The Apache Software Foundation
Quick Start
The Apache Software Foundation
The Apache Software Foundation
The Apache Software Foundation
Documentation
The Apache Software Foundation
Blogs and Articles About Derby
Tutorials, Tips and Tuning
Tools and Migration
Applications
Integration With Other Products
Eclipse Plug-ins
Papers and Presentations
Derby Engine
Derby Network Client
Derby Tutorial
Presentations
The Apache Software Foundation
Font size:
Apache Derby 10.5.1.1 Release
Distributions
Use the links below to download a distribution of Apache Derby. You should always
of distribution files downloaded from a mirror.
There are four different distributions:
bin distribution - contains the documentation, javadoc, and jar files for Derby.
lib distribution - contains only the jar files for Derby.
lib-debug distribution - contains jar files for Derby with source line numbers.
src distribution - contains the Derby source tree at the point which the binaries were built.
[] [] (Note that, due to long filenames, you will need gnu tar to unravel this tarball.)
There are two separate Eclipse plugins for Derby:
derby_core_plugin - provides the Derby jar files to other plugins in Eclipse.
derby_ui_plugin - provides an Apache Derby Nature in Eclipse for easy database application development.
Please note: both plugins must be installed for full functionality. For information on installing and using the Derby plugins for Eclipse, please see the
Release Notes for Derby 10.5.1.1
These notes describe the difference between Derby release 10.5.1.1 and the preceding release 10.4.2.0.
Derby is a pure Java relational database engine using standard SQL and JDBC as its APIs.
Derby functionality includes:
Embedded engine with JDBC drivers
Network Server
Network client JDBC drivers
Command line tools: ij (SQL scripting), dblook (schema dump) and sysinfo (system info)
New Features
This is a feature release. The following features were added.
SQL Roles - SQL roles (as described in SQL 2003 and errata) are useful for administering privileges for groups of users. Administering roles is generally less error-prone than administering privileges for individual users. SQL Roles are defined in Feature T331 "Basic roles" and Feature T332 "Extended roles". Derby 10.5 implements a subset of T331, plus support for CURRENT_ROLE, which is a part of T332.
Generated Columns - Generated Columns is a feature which instructs Derby to fill a column with an expression built out of other columns in the row. Derby constructs these column values at INSERT and UPDATE time. The user declares indexes on these columns. This in turn improves SELECT performance since it lets users declare indexes on pre-computed pieces of the WHERE clause. This feature satisfies support for SQL Feature T175.
LOB Improvements - There were many performance and usability improvements for BLOBs and CLOBs.
Replication of encrypted databases - With 10.5 it is possible to replicate encrypted databases.
OFFSET/FETCH FIRST - SQL 2008 has added new syntax to support a direct way to limit the returned set of rows in a result set, through the fetch first and offset clauses.
In-memory back end - Initial implementation of a storage engine for Derby where all data is kept in memory. There is no documentation for this feature. This functionality itself is not yet fully implemented, but users are welcome to experiment with it. For details, see the .
Standard ALTER COLUMN syntax - Allow standard SQL "SET" keyword in ALTER COLUMN syntax, like so: "ALTER TABLE ALTER COLUMN columnname SET DEFAULT default-value"
SYSCS_UTIL.SYSCS_UPDATE_STATISTICS - New system procedure that updates cardinality statistics (or creates them if they do not exist) for a table's index or for all the indexes on a table, allowing a user to ensure that a query plan based on the most recent state of the table can be created.
The following issues are addressed by Derby release 10.5.1.1. These issues are not addressed in the preceding 10.4.2.0 release.
