time to retain old data. The default is 45 seconds.
changelog_1062_li=The method TableEngine.createTable() now returns a Table object.
changelog_1063_li=For read-only databases, for the trace level "debug", the trace info is written to the temp directory.
changelog_1064_li=Closing the file lock will now wait until the background thread is stopped.
changelog_1065_li=In version 1.3.172, a performance regression was introduced when fixing the issue 389 (when there is a multi-column primary key, H2 does not seem to always pick the right index). This was related to boosting an index that matches the "order by" column list (the wrong index was used in some cases).
changelog_1066_li=Improved spatial index and data type.
changelog_1067_li=Issue 467\:OSGi Class Loader (ability to create reference to class in other ClassLoader, for example in another OSGi bundle).
changelog_1068_li=Queries "group by" that contain a subquery with an aggregate function returned the wrong result in some cases.
changelog_1069_li=Fix bug in unique and non-unique hash indexes which manifested as incorrect results when the search key was a different cardinal type from the table index key. e.g. where the one was INT and the other was LONG
changelog_1070_li=Bug\:Changes to the database structure did not result in the Session query cache being invalidated.
changelog_1071_li=New feature from Davide Cavestro - allow using custom Java object serialization engines on a per-DB basis.
changelog_1072_li=When running the Recover tool on very large (>6G) databases, some statistics were reported with negative numbers.
changelog_1073_li=Add a CONTAINS_UNCOMMITTED column to the SESSIONS metadata table, to allow detecting when rogue sessions are creating large transactions.
changelog_1074_li=Some small fixes to the GEOMETRY support, patches by Nicolas Fortin.
changelog_1075_li=The BNF tool and the autocomplete feature are exported in OSGi, which makes it possible for smart editors to do autocomplete. Patch from Nicolas Fortin.
changelog_1076_li=Fix DROP ALL OBJECTS and DROP SCHEMA in the presence of tables with computed column dependencies.
changelog_1077_li=Session-temporary LOB's could sometimes accumulate, increasing the size of the DB file until shutdown. Now they are cleared out at every commit.
changelog_1078_li=There was a bug where a hash index with more than one column would be silently converted to a regular index. It will now throw an exception.
changelog_1079_li=Query Statistics\:new feature which stores the newest 100 SQL queries executed and their performance data. Useful for tracking down badly performing queries.
changelog_1080_li=Fix an LOB deadlock between reading and updating LOB columns.
changelog_1081_li=Support the JDBC DatabaseMetaData\#getClientInfoProperties() method, in the sense of returning an empty result, in order to make WebSphere happy.
changelog_1082_li=Make Server\#openBrowser respect the $BROWSER environment variable. Add "chromium" to the list of browsers tried. Patch from Felix Kaiser.
changelog_1083_li=Fix to org.h2.util.ScriptReader when handling unclosed block comments.
changelog_1084_li=Make org.h2.util.ScriptReader throw a better exception when handling broken scripts which generate extremely large statements.
changelog_1085_li=Fix bug with ALLOW_LITERALS\=NONE, where the periodic analyze table on insert would throw an exception. A similar problem was fixed in the Console tool.
changelog_1086_li=Issue 510\:Make org.h2.bnf public for consumption by external projects, patch by Nicolas Fortin
changelog_1087_li=Issue 509\:Important fix on ValueGeometry, patch by Nicolas Fortin (with some tweaking) Make ValueGeometry\#getDimensionCount more reliable. Add unit test to check for illegal ValueGeometry comparison Add unit test for conversion of Geometry object into Object Add optional export to MANIFEST.MF for JTS Geometry classes Validate that geometry values can be represented in WKB.
changelog_1088_li=Issue 506\:RFE\:Include Thread.getName() in case of a deadlock
changelog_1089_li=Adding support for "GRANT ALTER ANY SCHEMA TO <user>", patch by John Yates
changelog_1090_h2=Version 1.3.173 (2013-07-28)
changelog_1091_li=Support empty statements that just contains a comment.
changelog_1092_li=Server mode\:if there was an error while reading from a LOB, the session was closed in some cases.
changelog_1093_li=Issue 463\:Driver name and version are now the same in OsgiDataSourceFactory and JdbcDatabaseMetaData.
changelog_1094_li=JaQu\:The data type VARCHAR is now (again) used for Strings (no longer TEXT, except when explicitly set).
changelog_1095_li=For in-memory databases, creating an index on a CLOB or BLOB column is no longer supported. This is to simplify the MVTableEngine.
changelog_1097_li=Issue 468\:trunc(timestamp) could return the wrong value (+12 hours), and trunc(number) throw a NullPointerException.
changelog_1098_li=The expression trunc(number) threw a NullPointerException.
changelog_1099_li=Fixed a deadlock when updating LOB's concurrently. See TestLob.testDeadlock2().
changelog_1100_li=Fixed a deadlock related to very large temporary result sets.
changelog_1101_li=Add "-list" command line option to Shell tool so that result-list-mode can be triggered when reading from a file.
changelog_1102_li=Issue 474\:H2 MySQL Compatibility code fails to ignore "COMMENT" in CREATE TABLE, patch from Aaron Azeckoski.
changelog_1103_li=Issue 476\:Broken link in jaqu.html
changelog_1104_li=Fix potential UTF8 encoding issue in org.h2.store.FileStore, reported by Juerg Spiess.
changelog_1105_li=Improve error message when check constraint is broken, test case from Gili (cowwoc).
changelog_1106_li=Improve error message when we have a unique constraint violation, displays the offending key in the error message.
changelog_1107_li=Issue 478\:Support for "SHOW TRANSACTION ISOLATION LEVEL", patch from Andrew Franklin.
changelog_1108_li=Issue 475\:PgServer\:add support for CancelRequest, patch from Andrew Franklin.
changelog_1109_li=Issue 473\:PgServer missing -key option, patch from Andrew Franklin.
changelog_1110_li=Issue 471\:CREATE VIEW does not check user rights, patch from Andrew Franklin.
changelog_1111_li=Issue 477\:PgServer binary transmission of query params is unimplemented, patch from Andrew Franklin.
changelog_1112_li=Issue 479\:Support for SUBSTRING without a FROM condition, patch from Andrew Franklin.
changelog_1113_li=Issue 472\:PgServer does not work with any recent Postgres JDBC driver, patch from Andrew Franklin.
changelog_1114_li=Add syntax for passing additional parameters into custom TableEngine implementations.
changelog_1115_li=Issue 480\:Bugfix post issue 475, 477, patch from Andrew Franklin.
changelog_1116_li=Issue 481\:Further extensions to PgServer to support better support PG JDBC, patch from Andrew Franklin.
changelog_1117_li=Add support for spatial datatype GEOMETRY.
changelog_1118_li=Add support for in-memory spatial index.
changelog_1119_li=change the PageStore\#changeCount field from an int to a long, to cope with databases with very high transaction rates.
changelog_1120_li=Fix a NullPointerException when attempting to add foreign key reference to a view.
changelog_1121_li=Add sufficient ClientInfo support to our javax.sql.Connection implementation to make WebSphere happy.
changelog_1122_li=Issue 482\:class LobStorageBackend$LobInputStream does not override the method InputStream.available().
changelog_1123_li=Fix corruption resulting from a mix of the "WRITE_DELAY\=0" option and "SELECT DISTINCT" queries that don't fit in memory.
changelog_1124_li=Fix the combination of updating a table which contains an LOB, and reading from the LOB at the same time. Previously it would throw an exception, now it works.
changelog_1125_li=Issue 484\:In the H2 Console tool, all schemas starting with "INFO" where hidden. Now they are hidden only if the database is not H2. Patch from "mgcodeact"/"cumer d"
changelog_1126_li=MySQL compatibility, support the "AUTO_INCREMENT\=3" part of the CREATE TABLE statement.
changelog_1127_li=Issue 486\:MySQL compatibility, support the "DEFAULT CHARSET" part of the CREATE TABLE statement.
changelog_1128_li=Issue 487\:support the MySQL "SET foreign_key_checks \=0" command
changelog_1129_li=Issue 490\:support MySQL "USING BTREE" index declaration
changelog_1130_li=Issue 485\:Database get corrupted when column is renamed for which check constraint was defined inside create table statement.
changelog_1131_li=Issue 499\:support MySQL "UNIQUE KEY (ID) USING BTREE" constraint syntax
changelog_1132_li=Issue 501\:"CREATE TABLE .. WITH"not serialized, patch from nico.devel
changelog_1133_li=Avoid problems with runtime-compiled ALIAS methods when people have set the JAVA_TOOL_OPTIONS environment variable.