Issue IdDescription
NullPointerException in system.sttest run with 10.5.1.0
Cannot insert row when a generated column references an identity column
NullPointerException on CREATE TABLE when generated column depends on another generated column
XAExceptions thrown by Derby can have errorCode 0
Failure in BlobClob4BlobTest due to wrong results being returned
Failure in ServerPropertiesTest due to java.security.AccessControlException on the server side, in 10.4 to 10.5.1. soft upgrade mode
ArrayIndexOutOfBoundsException in DatabaseMetaDataTest when run in Soft Upgrade mode
You can't invoke a table function which is stored in a jar file inside the database
The in-memory storage back end doesn't work on Windows
ClassCastException in SQLClob when running in soft upgrade mode (10.4.2.0 -& 10.5.1.0)
Documentation: more UPDATE_STATISTICS fixes needed for Reference Manual and Tuning Derby
Compress on a large table fails with IllegalArgumentException - Illegal Capacity
dblook script fails when URL contains special characters
SYSCS_UTIL.SYSCS_UPDATE_STATISTICS should update the store estimated row count for the table
Documentation: update documentation files to 10.5
Allow unlimited growth for in-memory conglomerates
'testRunTimeStatistics(org.apache.derbyTesting.functionTests.tests.lang.OffsetFetchNextTest)junit.framework.AssertionFailedError' on Windows
Trigger fails with ERROR 38000: The exception 'java.sql.SQLException: ResultSet not open during VTIResultSet.getNextRowCore()
Enforce absolute paths in the in-memory back end internal storage structure
Improve handling of the database temporary directory
DDMReader readBytes ArrayIndexOutOfBoundsException
Determine the subSubProtocol name for the in-memory back end
Add support for SQL:2008 &result offset clause& and &fetch first clause& to limit result set cardinality
Document the SQL Standard routine matching rules which Derby now supports
Creation/configuration of ClientXDataSource fails because of two setSsl methods
shutdown with incorrect permission on log files shows java.lang.NullPointerException at org.apache.derby.impl.store.raw.log.LogToFile.flush(LogToFile.java:3964). Should give bettter message.
AssertFailure when selecting rows from a table with CHARACTER and VARCHAR columns
Constraint causes wrong query result when using exists
InputStream returned from Blob.getBinaryStream(long, long) terminates the stream by returning 0, should return -1
Blob.getBinaryStream(long,long) is off by one for the pos+len check
If space reclamation cannot obtain container lock it will not retry getting the lock
Multithreaded clob update causes growth in table that does not get reclaimed
SELECT statement topic in Reference Manual should reflect current WHERE clause behavior
Update Statistics code in AlterTableConstantAction should use execute transaction rather than compile transaction
org.apache.derby.impl.load.Import needs to escape single quotes
SQLChar.getLength returns wrong length for some data values
two rows can be inserted with the same value in a column that a unique constraint on that column should prevent
An attempt was made to access an out of range slot on a page
CHAR data type documentation fails to describe the maximum allowed length
Replace felix.jar file with the felix sources
Remove dependency of junit targets on deprecated jdk16 variable
Removed deprecated jdk16 variable from the machinery which builds javadoc
Document syntax change for ALTER TABLE in Reference Manual
Update Derby Jira Report XML VTI to access more fields
Allow standard SQL syntax: ALTER TABLE ALTER COLUMN &col& SET DEFAULT &default&
'largedata/LobLimits.java' fails with "EXPECTED SQL Exception: (XJ001) Java exception: 'A truncation error was encountered trying to shrink CLOB 'XXXX' to length .: org.apache.derby.iapi.services.io.DerbyIOException'."
PassThroughException should not reimplement initCause()/getCause()
Only send RDBNAM on ACCSEC if EBCDIC conversion is possible
ALTER COLUMN ... WITH DEFAULT NULL does not change the default
Remove required RDBNAM from ACCSEC. Use SECCHK RDBNAM if none is provided on ACCSEC
ORDER BY causes column to be returned
Change eol style to native on jsr169 stubs
ResultSetStreamTest fails on case sensitive file systems, and isn't run as part of any suite
Set the Java 5 compilation path based on the Java 6 compilation path if a Java 5 environment can't be found but a Java 6 environment can be
Set the Java 6 complilation path based on the Java 5 compilation path so that the JDBC4 support is always built
Replace BUILDING.txt with a simlified up-to-date BUILDING.html
Separate out the small number of build targets which really require JDK1.4 features from the bulk of the targets which ought to compile cleanly against small device libraries
User Guide documentation on the limitations of small-device support is stale
Add a facility to override the default set of old versions to be tested in the upgrade tests.
Improve distribution of hash codes in SQLBinary and SQLChar
Clob.truncate(long) in the client driver doesn't update the cached Clob length
Clob.truncate with a value greater than the Clob length raises different exceptions in embedded and client driver
SELECT DISTINCT may return duplicates with territory-based collation
Update test harness to run with DesktopEE JRE
PositionedStoreStream doesn't initialize itself properly
NPE if you declare a constraint on a generated column and omit the datatype
Do not require JRE 1.4 libraries at compile time since SUN's JDK 1.4 went out of support and can be used on JRE 1.4 without any problem
NullPointerException when re-evaluating generated column during ON DELETE SET NULL referential action
Reference Manual needs information on generated columns
ImportExportProcedures imports ./testData/ImportExport/position_info.del using US-ASCII but the file contains non-ascii characters
Remove method TemplateRow.checkPartialColumnTypes
You should get an error if you try to override a generated column via an INSERT driven by an inner ResultSet
Don't allow references to generated columns in the NEW transition variables of triggered actions of BEFORE triggers
Generation clauses which mention user-coded functions may produce different resuls depending on who performs the triggering INSERT/UPDATE
testReplication_Local_3_p2_StateTests_smallInsert_immediateStopMaster_DISABLED() causes hang in teardown()
Document that you can change the length of varbinary as well as varchar columns
Add CharacterStreamDescriptor
Introduce interface for a position aware stream
Improve performance of reading modified Clobs
Add tests of generated columns when sql authorization is turned on
GeneratedColumnsTest fails with "'F_MINUS' is not recognized as a function or procedure." on cvm/phoneME
SQL roles: Add VTI for CONTAINED_ROLES
Make the datatype optional when declaring columns with generation clauses
Enable the adding of generated columns via ALTER TABLE
RolesConferredPrivilegesTest fails with 'java.lang.NoClassDefFoundError: java.sql.DriverManager' with weme 6.1
BTreeController.reclaim_deleted_rows() may hide exceptions
Race condition in NetXAResource.removeXaresFromSameRMchain()
Save useful length information for Clobs in store
Failed tests should save the database off to the fail directory
NPE on left join with aggregate
'The class 'org.apache.derby.diag.EnabledRoles' does not exist or is inaccessible.' in new roles tests (DERBY-3886)
some messages are referenced but never used - MessageBundleTest output
SQL roles: Move upgrade logic from 10.4 to 10.5
SQLSessionContext not correctly initialized in some non-method call nested contexts
Needs update of master for 'derbyall/derbynetmats/DerbyNet/derbynetmats/holdCursorJDBC30'?
Reference Manual: supported functions should be listed in JDBC escape syntax topic
Replication: NPE for startSlave of encrypted database
LOBStreamControl.truncate() doesn't delete temporary files
SQL roles: ij show enabled and settable roles
LOBStreamControl.replaceBytes() leaves temporary files open
NPE on a query with having clause involving a join
Getting Started Guide has typo and version problems
Replication: stopSlave does not close serversocket when master has crashed.
SQL roles: build support for dblook
Derby cannot replace a database after encountering corruption
Document numeric function ATAN2(float1,float2)
NullPoinerException thrown when INTEGER function used as a predicate in a WHERE clause of a SELECT .. GROUP BY .. HAVING statement
EmbedBlob.setBytes returns incorrect insertion count
intermittent hang pinging server on Linux
NPE in testSemantics(.....lang.RolesTest) on JavaME / CVM
Remove utility class PrivilegedFileOps
improve test importExportIJ.sql
Reset the current role when EmbedPooledConnection creates a new logical connection
The Tuning guide should mention that LIKE transformations/optimizations are disabled when using territory-based collations
Remove unneeded workarounds for DERBY-177 and DERBY-3693
Tuning Guide lacks information on turning off Derby property settings
Problems running org.apache.derbyTesting.system.optimizer.RunOptimizerTest
'testTimeSlice_*(NetworkServerControlApiTest.java)' fails on Jvm1.5/vista, Jvm1.4/SunOS 5.11
The test code executes java processes by just executing java instead of using a full path. This may cause the wrong java to be picked up.
need to test getting appropriate error message when os files are read-only
Convert derbynet/getCurrentProperties to JUnit
Convert derbynet/timeslice.java to Junit
RuntimeStatisticsParser can not distinguish names of table or index well and truly
Convert jdbcapi/SURTest_ij.sql to JUnit.