changelog_1134_h2=Version 1.3.172 (2013-05-25)
changelog_1135_li=Referential integrity\:when adding a referential integrity constraint failed, and if creating the constraint automatically created an index, this index was not removed.
changelog_1136_li=The auto-analyze feature now only reads 1000 rows per table instead of 10000.
changelog_1137_li=The optimization for IN(...) queries combined with OR could result in a strange exception of the type "column x must be included in the group by list".
changelog_1138_li=Issue 454\:Use Charset for type-safety.
changelog_1139_li=Queries with both LIMIT and OFFSET could throw an IllegalArgumentException.
changelog_1140_li=MVStore\:multiple issues were fixed\:460, 461, 462, 464, 466.
changelog_1141_li=MVStore\:larger stores (multiple GB) are now much faster.
changelog_1142_li=When using local temporary tables and not dropping them manually before closing the session, and then killing the process could result in a database that couldn't be opened (except when using the recover tool).
changelog_1143_li=Support TRUNC(timestamp) for improved Oracle compatibility.
changelog_1144_li=Add support for CREATE TABLE TEST (ID BIGSERIAL) for PostgreSQL compatibility. Patch from Jesse Long.
changelog_1145_li=Add new collation command SET BINARY_COLLATION UNSIGNED, helps with people testing BINARY columns in MySQL mode.
changelog_1146_li=Issue 453\:ABBA race conditions in TABLE LINK connection sharing.
changelog_1147_li=Issue 449\:Postgres Serial data type should not automatically be marked as primary key
changelog_1158_li=Security\:the TCP server did not correctly restrict access rights of clients in some cases. This was specially a problem when using the flag "tcpAllowOthers".
changelog_1159_li=H2 Console\:the session timeout can now be configured using the system property "h2.consoleTimeout".
changelog_1160_li=Issue 431\:Improved compatibility with MySQL\:support for "ENGINE\=InnoDB charset\=UTF8" when creating a table.
changelog_1161_li=Issue 249\:Improved compatibility with MySQL in the MySQL mode\:now the methods DatabaseMetaData methods stores*Case*Identifiers return the same as MySQL when using the MySQL mode.
changelog_1162_li=Issue 434\:H2 Console didn't work in the Chrome browser due to a wrong viewport argument.
changelog_1163_li=There was a possibility that the .lock.db file was not deleted when the database was closed, which could slow down opening the database.
changelog_1164_li=The SQL script generated by the "script" command contained inconsistent newlines on Windows.
changelog_1165_li=When using trace level 4 (SLF4J) in the server mode, a directory "trace.db" and an empty file was created on the client side. This is no longer made.
changelog_1166_li=Optimize IN(...) queries\:there was a bug in version 1.3.170 if the type of the left hand side didn't match the type of the right hand side. Fixed.
changelog_1167_li=Optimize IN(...) queries\:there was a bug in version 1.3.170 for comparison of the type "X IN(NULL, NULL)". Fixed.
changelog_1168_li=Timestamps with timezone that were passed as a string were not always converted correctly. For example "2012-11-06T23\:00\:00.000Z" was converted to "2012-11-06" instead of to "2012-11-07" in the timezone CET. Thanks a lot to Steve Hruda for reporting the problem\!
changelog_1169_li=New table engine "org.h2.mvstore.db.MVTableEngine" that internally uses the MVStore to persist data. To try it out, append ";DEFAULT_TABLE_ENGINE\=org.h2.mvstore.db.MVTableEngine" to the database URL. This is still very experimental, and many features are not supported yet. The data is stored in a file with the suffix ".mv.db".
changelog_1170_li=New connection setting "DEFAULT_TABLE_ENGINE" to use a specific table engine if none is set explicitly. This is to simplify testing the MVStore table engine.
changelog_1171_li=MVStore\:encrypted stores are now supported. Only standardized algorithms are used\:PBKDF2, SHA-256, XTS-AES, AES-128.
changelog_1172_li=MVStore\:improved API thanks to Simo Tripodi.
changelog_1173_li=MVStore\:maps can now be renamed.
changelog_1174_li=MVStore\:store the file header also at the end of each chunk, which results in a further reduced number of write operations.
changelog_1175_li=MVStore\:a map implementation that supports concurrent operations.
changelog_1176_li=MVStore\:unified exception handling; the version is included in the messages.
changelog_1177_li=MVStore\:old data is now retained for 45 seconds by default.
changelog_1178_li=MVStore\:compress is now disabled by default, and can be enabled on request.
changelog_1179_li=Support ALTER TABLE ADD ... AFTER. Patch from Andrew Gaul (argaul at gmail.com). Fixes issue 401.
changelog_1180_li=Improved OSGi support. H2 now registers itself as a DataSourceFactory service. Fixes issue 365.
changelog_1181_li=Add a DISK_SPACE_USED system function. Fixes issue 270.
changelog_1182_li=Fix a compile-time ambiguity when compiling with JDK7, thanks to a patch from Lukas Eder.
changelog_1183_li=Supporting dropping an index for Lucene full-text indexes.
changelog_1184_li=Optimized performance for SELECT ... ORDER BY X LIMIT Y OFFSET Z queries for in-memory databases using partial sort (by Sergi Vladykin).
changelog_1185_li=Experimental off-heap memory storage engine "nioMemFS\:" and "nioMemLZF\:", suggestion from Mark Addleman.
changelog_1186_li=Issue 438\: JdbcDatabaseMetaData.getSchemas() is no longer supported as of 1.3.169.
changelog_1187_li=MySQL compatibility\: support for ALTER TABLE tableName MODIFY [COLUMN] columnName columnDef. Patch from Ville Koskela.
changelog_1188_li=Issue 404\: SHOW COLUMNS FROM tableName does not work with ALLOW_LITERALS\=NUMBERS.
changelog_1189_li=Throw an explicit error to make it clear we don't support the TRIGGER combination of SELECT and FOR EACH ROW.
changelog_1190_li=Issue 439\: Utils.sortTopN does not handle single-element arrays.
changelog_1191_h2=Version 1.3.170 (2012-11-30)
changelog_1192_li=Issue 407\: The TriggerAdapter didn't work with CLOB and BLOB columns.
changelog_1193_li=PostgreSQL compatibility\: support for data types BIGSERIAL and SERIAL as an alias for AUTO_INCREMENT.
changelog_1194_li=Issue 417\: H2 Console\: the web session timeout didn't work, resulting in a memory leak. This was only a problem if the H2 Console was run for a long time and many sessions were opened.
changelog_1195_li=Issue 412\: Running the Server tool with just the option "-browser" will now log a warning.
changelog_1196_li=Issue 411\:CloseWatcher registration was not concurrency-safe.
changelog_1197_li=MySQL compatibility\:support for CONCAT_WS. Thanks a lot to litailang for the patch\!
changelog_1198_li=PostgreSQL compatibility\:support for EXTRACT(WEEK FROM dateColumn). Thanks to Prashant Bhat for the patch\!
changelog_1199_li=Fix for a bug where we would sometimes use the wrong unique constraint to validate foreign key constraints.
changelog_1200_li=Support BOM at the beginning of files for the RUNSCRIPT command
changelog_1201_li=Fix in calling SET @X \=IDENTITY() where it would return NULL incorrectly
changelog_1202_li=Fix ABBA deadlock between adding a constraint and the H2-Log-Writer thread.
changelog_1203_li=Optimize IN(...) queries where the values are constant and of the same type.
changelog_1204_li=Restore tool\:the parameter "quiet" was not used and is now removed.
changelog_1205_li=Fix ConcurrentModificationException when creating tables and executing SHOW TABLES in parallel. Reported by Viktor Voytovych.
changelog_1206_li=Serialization is now pluggable using the system property "h2.javaObjectSerializer". Thanks to Sergi Vladykin for the patch\!
changelog_1207_h2=Version 1.3.169 (2012-09-09)
changelog_1208_li=The default jar file is now compiled for Java 6.
changelog_1209_li=The new jar file will probably not end up in the central Maven repository in the next few weeks because Sonatype has disabled automatic synchronization from SourceForge (which they call 'legacy sync' now). It will probably take some time until this is sorted out. The H2 jar files are deployed to http\://h2database.com/m2-repo/com/h2database/h2/maven-metadata.xml and http\://hsql.sourceforge.net/m2-repo/com/h2database/h2/maven-metadata.xml as usual.
changelog_1210_li=A part of the documentation and the H2 Console has been changed to support the Apple retina display.