Add Apache 2.0 license headers to a number of files
Remove the update-with-jdbc4 script, which is dead code now
StoreStreamClob.getReader(charPos) performs poorly
client Insert/retrieval of 18MB Clob is extremely slow in MultiByteClobTest
Administration Guide topics on unsupported DB2 driver should be removed
Derby tests for the existance of BigDecimal methods toPlainString and bdPrecison but does not check if they were found before using them.
failure in testAttributeDrdaStreamOutBufferSize(org.apache.derbyTesting.functionTests.tests.management.NetworkServerMBeanTest)junit.framework.AssertionFailedError: expected:&0& but was:&131072&
Create a simple Clob performance regression test
traces of removed ij functionality of ij.unicodeEscape should be removed
Create a simple Blob performance regression test
'org.apache.derbyTesting.functionTests.tests.lang.XMLTypeAndOpsTest' failures on JVM 1.5 on trunk, 10.4 and 10.3
Convert "org.apache.derbyTesting.functionTests.tests.lang.optimizerOverrides.sql" to junit.
Convert derbynet/maxthreads to JUnit
NullPointerException when accessing a clob through a pooled connection
Remove unnecessary call to Hashtable.get() in TableScanResultSet.getNextRowCore()
Convert jdbcapi/metadataMultiConn to JUnit.
Convert derbynet/OutBufferedStream to JUnit
Remove unnecessary methods from InternalClob interface
Excessive memory usage when fetching small Clobs
Document the new system stored procedure SYSCS_UTIL.SYSCS_UPDATE_STATISTICS which was added as part of DERBY-269
Assert failure in CacheEntry.unkeepForRemove when running stress.multi
ImportExportTest doesn't detect lack of expected errors
LOBStreamControl shouldn't throw SQLException
Client Configuration.java imports engine class org.apache.derby..JVMInfo
PositionedStoreStream.reposition(pos) with pos greater than length leaves the stream object in an inconsistent state
Run junit tests with -Dfile.encoding="UTF-16" to expose encoding issues and analyze failures
Add client side JDBC statement pool documentation
SecureServerTest, SSLTest are failed on Zos with exit code 143 starting network server
testGetBytes under BlobClob4BlobTest failed on Zos with encoding issue
BlobStoredProcedureTest failed on Zos: AssertionFailedError: Error SYSIBM.BLOBGETPOSITIONFROMLOCATOR returns the wrong value for the position of the Blob expected:&8& but was:&-1&
jdbc4/ClobTest fails on Zos with AssertionFailedError: Streams differ at index 0 expected:&200& but was:&72&
ImportExportLobTest failed on Zos Clobs differ at index 1 expected:&99& but was:&196&
testClasspathChecker under SysinfoCPCheckTest failed on Zos
Create a utility class for skipping data in an InputStream
Make LOBStoredProcedure on the server side smarter about the read buffer size
Make EmbedBlob.length use skip instead of read
Convert "org.apache.derbyTesting.functionTests.tests.lang.nulls.sql" to junit.
EmbedBlob.setPosition is highly ineffective for streams
Union Query fail on Derby 10.4.1.3
Rename BaseJDBCTestCase.usingDerbyNet
Convert "org.apache.derbyTesting.functionTests.tests.lang.arithmetic.sql" to junit.
Convert "org.apache.derbyTesting.functionTests.tests.lang.miscerrors.sql" to junit.
Convert "org.apache.derbyTesting.functionTests.tests.lang.ungroupedAggregatesNegative.sql" to junit.
Convert "org.apache.derbyTesting.functionTests.tests.lang.precedence.sql" to junit
ij's help text lacks the optional HOLD | NOHOLD syntax for GET CURSOR
Convert "org.apache.derbyTesting.functionTests.tests.lang.connect.sql" to junit
Convert case.sql to junit
Convert "org.apache.derbyTesting.functionTests.tests.lang.constant Expression.sql" to junit
Derby can leak classloaders in an app server environment
Revoking EXECUTE privilege on a function if used in a CHECK constraint: implementation problem
Move test code from CompressedNumber to a unit test
SQL LENGTH function materializes CLOB into memory
Skip and read methods in ArrayInputStream may overflow
Add more tests for legal/illegal commands in the different replication states
Revoking a column level privilege from a user, a prepared statement relying on that privilege can still be executed
Incorrect position calculation in PositionedStoreStream with read(bytearray,...)
Maximum value allowed for derby.storage.fileCacheSize (100) is too low for large system. Increase the maximum value and redocument the property.
SQL Length function materializes BLOB into memory
Improve calculation of refSize in ClassSize.java
Bundle-SymbolicName: needed in Derby manifest for OSGi 4 environment
Fix error handling in PrivilegedFileOpsForTests
Don't call RAFContainer.padFile() from instances of RAFContainer4
add more information to the XSDB1:ERROR XSDB1: Unknown page format at page error
Reset current schema to default (user name) when creating a new logical connection in the client driver
Add circularity check for the GRANT role statement
NPE when firing a trigger
convert derbynet/getCurrentProperties.java to junit
java/testing/README.htm refers to a non-existing test
convert store/aes.sql to junit test & add unrestricted test cases.
setting tracedirectory from the command line does not work
STATUS file changes June 2008
NetworkServer console messages should print a time stamp
In Net Client mode, negative values for stream length are accepted without an exception for PreparedStatement.setAsciiStream()
If an IOException is encountered during establishment of the connection, Network Server should print the root exception to the console instead of a generic message
Make it possible to build the JSR169 support with the jdk1.4 libraries
Remove xml-apis.jar because we no longer have to compile against jdk1.3
java.lang.Exception: DRDA_UnableToAccept.S:Unable to accept connections and client hang if tracing is turned on but traceDirectory does not exist
Get rid of tools/java/empty.jar
Delete superfluous test lang/longStringColumn.java
NullPointerException when invoking statement event listeners if one of the listeners is null
'Method resolution for signature java.lang.String.valueOf(int) was ambiguous. (No single maximally specific method.)' in 'testFunctionNullHandling(org.apache.derbyTesting.functionTests.tests.lang.RoutineTest)' on JavaME
Deadlocks accessing DB metadata
'javax.transaction.xa.XAException' ++ in 'J2EEDataSourceTest'
EmbedPooledConnection doesn't reset schema when creating a new logical connection
convert predicatePushdown.sql to JUnit
weme 6.1 Base connection pooling suitejunit.framework.AssertionFailedError: java.lang.NullPointerException
Wire the vti demo into the build.
SYSCS_BULK_INSERT doesn't quote identifiers or strings properly
When authenticating a user at connect time, verify that the user provided is not also a defined role name.
StackOverflowException in deadlock trace
'ALTER TABLE' is not allowed on the System table '"SYSCS_DIAG"."TRANSACTION_TABLE"' in derbyall - store/TransactionTable.sql
Add checks that a new role isn't already a user authorization id
Remove JDBC 3.0-specific topics from Reference Manual and merge implementation notes as needed
SQL roles: Make CURRENT_ROLE check that the role is still valid
Make role descriptor a dependency Provider
Master file for derbynet/testProperties.java needs update for 10.3.2.2 to 10.3.3.0 version change.
Convert store/streamingColumn to JUnit
Typo yields redundant test case in GrantRevokeTest
revive tests with ldap references to be runnable as junit tests
LOBStateTracker should not use SYSIBM.CLOBRELEASELOCATOR when the database is soft-upgraded from 10.2
Comment in template security policy incorrectly says that JMX is not enabled by default
errror in nightly regression test: LobStreamsTest:encryptedjunit.framework.AssertionFailedError: f:\jartest\JarResults.\ibm16_suites.All\system\singleUse\oneuse1e\tmp\lob6165.tmp
Convert datetime.sql to JUnit
Derby does not follow the SQL Standard when trying to map SQL routines to Java methods.
can't call a stored function with an aggregate argument without getting the following error: ERROR 42Y29
Update and rectify the Derby tutorial
Add an example for table function in Derby Reference Manual
java/testing/Readme.htm location of derbyTesting.jar
tools/derbyrunjartest.java fails with 'No command given.' (expecting 'No arguments given.')