changelog_1211_li=The CreateCluster tool could not be used if the source database contained a CLOB or BLOB. The root cause was that the TCP server did not synchronize on the session, which caused a problem when using the exclusive mode.
changelog_1212_li=Statement.getQueryTimeout()\:only the first call to this method will query the database. If the query timeout was changed in another way than calling setQueryTimeout, this method will always return the last value. This was changed because Hibernate calls getQueryTimeout() a lot.
changelog_1213_li=Issue 416\:PreparedStatement.setNString throws AbstractMethodError. All implemented JDBC 4 methods that don't break compatibility with Java 5 are now included in the default jar file.
changelog_1214_li=Issue 414\:for some functions, the parameters were evaluated twice (for example "char(nextval(..))" ran "nextval(..)" twice).
changelog_1215_li=The ResultSetMetaData methods getSchemaName and getTableName could return null instead of "" (an empty string) as specified in the JDBC API.
changelog_1216_li=Added compatibility for "SET NAMES" query in MySQL compatibility mode.
changelog_1217_h2=Version 1.3.168 (2012-07-13)
changelog_1218_li=The message "Transaction log could not be truncated" was sometimes written to the .trace.db file even if there was no problem truncating the transaction log.
changelog_1219_li=New system property "h2.serializeJavaObject" (default\:true) that allows to disable serializing Java objects, so that the objects compareTo and toString methods can be used.
changelog_1220_li=Dylan has translated the H2 Console tool to Korean. Thanks a lot\!
changelog_1221_li=Executing the statement CREATE INDEX IF ALREADY EXISTS if the index already exists no longer fails for a read only database.
changelog_1222_li=MVCC\:concurrently updating a row could result in the row to appear deleted in the second connection, if there are multiple unique indexes (or a primary key and at least one unique index). Thanks a lot to Teruo for the patch\!
changelog_1223_li=Fulltext search\:in-memory Lucene indexes are now supported.
changelog_1224_li=Fulltext search\:UUID primary keys are now supported.
changelog_1225_li=Apache Tomcat 7.x will now longer log a warning when unloading the web application, if using a connection pool.
changelog_1226_li=H2 Console\:support the Midori browser (for Debian / Raspberry Pi)
changelog_1227_li=When opening a remote session, don't open a temporary file if the trace level is set to zero
changelog_1228_li=Use HMAC for authenticating remote LOB id's, removing the need for maintaining a cache, and removing the limit on the number of LOBs per result set.
changelog_1229_li=H2 Console\:HTML and XML documents can now be edited in an updatable result set. There is (limited) support for editing multi-line documents.
changelog_1002_li=H2 Console\:auto-complete is now triggered by the tab key.
changelog_1003_li=H2 Console\:auto-complete did not work with multi-line statements.
changelog_1004_li=CLOB and BLOB data was not immediately removed after a rollback.
changelog_1005_li=There is a new Aggregate API that supports the internal H2 data types (GEOMETRY for example). Thanks a lot to Nicolas Fortin for the patch\!
changelog_1006_li=Referential integrity constraints sometimes used the wrong index, such that updating a row in the referenced table incorrectly failed with a constraint violation.
changelog_1007_li=The Polish translation was completed and corrected by Wojtek Jurczyk. Thanks a lot\!
changelog_1008_li=Issue 545\:Unnecessary duplicate code was removed.
changelog_1009_li=The profiler tool can now process files with full thread dumps.
changelog_1010_li=MVStore\:the file format was changed slightly.
changelog_1011_li=MVStore mode\:the CLOB and BLOB storage was re-implemented and is now much faster than with the PageStore (which is still the default storage).
changelog_1012_li=MVStore mode\:creating indexes is now much faster (in many cases faster than with the default PageStore).
changelog_1013_li=Various bugs in the MVStore storage and have been fixed, including a bug in the R-tree implementation. The database could get corrupt if there were transient IO exceptions while storing.
changelog_1014_li=The method org.h2.expression.Function.getCost could throw a NullPointException.
changelog_1015_li=Storing LOBs in separate files (outside of the main database file) is no longer supported for new databases.
changelog_1016_li=Lucene 2 is no longer supported.
changelog_1017_li=Fix bug in calculating default MIN and MAX values for SEQUENCE.
changelog_1018_li=Fix bug in performing IN queries with multiple values when IGNORECASE\=TRUE
changelog_1019_li=Add entry-point to org.h2.tools.Shell so it can be called from inside an application. patch by Thomas Gillet.
changelog_1020_li=Fix bug that prevented the PgServer from being stopped and started multiple times.
changelog_1021_h2=Version 1.3.175 (2014-01-18)
changelog_1022_li=EXPLAIN was incorrect for queries with "ANY" or "SOME" conditions.
changelog_1023_li=CallableStatement with "out" parameters\:running the same statement twice could result in an exception ("parameter not set").
changelog_1024_li=Union queries\:duplicate rows could be returned if the sub-queries contained "order by".
changelog_1025_li=The GEOMETRY data type now works for user defined functions that return a result set.
changelog_1026_li=PostgreSQL compatibility\:the PgServer was not working properly when the setting database_to_upper was set to false.
changelog_1027_li=JdbcDataSource\:the methods setUrl and getUrl where added as aliases for setURL and getURL. This should solve problems with the HikariCP (Hikari connection pool).
changelog_1028_li=Issue 537\:H2 Console\:when loading the schema, incorrect JDBC calls where issued, which caused the exception "Column PRECISION not found".
changelog_1030_li=The collation can now be set in the database URL, even if there are data tables, if the collection is the same as the current collation.
changelog_1031_li=Improved Oracle compatibility for CASE WHEN and DECODE.
changelog_1032_li=The statement "drop all objects" did not work if a table depends on a view via a constraint.
changelog_1033_li=Subqueries or views with "order by" an alias expression could not be executed due to a regression introduced in version 1.3.174.
changelog_1034_li=Issue 73\:MySQL compatibility\:support REPLACE, patch by Cemo Koc.
changelog_1035_li=The spatial index now works in MVCC mode when using the MVStore storage.
changelog_1036_li=MVStore\:concurrency problems have been fixed. The API has been simplified.
changelog_1037_li=Improve error message when dropping an index that belongs to a constraint, specify constraint in error message.
changelog_1038_li=Issue 518\:java.sql.Connection.commit() freezes after LOB modification with EXCLUSIVE connection
changelog_1039_li=Issue 517\:Create or replace view statement has no effect on the others already existing JDBC connection
changelog_1040_li=Support 123L syntax as in Java; example\:SELECT (2000000000L*2).
changelog_1041_li=Issue 520\:Add support for sequence max value, min value and cycle, patch by Daniel Gredler.
changelog_1042_li=Issue 521\:ScriptReader should implement Closeable
changelog_1043_li=Issue 524\:RunScript.execute does not close its Statement, patch from Gaul.
changelog_1044_li=Add support for DB2 "WITH UR" clause, patch from litailang
changelog_1045_li=Added support for ON DUPLICATE KEY UPDATE like MySQL with the values() function to update with the value that was to be inserted. Patch from Jean-Francois Noel.
changelog_1046_li=Issue 522\:Treat empty strings like NULL in Oracle compatibility mode, patch by Daniel Gredler.
changelog_1047_li=Issue 527\:Oracle compatibility mode\:incorrect scale behavior, patch by Daniel Gredler.
changelog_1048_li=Slightly reduce the memory cost of View metadata.
changelog_1049_li=Extend support of "GRANT ALTER ANY SCHEMA TO <user>" to allow grantee ability to manipulate tables
changelog_1050_li=Issue 532\:Javadoc for ErrorCode.ROLES_AND_RIGHT_CANNOT_BE_MIXED looks wrong
changelog_1051_li=Issue 528\:Add Oracle-compatible TO_CHAR function, patch by Daniel Gredler.
changelog_1052_li=Issue 534\:CREATE ALIAS NOCACHE, Patch by Nicolas Fortin
changelog_1053_li=Fix an issue with storing Unicode surrogate pairs in CLOB columns.
changelog_1055_li=Fix bug in DB2 syntax "select * from test with ur", patch from litailang
changelog_1056_li=Fix bug in DROP ALL OBJECTS when dealing with inter-schema dependencies.
changelog_1057_h2=Version 1.3.174 (2013-10-19)
changelog_1058_li=LIRS cache\:bugfix for caches that only contain non-resident entries.
changelog_1059_li=For in-memory databases, queries with a "group by" column that is also a hash index threw a RuntimeException.
changelog_1060_li=Improved error message for some syntax errors.
changelog_1061_li=File system abstraction\:if used directly, some file systems did not work correctly with spliced byte buffers (the database engine doesn't use those).