UDF used with aggregate arguments results in error 30000
Tools Guide should document continuation marker for ij
XSDA3 error in concateTests in lang.LangHarnessJavaTest caused by bug in SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE()
Convert deadlockMode.java to JUnit
SYSCS_UTIL.SYSCS_EMPTY_STATEMENT_CACHE needs a better description in the reference manual
Implement more load types for org.apache.derbyTesting.perf.clients.Runner
Perform thread dump with ASSERTS with jdk 1.5 or higher
Making the instructions more clear in documentation(BUILDING.txt) regarding setting JRE lib path for a Windows system
SELECT DISTINCT field FROM TABLE_NAME GROUP BY field, field2
Developer's Guide needs correction on garbage collection
If derbytesting.jar is in a different directory than the derby jars SystemPrivilegesPermissionTest fails with java.security.AccessControlException
Optimize LOBStateTracker for non-locator servers
Incorporate DERBY-3310 and DERBY-3494 write-ups into NormalizeResultSetNode javadoc
Creation of logical connections from a pooled connection causes resource leak on the server
AllocPage.createPage() doesn't initialize minimumRecordSize correctly
suites.All fails to run on Jvm 1.5 when built with JDK 1.5 (Failed to invoke suite(): .jdbc4._Suite)
Convert jdbcapi/testRelative.java to JUnit
Remove am.Connection.reset(LogWriter,ClientBaseDataSource,boolean) and called methods
Replication tests shall not be run on Java ME
IndexOutOfBoundsError in ClockPolicy.moveHand
Changing certain properties on client DataSource objects causes existing connections to reflect the new values
Remove unused method Connection.resetConnection(LogWriter, String, Properties)
The Developer's Guide incorrectly describes the behavior of transactions inside procedures and functions
DrdaStreamOutBufferSize attribute of NetworkServerMBean should be of type int instead of String
Merge EngineBlob and EngineClob into a single interface
With client, attempting to get the lob length after commit or connection close if there was a call to length() before commit does not throw an exception
directory error in the file "java/testing/README.htm"
LOB locators are not released if the LOB columns are not accessed by the client
Implement DETERMINISTIC keyword for procedures and functions
Convert jdbcapi/savepointJdbc30_JSR169.java and jdbcapi/savepointJdbc30_XA.java to JUnit
AsynchronousLogShipper#forceFlush should time out
Alter column set data type not allowed in soft upgrade with unique constraint
Number of log files (and log dir size) on the slave increases continuously
testStartStopManagementFromApplication(org.apache.derbyTesting.functionTests.tests.management.ManagementMBeanTest)junit.framework.AssertionFailedError: expected:&2& but was:&5&
regression test failure in testDerbyJarAttributeAlpha - Security Exception
change derby.tests.trace property to print the name of the test before it runs it
Change Collation test to run DatabaseMetaDataTest, BatchUpdateTest,GroupByExpressionTest, and UpdateableResultSetTest for only one locale
Implement procedure SYSCS_UTIL.SYSCS_PREPARE_REPLICATION()
Unable to start slave mode after authentication failure on a previous startSlave attempt
NoClassDefFoundError failure in SystemPrivilegesPermissionTest for weme6.1
Failed to get database schemas of a JAR database
If NetworkServer fails to shutdown when run as a separate process by NetworkServerTestSetup, test run will hang
NetworkServerControl with options but no command does not give usage message
NullPointerException during execution for query with LEFT OUTER JOIN whose inner table selects all constants.
DatabasePermission uses a JDK 1.4 method that is not present in J2ME/CDC/Foundation.
The slave will not notice that a network cable is unplugged and will therefore reject failover/stopSlave commands
AsynchronousLogShipper#workToDo is blocked while the log shipper sends a log chunk
Remove unneeded code to get JDBC level in BrokeredConnection and BrokeredStatement classes
Functionality for skipping testsuites on certain platforms fails for the new management testsuite on phoneME advanced
convert views.sql to junit
23 timeouts encountered while running management junit suite as part of full nightly runs.
The replication log shipper is not notified when a new replication transmitter is instantiated in MC#handleException.
Log receiver thread fails with NPE at failover when master has died
Add JMX testing with a security manager and investigate testing with JMX authentication.
Change stress.multi to dump thread stacks before killing off testers with jdk 1.5 and higher
Move the setup of NormalizeResultSetNode into the NormalizeResultSetNode
Error message XRE04 does not include the right port number.
Clob.length() doesn't detect a closed underlying connection in a consistent way
Require new permissions in o.a.d.security.SystemPermission to allow control to Derby's JMX management and to ensure information is not leaked through JMX
Shutdown on a database without stopping replication hangs
Make ResultSet.getStatement return the correct statement when created by a logical statement
DatabaseMetaData.getConnection returns the wrong connection when using connection pooling
The -ca option should be removed from ij and from the ij command usage line
Remove JDBC 2.0-specific topics from Reference Manual and merge implementation notes as needed
Wrong message when using SHOW in ij.