changelog_1062_li=To use the MVStore storage engine (which is still work in progress), append ";mv_store\=true" to the database URL. Using the MVTableEngine when creating the table is no longer recommended.
changelog_1063_li=To compile user defined functions, the javax.tools.JavaCompiler is now used if available, and no temporary files are created. This should solve problems when multiple H2 database concurrently compile the same user defined functions. To disable, system the system property "h2.javaSystemCompiler" to false.
changelog_1064_li=Concurrently creating function aliases in different databases could result in the exception "javac\:file not found".
changelog_1065_li=The function "regexp_replace" threw the wrong kind of exception if the replacement string was invalid.
changelog_1066_li=A checkpoint is now done every MAX_LOG_SIZE / 2 instead of every MAX_LOG_SIZE, so that the transaction log doesn't grow as large.
changelog_1067_li=MVStore table engine\:new setting "retention_time" to configure the time to retain old data. The default is 45 seconds.
changelog_1068_li=The method TableEngine.createTable() now returns a Table object.
changelog_1069_li=For read-only databases, for the trace level "debug", the trace info is written to the temp directory.
changelog_1070_li=Closing the file lock will now wait until the background thread is stopped.
changelog_1071_li=In version 1.3.172, a performance regression was introduced when fixing the issue 389 (when there is a multi-column primary key, H2 does not seem to always pick the right index). This was related to boosting an index that matches the "order by" column list (the wrong index was used in some cases).
changelog_1072_li=Improved spatial index and data type.
changelog_1073_li=Issue 467\:OSGi Class Loader (ability to create reference to class in other ClassLoader, for example in another OSGi bundle).
changelog_1074_li=Queries "group by" that contain a subquery with an aggregate function returned the wrong result in some cases.
changelog_1075_li=Fix bug in unique and non-unique hash indexes which manifested as incorrect results when the search key was a different cardinal type from the table index key. e.g. where the one was INT and the other was LONG
changelog_1076_li=Bug\:Changes to the database structure did not result in the Session query cache being invalidated.
changelog_1077_li=New feature from Davide Cavestro - allow using custom Java object serialization engines on a per-DB basis.
changelog_1078_li=When running the Recover tool on very large (>6G) databases, some statistics were reported with negative numbers.
changelog_1079_li=Add a CONTAINS_UNCOMMITTED column to the SESSIONS metadata table, to allow detecting when rogue sessions are creating large transactions.
changelog_1080_li=Some small fixes to the GEOMETRY support, patches by Nicolas Fortin.
changelog_1081_li=The BNF tool and the autocomplete feature are exported in OSGi, which makes it possible for smart editors to do autocomplete. Patch from Nicolas Fortin.
changelog_1082_li=Fix DROP ALL OBJECTS and DROP SCHEMA in the presence of tables with computed column dependencies.
changelog_1083_li=Session-temporary LOB's could sometimes accumulate, increasing the size of the DB file until shutdown. Now they are cleared out at every commit.
changelog_1084_li=There was a bug where a hash index with more than one column would be silently converted to a regular index. It will now throw an exception.
changelog_1085_li=Query Statistics\:new feature which stores the newest 100 SQL queries executed and their performance data. Useful for tracking down badly performing queries.
changelog_1086_li=Fix an LOB deadlock between reading and updating LOB columns.
changelog_1087_li=Support the JDBC DatabaseMetaData\#getClientInfoProperties() method, in the sense of returning an empty result, in order to make WebSphere happy.
changelog_1088_li=Make Server\#openBrowser respect the $BROWSER environment variable. Add "chromium" to the list of browsers tried. Patch from Felix Kaiser.
changelog_1089_li=Fix to org.h2.util.ScriptReader when handling unclosed block comments.
changelog_1090_li=Make org.h2.util.ScriptReader throw a better exception when handling broken scripts which generate extremely large statements.
changelog_1091_li=Fix bug with ALLOW_LITERALS\=NONE, where the periodic analyze table on insert would throw an exception. A similar problem was fixed in the Console tool.
changelog_1092_li=Issue 510\:Make org.h2.bnf public for consumption by external projects, patch by Nicolas Fortin
changelog_1093_li=Issue 509\:Important fix on ValueGeometry, patch by Nicolas Fortin (with some tweaking) Make ValueGeometry\#getDimensionCount more reliable. Add unit test to check for illegal ValueGeometry comparison Add unit test for conversion of Geometry object into Object Add optional export to MANIFEST.MF for JTS Geometry classes Validate that geometry values can be represented in WKB.
changelog_1094_li=Issue 506\:RFE\:Include Thread.getName() in case of a deadlock
changelog_1095_li=Adding support for "GRANT ALTER ANY SCHEMA TO <user>", patch by John Yates
changelog_1096_h2=Version 1.3.173 (2013-07-28)
changelog_1097_li=Support empty statements that just contains a comment.
changelog_1098_li=Server mode\:if there was an error while reading from a LOB, the session was closed in some cases.
changelog_1099_li=Issue 463\:Driver name and version are now the same in OsgiDataSourceFactory and JdbcDatabaseMetaData.
changelog_1100_li=JaQu\:The data type VARCHAR is now (again) used for Strings (no longer TEXT, except when explicitly set).
changelog_1101_li=For in-memory databases, creating an index on a CLOB or BLOB column is no longer supported. This is to simplify the MVTableEngine.
changelog_1103_li=Issue 468\:trunc(timestamp) could return the wrong value (+12 hours), and trunc(number) throw a NullPointerException.
changelog_1104_li=The expression trunc(number) threw a NullPointerException.
changelog_1105_li=Fixed a deadlock when updating LOB's concurrently. See TestLob.testDeadlock2().
changelog_1106_li=Fixed a deadlock related to very large temporary result sets.
changelog_1107_li=Add "-list" command line option to Shell tool so that result-list-mode can be triggered when reading from a file.
changelog_1108_li=Issue 474\:H2 MySQL Compatibility code fails to ignore "COMMENT" in CREATE TABLE, patch from Aaron Azeckoski.
changelog_1109_li=Issue 476\:Broken link in jaqu.html
changelog_1110_li=Fix potential UTF8 encoding issue in org.h2.store.FileStore, reported by Juerg Spiess.
changelog_1111_li=Improve error message when check constraint is broken, test case from Gili (cowwoc).
changelog_1112_li=Improve error message when we have a unique constraint violation, displays the offending key in the error message.
changelog_1113_li=Issue 478\:Support for "SHOW TRANSACTION ISOLATION LEVEL", patch from Andrew Franklin.
changelog_1114_li=Issue 475\:PgServer\:add support for CancelRequest, patch from Andrew Franklin.
changelog_1115_li=Issue 473\:PgServer missing -key option, patch from Andrew Franklin.
changelog_1116_li=Issue 471\:CREATE VIEW does not check user rights, patch from Andrew Franklin.
changelog_1117_li=Issue 477\:PgServer binary transmission of query params is unimplemented, patch from Andrew Franklin.
changelog_1118_li=Issue 479\:Support for SUBSTRING without a FROM condition, patch from Andrew Franklin.
changelog_1119_li=Issue 472\:PgServer does not work with any recent Postgres JDBC driver, patch from Andrew Franklin.
changelog_1120_li=Add syntax for passing additional parameters into custom TableEngine implementations.
changelog_1121_li=Issue 480\:Bugfix post issue 475, 477, patch from Andrew Franklin.
changelog_1122_li=Issue 481\:Further extensions to PgServer to support better support PG JDBC, patch from Andrew Franklin.
changelog_1123_li=Add support for spatial datatype GEOMETRY.
changelog_1124_li=Add support for in-memory spatial index.
changelog_1125_li=change the PageStore\#changeCount field from an int to a long, to cope with databases with very high transaction rates.
changelog_1126_li=Fix a NullPointerException when attempting to add foreign key reference to a view.
changelog_1127_li=Add sufficient ClientInfo support to our javax.sql.Connection implementation to make WebSphere happy.
changelog_1128_li=Issue 482\:class LobStorageBackend$LobInputStream does not override the method InputStream.available().
changelog_1129_li=Fix corruption resulting from a mix of the "WRITE_DELAY\=0" option and "SELECT DISTINCT" queries that don't fit in memory.
changelog_1130_li=Fix the combination of updating a table which contains an LOB, and reading from the LOB at the same time. Previously it would throw an exception, now it works.
changelog_1131_li=Issue 484\:In the H2 Console tool, all schemas starting with "INFO" where hidden. Now they are hidden only if the database is not H2. Patch from "mgcodeact"/"cumer d"
changelog_1132_li=MySQL compatibility, support the "AUTO_INCREMENT\=3" part of the CREATE TABLE statement.
changelog_1133_li=Issue 486\:MySQL compatibility, support the "DEFAULT CHARSET" part of the CREATE TABLE statement.
changelog_1134_li=Issue 487\:support the MySQL "SET foreign_key_checks \=0" command
changelog_1135_li=Issue 490\:support MySQL "USING BTREE" index declaration
changelog_1136_li=Issue 485\:Database get corrupted when column is renamed for which check constraint was defined inside create table statement.
changelog_1137_li=Issue 499\:support MySQL "UNIQUE KEY (ID) USING BTREE" constraint syntax
changelog_1138_li=Issue 501\:"CREATE TABLE .. WITH"not serialized, patch from nico.devel
changelog_1139_li=Avoid problems with runtime-compiled ALIAS methods when people have set the JAVA_TOOL_OPTIONS environment variable.
changelog_1140_h2=Version 1.3.172 (2013-05-25)
changelog_1141_li=Referential integrity\:when adding a referential integrity constraint failed, and if creating the constraint automatically created an index, this index was not removed.
changelog_1142_li=The auto-analyze feature now only reads 1000 rows per table instead of 10000.
changelog_1143_li=The optimization for IN(...) queries combined with OR could result in a strange exception of the type "column x must be included in the group by list".
changelog_1144_li=Issue 454\:Use Charset for type-safety.
changelog_1145_li=Queries with both LIMIT and OFFSET could throw an IllegalArgumentException.
changelog_1146_li=MVStore\:multiple issues were fixed\:460, 461, 462, 464, 466.
changelog_1147_li=MVStore\:larger stores (multiple GB) are now much faster.
changelog_1148_li=When using local temporary tables and not dropping them manually before closing the session, and then killing the process could result in a database that couldn't be opened (except when using the recover tool).
changelog_1149_li=Support TRUNC(timestamp) for improved Oracle compatibility.
changelog_1150_li=Add support for CREATE TABLE TEST (ID BIGSERIAL) for PostgreSQL compatibility. Patch from Jesse Long.
changelog_1151_li=Add new collation command SET BINARY_COLLATION UNSIGNED, helps with people testing BINARY columns in MySQL mode.
changelog_1152_li=Issue 453\:ABBA race conditions in TABLE LINK connection sharing.
changelog_1153_li=Issue 449\:Postgres Serial data type should not automatically be marked as primary key
changelog_1164_li=Security\:the TCP server did not correctly restrict access rights of clients in some cases. This was specially a problem when using the flag "tcpAllowOthers".
changelog_1165_li=H2 Console\:the session timeout can now be configured using the system property "h2.consoleTimeout".
changelog_1166_li=Issue 431\:Improved compatibility with MySQL\:support for "ENGINE\=InnoDB charset\=UTF8" when creating a table.
changelog_1167_li=Issue 249\:Improved compatibility with MySQL in the MySQL mode\:now the methods DatabaseMetaData methods stores*Case*Identifiers return the same as MySQL when using the MySQL mode.
changelog_1168_li=Issue 434\:H2 Console didn't work in the Chrome browser due to a wrong viewport argument.
changelog_1169_li=There was a possibility that the .lock.db file was not deleted when the database was closed, which could slow down opening the database.
changelog_1170_li=The SQL script generated by the "script" command contained inconsistent newlines on Windows.
changelog_1171_li=When using trace level 4 (SLF4J) in the server mode, a directory "trace.db" and an empty file was created on the client side. This is no longer made.
changelog_1172_li=Optimize IN(...) queries\:there was a bug in version 1.3.170 if the type of the left hand side didn't match the type of the right hand side. Fixed.
changelog_1173_li=Optimize IN(...) queries\:there was a bug in version 1.3.170 for comparison of the type "X IN(NULL, NULL)". Fixed.
changelog_1174_li=Timestamps with timezone that were passed as a string were not always converted correctly. For example "2012-11-06T23\:00\:00.000Z" was converted to "2012-11-06" instead of to "2012-11-07" in the timezone CET. Thanks a lot to Steve Hruda for reporting the problem\!
changelog_1175_li=New table engine "org.h2.mvstore.db.MVTableEngine" that internally uses the MVStore to persist data. To try it out, append ";DEFAULT_TABLE_ENGINE\=org.h2.mvstore.db.MVTableEngine" to the database URL. This is still very experimental, and many features are not supported yet. The data is stored in a file with the suffix ".mv.db".
changelog_1176_li=New connection setting "DEFAULT_TABLE_ENGINE" to use a specific table engine if none is set explicitly. This is to simplify testing the MVStore table engine.
changelog_1177_li=MVStore\:encrypted stores are now supported. Only standardized algorithms are used\:PBKDF2, SHA-256, XTS-AES, AES-128.
changelog_1178_li=MVStore\:improved API thanks to Simo Tripodi.
changelog_1179_li=MVStore\:maps can now be renamed.
changelog_1180_li=MVStore\:store the file header also at the end of each chunk, which results in a further reduced number of write operations.
changelog_1181_li=MVStore\:a map implementation that supports concurrent operations.
changelog_1182_li=MVStore\:unified exception handling; the version is included in the messages.
changelog_1183_li=MVStore\:old data is now retained for 45 seconds by default.
changelog_1184_li=MVStore\:compress is now disabled by default, and can be enabled on request.
changelog_1185_li=Support ALTER TABLE ADD ... AFTER. Patch from Andrew Gaul (argaul at gmail.com). Fixes issue 401.
changelog_1186_li=Improved OSGi support. H2 now registers itself as a DataSourceFactory service. Fixes issue 365.
changelog_1187_li=Add a DISK_SPACE_USED system function. Fixes issue 270.
changelog_1188_li=Fix a compile-time ambiguity when compiling with JDK7, thanks to a patch from Lukas Eder.
changelog_1189_li=Supporting dropping an index for Lucene full-text indexes.
changelog_1190_li=Optimized performance for SELECT ... ORDER BY X LIMIT Y OFFSET Z queries for in-memory databases using partial sort (by Sergi Vladykin).
changelog_1191_li=Experimental off-heap memory storage engine "nioMemFS\:" and "nioMemLZF\:", suggestion from Mark Addleman.
changelog_1192_li=Issue 438\: JdbcDatabaseMetaData.getSchemas() is no longer supported as of 1.3.169.
changelog_1193_li=MySQL compatibility\: support for ALTER TABLE tableName MODIFY [COLUMN] columnName columnDef. Patch from Ville Koskela.
changelog_1194_li=Issue 404\: SHOW COLUMNS FROM tableName does not work with ALLOW_LITERALS\=NUMBERS.
changelog_1195_li=Throw an explicit error to make it clear we don't support the TRIGGER combination of SELECT and FOR EACH ROW.
changelog_1196_li=Issue 439\: Utils.sortTopN does not handle single-element arrays.
changelog_1197_h2=Version 1.3.170 (2012-11-30)
changelog_1198_li=Issue 407\: The TriggerAdapter didn't work with CLOB and BLOB columns.
changelog_1199_li=PostgreSQL compatibility\: support for data types BIGSERIAL and SERIAL as an alias for AUTO_INCREMENT.
changelog_1200_li=Issue 417\: H2 Console\: the web session timeout didn't work, resulting in a memory leak. This was only a problem if the H2 Console was run for a long time and many sessions were opened.
changelog_1201_li=Issue 412\: Running the Server tool with just the option "-browser" will now log a warning.
changelog_1202_li=Issue 411\:CloseWatcher registration was not concurrency-safe.
changelog_1203_li=MySQL compatibility\:support for CONCAT_WS. Thanks a lot to litailang for the patch\!
changelog_1204_li=PostgreSQL compatibility\:support for EXTRACT(WEEK FROM dateColumn). Thanks to Prashant Bhat for the patch\!
changelog_1205_li=Fix for a bug where we would sometimes use the wrong unique constraint to validate foreign key constraints.
changelog_1206_li=Support BOM at the beginning of files for the RUNSCRIPT command
changelog_1207_li=Fix in calling SET @X \=IDENTITY() where it would return NULL incorrectly
changelog_1208_li=Fix ABBA deadlock between adding a constraint and the H2-Log-Writer thread.
changelog_1209_li=Optimize IN(...) queries where the values are constant and of the same type.
changelog_1210_li=Restore tool\:the parameter "quiet" was not used and is now removed.
changelog_1211_li=Fix ConcurrentModificationException when creating tables and executing SHOW TABLES in parallel. Reported by Viktor Voytovych.
changelog_1212_li=Serialization is now pluggable using the system property "h2.javaObjectSerializer". Thanks to Sergi Vladykin for the patch\!
changelog_1213_h2=Version 1.3.169 (2012-09-09)
changelog_1214_li=The default jar file is now compiled for Java 6.
changelog_1215_li=The new jar file will probably not end up in the central Maven repository in the next few weeks because Sonatype has disabled automatic synchronization from SourceForge (which they call 'legacy sync' now). It will probably take some time until this is sorted out. The H2 jar files are deployed to http\://h2database.com/m2-repo/com/h2database/h2/maven-metadata.xml and http\://hsql.sourceforge.net/m2-repo/com/h2database/h2/maven-metadata.xml as usual.
changelog_1216_li=A part of the documentation and the H2 Console has been changed to support the Apple retina display.
changelog_1217_li=The CreateCluster tool could not be used if the source database contained a CLOB or BLOB. The root cause was that the TCP server did not synchronize on the session, which caused a problem when using the exclusive mode.
changelog_1218_li=Statement.getQueryTimeout()\:only the first call to this method will query the database. If the query timeout was changed in another way than calling setQueryTimeout, this method will always return the last value. This was changed because Hibernate calls getQueryTimeout() a lot.
changelog_1219_li=Issue 416\:PreparedStatement.setNString throws AbstractMethodError. All implemented JDBC 4 methods that don't break compatibility with Java 5 are now included in the default jar file.
changelog_1220_li=Issue 414\:for some functions, the parameters were evaluated twice (for example "char(nextval(..))" ran "nextval(..)" twice).
changelog_1221_li=The ResultSetMetaData methods getSchemaName and getTableName could return null instead of "" (an empty string) as specified in the JDBC API.
changelog_1222_li=Added compatibility for "SET NAMES" query in MySQL compatibility mode.
changelog_1223_h2=Version 1.3.168 (2012-07-13)
changelog_1224_li=The message "Transaction log could not be truncated" was sometimes written to the .trace.db file even if there was no problem truncating the transaction log.
changelog_1225_li=New system property "h2.serializeJavaObject" (default\:true) that allows to disable serializing Java objects, so that the objects compareTo and toString methods can be used.
changelog_1226_li=Dylan has translated the H2 Console tool to Korean. Thanks a lot\!
changelog_1227_li=Executing the statement CREATE INDEX IF ALREADY EXISTS if the index already exists no longer fails for a read only database.
changelog_1228_li=MVCC\:concurrently updating a row could result in the row to appear deleted in the second connection, if there are multiple unique indexes (or a primary key and at least one unique index). Thanks a lot to Teruo for the patch\!
changelog_1229_li=Fulltext search\:in-memory Lucene indexes are now supported.
changelog_1230_li=Fulltext search\:UUID primary keys are now supported.
changelog_1231_li=Apache Tomcat 7.x will now longer log a warning when unloading the web application, if using a connection pool.
changelog_1232_li=H2 Console\:support the Midori browser (for Debian / Raspberry Pi)
changelog_1233_li=When opening a remote session, don't open a temporary file if the trace level is set to zero
changelog_1234_li=Use HMAC for authenticating remote LOB id's, removing the need for maintaining a cache, and removing the limit on the number of LOBs per result set.
changelog_1235_li=H2 Console\:HTML and XML documents can now be edited in an updatable result set. There is (limited) support for editing multi-line documents.
cheatSheet_1000_h1=H2 Database Engine Cheat Sheet
cheatSheet_1001_h2=Using H2
cheatSheet_1002_a=H2
...
...
@@ -2508,70 +2514,80 @@ mvstore_1114_li=\ if a map is modified concurrently.
mvstore_1115_h3=Storage Engine for H2
mvstore_1116_p=\ The plan is to use the MVStore as the default storage engine for the H2 database in the future (supporting SQL, JDBC, transactions, MVCC, and so on). This is work in progress. To try it out, append <code>;MV_STORE\=TRUE</code> to the database URL. In general, performance should be similar than the current default storage engine (the page store). Even though it can be used with the default table level locking, it is recommended to use it together with the MVCC mode (to do that, append <code>;MVCC\=TRUE</code> to the database URL).
mvstore_1117_h2=File Format
mvstore_1118_p=\ The data is stored in one file. The file contains two file headers (to be safe), and a number of chunks. The file headers are one block each; a block is 4096 bytes. Each chunk is at least one block, but typically 200 blocks or more. There might be a number of free blocks in front of every chunk. There is one chunk for every version.
mvstore_1119_h3=File Header
mvstore_1120_p=\ There are two file headers, which normally contain the exact same data. But once in a while, the file headers are updated, and writing could partially fail, which could corrupt a header. That's why there is a second header. Only the file headers are updated in this way (called "in-place update"). The headers contain the following data\:
mvstore_1121_p=\ The data is stored in the form of a key-value pair. Each value is stored as a hexadecimal number. The entries are\:
mvstore_1122_li=H\:The entry "H\:2" stands for the the H2 database.
mvstore_1123_li=block\:The block number where one of the newest chunks starts (but not necessarily the newest).
mvstore_1124_li=blockSize\:The block size of the file; currently always hex 1000, which is decimal 4096, to match the <a href\="https\://en.wikipedia.org/wiki/Disk_sector">disk sector</a> length of modern hard disks.
mvstore_1125_li=chunk\:The chunk id, which is normally the same value as the version; however, the chunk id might roll over to 0, while the version doesn't.
mvstore_1126_li=created\:The number of milliseconds since 1970 when the file was created.
mvstore_1127_li=format\:The file format number. Currently 1.
mvstore_1128_li=version\:The version number of the chunk.
mvstore_1129_li=fletcher\:The <a href\="https\://en.wikipedia.org/wiki/Fletcher's_checksum"> Fletcher-32 checksum</a> of the header.
mvstore_1130_p=\ When opening the file, both headers are read and the checksum is verified. If both headers are valid, the one with the newer version is used. The chunk with the latest version is then detected (details about this see below), and the rest of the metadata is read from there. If the chunk id, block and version are not stored in the file header, then the latest chunk lookup starts with the last chunk in the file.
mvstore_1131_h3=Chunk Format
mvstore_1132_p=\ There is one chunk per version. Each chunk consists of a header, the pages that were modified in this version, and a footer. The pages contain the actual data of the maps. The pages inside a chunk are stored right after the header, next to each other (unaligned). The size of a chunk is a multiple of the block size. The footer is stored in the last 128 bytes of the chunk.
mvstore_1133_p=\ The footer allows to verify that the chunk is completely written (a chunk is written as one write operation), and allows to find the start position of the very last chunk in the file. The chunk header and footer contain the following data\:
mvstore_1134_p=\ The fields of the chunk header and footer are\:
mvstore_1135_li=chunk\:The chunk id.
mvstore_1136_li=block\:The first block of the chunk (multiply by the block size to get the position in the file).
mvstore_1137_li=len\:The size of the chunk in number of blocks.
mvstore_1138_li=map\:The id of the newest map; incremented when a new map is created.
mvstore_1139_li=max\:The sum of all maximum page sizes (see page format).
mvstore_1140_li=next\:The predicted start block of the next chunk.
mvstore_1141_li=pages\:The number of pages in the chunk.
mvstore_1142_li=root\:The position of the metadata root page (see page format).
mvstore_1143_li=time\:The time the chunk was written, in milliseconds after the file was created.
mvstore_1144_li=version\:The version this chunk represents.
mvstore_1145_li=fletcher\:The checksum of the footer.
mvstore_1146_p=\ Chunks are never updated in-place. Each chunk contains the pages that were changed in that version (there is one chunk per version, see above), plus all the parent nodes of those pages, recursively, up to the root page. If an entry in a map is changed, removed, or added, then the respective page is copied to be stored in the next chunk, and the number of live pages in the old chunk is decremented. This mechanism is called copy-on-write, and is similar to how the <a href\="https\://en.wikipedia.org/wiki/Btrfs">Btrfs</a> file system works. Chunks without live pages are marked as free, so the space can be re-used by more recent chunks. Because not all chunks are of the same size, there can be some unused space in front of a chunk for some time (until a small chunk is written or the chunks are compacted). There is a <a href\="http\://stackoverflow.com/questions/13650134/after-how-many-seconds-are-file-system-write-buffers-typically-flushed"> delay of 45 seconds</a> (by default) before a free chunk is overwritten, to ensure new versions are persisted first, as hard disks sometimes re-order write operations.
mvstore_1147_p=\ How the newest chunk is located when opening a store\:The file header contains the position of a recent chunk, but not always the newest one. This is to reduce the number of file header updates. After opening the file, the file headers, and the chunk footer of the very last chunk (at the end of the file) are read. From those candidates, the header of the most recent chunk is read. If it contains a "next" pointer (see above), those chunk's header and footer are read as well. If it turned out to be a newer valid chunk, this is repeated, until the newest chunk was found. Before writing a chunk, the position of the next chunk is predicted based on the assumption that the next chunk will be of the same size as the current one. When the next chunk is written, and the previous prediction turned out to be incorrect, the file header is updated as well. In any case, the file header is updated if the next chain gets longer than 20 hops.
mvstore_1148_h3=Page Format
mvstore_1149_p=\ Each map is a <a href\="https\://en.wikipedia.org/wiki/B-tree">B-tree</a>, and the map data is stored in (B-tree-) pages. There are leaf pages that contain the key-value pairs of the map, and internal nodes, which only contain keys and pointers to leaf pages. The root of a tree is either a leaf or an internal node. Unlike file header and chunk header and footer, the page data is not human readable. Instead, it is stored as byte arrays, with long (8 bytes), int (4 bytes), short (2 bytes), and <a href\="https\://en.wikipedia.org/wiki/Variable-length_quantity">variable size int and long</a> (1 to 5 / 10 bytes). The page format is\:
mvstore_1150_li=length (int)\:Length of the page in bytes.
mvstore_1151_li=checksum (short)\:Checksum (chunk id xor offset within the chunk xor page length).
mvstore_1152_li=mapId (variable size int)\:The id of the map this page belongs to.
mvstore_1153_li=len (variable size int)\:The number of keys in the page.
mvstore_1154_li=type (byte)\:The page type (0 for leaf page, 1 for internal node; plus 2 if the page data is compressed).
mvstore_1155_li=keys (byte array)\:All keys, stored depending on the data type.
mvstore_1156_li=children (array of long; internal nodes only)\:The position of the children.
mvstore_1157_li=childCounts (array of variable size long; internal nodes only)\:The total number of entries for the given child page.
mvstore_1158_li=values (byte array; leaf pages only)\:All values, stored depending on the data type.
mvstore_1159_p=\ Even though this is not required by the file format, each B-tree is stored "upside down", that means the leaf pages first, then the internal nodes, and lastly the root page.
mvstore_1160_p=\ Pointers to pages are stored as a long, using a special format\:26 bits for the chunk id, 32 bits for the offset within the chunk, 5 bits for the length code, 1 bit for the page type (leaf or internal node). The page type is encoded so that when clearing or removing a map, leaf pages don't have to be read (internal nodes do have to be read in order to know where all the pages are; but in a typical B-tree the vast majority of the pages are leaf pages). The absolute file position is not included so that chunks can be moved within the file without having to change page pointers; only the chunk metadata needs to be changed. The length code is a number from 0 to 31, where 0 means the maximum length of the page is 32 bytes, 1 means 48 bytes, 2\:64, 3\:96, 4\:128, 5\:192, and so on until 31 which means longer than 1 MB. That way, reading a page only requires one read operation (except for very large pages, in which case two read operations might be needed). The sum of the maximum length of all pages is stored in the chunk metadata (field "max"), and when a page is marked as removed, the live maximum length is adjusted. This allows to estimate the amount of free space within a block, in addition to the number of free pages.
mvstore_1161_p=\ The total number of entries in a child nodes is kept to allow efficient range counting, lookup by index, and skip operations. The pages form a <a href\="http\://www.chiark.greenend.org.uk/~sgtatham/algorithms/cbtree.html">counted B-tree</a>.
mvstore_1162_p=\ Data compression\:The data after the page type are optionally compressed using the LZF algorithm.
mvstore_1163_h3=Metadata Map
mvstore_1164_p=\ In addition to the user maps, there is one metadata map that contains names and positions of user maps, and chunk metadata. The very last page of a chunk contains the root page of that metadata map. The exact position of this root page is stored in the chunk header. This page (directly or indirectly) points to the root pages of all other maps. The metadata map of a store with a map named "data", and one chunk, contains the following entries\:
mvstore_1165_li=chunk.1\:The metadata of chunk 1. This is the same data as the chunk header, plus the number of live pages, and the maximum live length.
mvstore_1166_li=setting.storeVersion\:The store version (a user defined value).
mvstore_1167_li=map.1\:The metadata of map 1. The entries are\:name, createVersion, and type.
mvstore_1168_li=name.data\:The map id of the map named "data". The value is "1".
mvstore_1169_li=root.1\:The root position of map 1.
mvstore_1170_h2=Similar Projects and Differences to Other Storage Engines
mvstore_1171_p=\ Unlike similar storage engines like LevelDB and Kyoto Cabinet, the MVStore is written in Java and can easily be embedded in a Java and Android application.
mvstore_1172_p=\ The MVStore is somewhat similar to the Berkeley DB Java Edition because it is also written in Java, and is also a log structured storage, but the H2 license is more liberal.
mvstore_1173_p=\ Like SQLite 3, the MVStore keeps all data in one file. Unlike SQLite 3, the MVStore uses is a log structured storage. The plan is to make the MVStore both easier to use as well as faster than SQLite 3. In a recent (very simple) test, the MVStore was about twice as fast as SQLite 3 on Android.
mvstore_1174_p=\ The API of the MVStore is similar to MapDB (previously known as JDBM) from Jan Kotek, and some code is shared between MVStore and MapDB. However, unlike MapDB, the MVStore uses is a log structured storage. The MVStore does not have a record size limit.
mvstore_1175_h2=Current State
mvstore_1176_p=\ The code is still experimental at this stage. The API as well as the behavior may partially change. Features may be added and removed (even though the main features will stay).
mvstore_1177_h2=Requirements
mvstore_1178_p=\ The MVStore is included in the latest H2 jar file.
mvstore_1179_p=\ There are no special requirements to use it. The MVStore should run on any JVM as well as on Android.
mvstore_1180_p=\ To build just the MVStore (without the database engine), run\:
mvstore_1181_p=\ This will create the file <code>bin/h2mvstore-1.3.175.jar</code> (about 130 KB).
mvstore_1118_p=\ The data is stored in one file. The file contains two file headers (for safety), and a number of chunks. The file headers are one block each; a block is 4096 bytes. Each chunk is at least one block, but typically 200 blocks or more. Data is stored in the chunks in the form of a <a href\="https\://en.wikipedia.org/wiki/Log-structured_file_system">log structured storage</a>. There is one chunk for every version.
mvstore_1119_p=\ Each chunk contains a number of B-tree pages. As an example, the following code\:
mvstore_1120_p=\ will result in the following two chunks (excluding metadata)\:
mvstore_1127_p=\ - Page 5\:(root) node with 2 entries pointing to page 4 and 2
mvstore_1128_p=\ That means each chunk contains the changes of one version\:the new version of the changed pages and the parent pages, recursively, up to the root page. Pages in subsequent chunks refer to pages in earlier chunks.
mvstore_1129_h3=File Header
mvstore_1130_p=\ There are two file headers, which normally contain the exact same data. But once in a while, the file headers are updated, and writing could partially fail, which could corrupt a header. That's why there is a second header. Only the file headers are updated in this way (called "in-place update"). The headers contain the following data\:
mvstore_1131_p=\ The data is stored in the form of a key-value pair. Each value is stored as a hexadecimal number. The entries are\:
mvstore_1132_li=H\:The entry "H\:2" stands for the the H2 database.
mvstore_1133_li=block\:The block number where one of the newest chunks starts (but not necessarily the newest).
mvstore_1134_li=blockSize\:The block size of the file; currently always hex 1000, which is decimal 4096, to match the <a href\="https\://en.wikipedia.org/wiki/Disk_sector">disk sector</a> length of modern hard disks.
mvstore_1135_li=chunk\:The chunk id, which is normally the same value as the version; however, the chunk id might roll over to 0, while the version doesn't.
mvstore_1136_li=created\:The number of milliseconds since 1970 when the file was created.
mvstore_1137_li=format\:The file format number. Currently 1.
mvstore_1138_li=version\:The version number of the chunk.
mvstore_1139_li=fletcher\:The <a href\="https\://en.wikipedia.org/wiki/Fletcher's_checksum"> Fletcher-32 checksum</a> of the header.
mvstore_1140_p=\ When opening the file, both headers are read and the checksum is verified. If both headers are valid, the one with the newer version is used. The chunk with the latest version is then detected (details about this see below), and the rest of the metadata is read from there. If the chunk id, block and version are not stored in the file header, then the latest chunk lookup starts with the last chunk in the file.
mvstore_1141_h3=Chunk Format
mvstore_1142_p=\ There is one chunk per version. Each chunk consists of a header, the pages that were modified in this version, and a footer. The pages contain the actual data of the maps. The pages inside a chunk are stored right after the header, next to each other (unaligned). The size of a chunk is a multiple of the block size. The footer is stored in the last 128 bytes of the chunk.
mvstore_1143_p=\ The footer allows to verify that the chunk is completely written (a chunk is written as one write operation), and allows to find the start position of the very last chunk in the file. The chunk header and footer contain the following data\:
mvstore_1144_p=\ The fields of the chunk header and footer are\:
mvstore_1145_li=chunk\:The chunk id.
mvstore_1146_li=block\:The first block of the chunk (multiply by the block size to get the position in the file).
mvstore_1147_li=len\:The size of the chunk in number of blocks.
mvstore_1148_li=map\:The id of the newest map; incremented when a new map is created.
mvstore_1149_li=max\:The sum of all maximum page sizes (see page format).
mvstore_1150_li=next\:The predicted start block of the next chunk.
mvstore_1151_li=pages\:The number of pages in the chunk.
mvstore_1152_li=root\:The position of the metadata root page (see page format).
mvstore_1153_li=time\:The time the chunk was written, in milliseconds after the file was created.
mvstore_1154_li=version\:The version this chunk represents.
mvstore_1155_li=fletcher\:The checksum of the footer.
mvstore_1156_p=\ Chunks are never updated in-place. Each chunk contains the pages that were changed in that version (there is one chunk per version, see above), plus all the parent nodes of those pages, recursively, up to the root page. If an entry in a map is changed, removed, or added, then the respective page is copied to be stored in the next chunk, and the number of live pages in the old chunk is decremented. This mechanism is called copy-on-write, and is similar to how the <a href\="https\://en.wikipedia.org/wiki/Btrfs">Btrfs</a> file system works. Chunks without live pages are marked as free, so the space can be re-used by more recent chunks. Because not all chunks are of the same size, there can be a number of free blocks in front of a chunk for some time (until a small chunk is written or the chunks are compacted). There is a <a href\="http\://stackoverflow.com/questions/13650134/after-how-many-seconds-are-file-system-write-buffers-typically-flushed"> delay of 45 seconds</a> (by default) before a free chunk is overwritten, to ensure new versions are persisted first.
mvstore_1157_p=\ How the newest chunk is located when opening a store\:The file header contains the position of a recent chunk, but not always the newest one. This is to reduce the number of file header updates. After opening the file, the file headers, and the chunk footer of the very last chunk (at the end of the file) are read. From those candidates, the header of the most recent chunk is read. If it contains a "next" pointer (see above), those chunk's header and footer are read as well. If it turned out to be a newer valid chunk, this is repeated, until the newest chunk was found. Before writing a chunk, the position of the next chunk is predicted based on the assumption that the next chunk will be of the same size as the current one. When the next chunk is written, and the previous prediction turned out to be incorrect, the file header is updated as well. In any case, the file header is updated if the next chain gets longer than 20 hops.
mvstore_1158_h3=Page Format
mvstore_1159_p=\ Each map is a <a href\="https\://en.wikipedia.org/wiki/B-tree">B-tree</a>, and the map data is stored in (B-tree-) pages. There are leaf pages that contain the key-value pairs of the map, and internal nodes, which only contain keys and pointers to leaf pages. The root of a tree is either a leaf or an internal node. Unlike file header and chunk header and footer, the page data is not human readable. Instead, it is stored as byte arrays, with long (8 bytes), int (4 bytes), short (2 bytes), and <a href\="https\://en.wikipedia.org/wiki/Variable-length_quantity">variable size int and long</a> (1 to 5 / 10 bytes). The page format is\:
mvstore_1160_li=length (int)\:Length of the page in bytes.
mvstore_1161_li=checksum (short)\:Checksum (chunk id xor offset within the chunk xor page length).
mvstore_1162_li=mapId (variable size int)\:The id of the map this page belongs to.
mvstore_1163_li=len (variable size int)\:The number of keys in the page.
mvstore_1164_li=type (byte)\:The page type (0 for leaf page, 1 for internal node; plus 2 if the page data is compressed).
mvstore_1165_li=children (array of long; internal nodes only)\:The position of the children.
mvstore_1166_li=childCounts (array of variable size long; internal nodes only)\:The total number of entries for the given child page.
mvstore_1167_li=keys (byte array)\:All keys, stored depending on the data type.
mvstore_1168_li=values (byte array; leaf pages only)\:All values, stored depending on the data type.
mvstore_1169_p=\ Even though this is not required by the file format, each B-tree is stored "upside down", that means the leaf pages first, then the internal nodes, and lastly the root page.
mvstore_1170_p=\ Pointers to pages are stored as a long, using a special format\:26 bits for the chunk id, 32 bits for the offset within the chunk, 5 bits for the length code, 1 bit for the page type (leaf or internal node). The page type is encoded so that when clearing or removing a map, leaf pages don't have to be read (internal nodes do have to be read in order to know where all the pages are; but in a typical B-tree the vast majority of the pages are leaf pages). The absolute file position is not included so that chunks can be moved within the file without having to change page pointers; only the chunk metadata needs to be changed. The length code is a number from 0 to 31, where 0 means the maximum length of the page is 32 bytes, 1 means 48 bytes, 2\:64, 3\:96, 4\:128, 5\:192, and so on until 31 which means longer than 1 MB. That way, reading a page only requires one read operation (except for very large pages). The sum of the maximum length of all pages is stored in the chunk metadata (field "max"), and when a page is marked as removed, the live maximum length is adjusted. This allows to estimate the amount of free space within a block, in addition to the number of free pages.
mvstore_1171_p=\ The total number of entries in a child nodes is kept to allow efficient range counting, lookup by index, and skip operations. The pages form a <a href\="http\://www.chiark.greenend.org.uk/~sgtatham/algorithms/cbtree.html">counted B-tree</a>.
mvstore_1172_p=\ Data compression\:The data after the page type are optionally compressed using the LZF algorithm.
mvstore_1173_h3=Metadata Map
mvstore_1174_p=\ In addition to the user maps, there is one metadata map that contains names and positions of user maps, and chunk metadata. The very last page of a chunk contains the root page of that metadata map. The exact position of this root page is stored in the chunk header. This page (directly or indirectly) points to the root pages of all other maps. The metadata map of a store with a map named "data", and one chunk, contains the following entries\:
mvstore_1175_li=chunk.1\:The metadata of chunk 1. This is the same data as the chunk header, plus the number of live pages, and the maximum live length.
mvstore_1176_li=map.1\:The metadata of map 1. The entries are\:name, createVersion, and type.
mvstore_1177_li=name.data\:The map id of the map named "data". The value is "1".
mvstore_1178_li=root.1\:The root position of map 1.
mvstore_1179_li=setting.storeVersion\:The store version (a user defined value).
mvstore_1180_h2=Similar Projects and Differences to Other Storage Engines
mvstore_1181_p=\ Unlike similar storage engines like LevelDB and Kyoto Cabinet, the MVStore is written in Java and can easily be embedded in a Java and Android application.
mvstore_1182_p=\ The MVStore is somewhat similar to the Berkeley DB Java Edition because it is also written in Java, and is also a log structured storage, but the H2 license is more liberal.
mvstore_1183_p=\ Like SQLite 3, the MVStore keeps all data in one file. Unlike SQLite 3, the MVStore uses is a log structured storage. The plan is to make the MVStore both easier to use as well as faster than SQLite 3. In a recent (very simple) test, the MVStore was about twice as fast as SQLite 3 on Android.
mvstore_1184_p=\ The API of the MVStore is similar to MapDB (previously known as JDBM) from Jan Kotek, and some code is shared between MVStore and MapDB. However, unlike MapDB, the MVStore uses is a log structured storage. The MVStore does not have a record size limit.
mvstore_1185_h2=Current State
mvstore_1186_p=\ The code is still experimental at this stage. The API as well as the behavior may partially change. Features may be added and removed (even though the main features will stay).
mvstore_1187_h2=Requirements
mvstore_1188_p=\ The MVStore is included in the latest H2 jar file.
mvstore_1189_p=\ There are no special requirements to use it. The MVStore should run on any JVM as well as on Android.
mvstore_1190_p=\ To build just the MVStore (without the database engine), run\:
mvstore_1191_p=\ This will create the file <code>bin/h2mvstore-1.3.175.jar</code> (about 130 KB).