Removing a ConnectionEventListener from a PooledConnection during its connectionClosed() callback causes other ConnectionEventListener callbacks to be missed
Derby 10.3.1.4 and 10.3.2.1 break scrollable result sets? Hibernate Query.setFirstResult and/or setMaxResults
SQLException thrown from user function kills network connection
Tests for JMX Management and Monitoring
"ERROR XSDA3: Limitation: Record cannot be updated or inserted due to lack of space on the page...." in suites.All
"No Current connection" on PooledConnection.getConnection() if pooled connection is reused during connectionClosed processing
SQL "distinct" and "order by" needed together
Invalid method java.lang.Integer && void &init&(short) because java.lang.NoSuchMethodException: java.lang.Integer.&init&(short)
Select from large lob table with embedded gives OutOfMemoryError
ERROR XSDB3: Container information cannot change once written
CancelQueryTask.forgetContext() could be simplified.
SQL roles: Implement authorization stack (and SQL session context to hold it)
Database creation and boot should fail if collation=TERRITORY_BASED and the selected locale is not supported
Logical connections do not check if a transaction is active on close
ASSERT in MergeSort.checkColumnTypes() disallow legal type conversions
NPE in PooledConnction event notification handling if a null listener is added
Delayed (on-demand) creation of current user schema makes select from view belonging to other schema fail.
Not possible for non-db-owner to create a temporary table. "Get ERROR 42507: User 'USERB' can not perform the operation in schema 'SESSION'."
Remove final from all getConnection() methods in EmbeddedDataSource
SQL roles: make use of privileges granted to roles in actual privilege checking
Group by query with many aggregate columns and case statements fails with: ERROR XSDA7: Restore of a serializable or SQLData object of class , attempted to read more data than was originally stored
Developer's Guide: Add examples showing use of SQL authorization with user authentication
SQL roles: Add documentation
error XSDA3 when test is executing SYSCS_INPLACE_COMPRESS_TABLE in specific situation
Reduce dependency on NetworkServerControl in TestConfiguration
SQL roles: add catalog support
Reduce memory footprint of StoredRecordHeader
totalSpace not properly initialized in AllocPage
Unnecessary if statement can be removed from BaseActivation.getColumnFromNow
Extend the system function mechanism to support functions with multiple arguments
'Schema &schemaname& does not exist' when constraint used in table definition
Cannot restore backups without BACKUP.HISTORY file
Index split deadlock
need to mention ClientDriver In Derby reference manual's java.sql.Driver interface section
Closing a resultset after retrieving a large & 32665 bytes value with Network Server does not release locks
Thread safety issue in TableDescriptor
Add caching of store stream length in StoreStreamClob, if appropriate
Convert lang/declareGlobalTempTableJava.java to JUnit
Reply incorrectly handles read() returning -1
convert lang/closed.java to junit
NullPointerException on ClientDataSource.getConnection() when ds.setdatabaseName was invalid
DOCS - Move Derby system and properties info from Tuning Guide into Reference Manual
DOCS - Reorder JDBC Reference section in Ref Manual
intermittent NPEs during DELETE ops in a reasonably large transaction
ORDER BY with expression with distinct in the select list returns incorrect result
Example in DEGREES function is not appropriate.
Improve usability of Derby's client/server security by implementing ANSI Roles
Misleading error message for non-matching ORDER BY clause in queries with GROUP BY.
jdbcapi/ParameterMappingTest.java test does not execute test for setObject(Blob/Clob) in DerbyNetClient
Reference Manual updates - JDBC Reference section
jdbcapi/SetQueryTimeoutTest.java fails on IBM wctme 5.7
Rewrite stress.multi as a JUnit test
Make i18n/LocalizedDisplay.sql and i18n/LocalizedConnectionAttribute.sql behave equally on different platforms
Remove optional osgi.jar download by including equivalent jar from the apache incubator felix project in the Derby's svn
Add MessageBundleTest to build script to verify that there are no orphaned message ids
Message id J029 (CONN_RESTORE_FROM) does not have a matching message text
Message id J030 in reference.MessageId.java does not have a matching message
Possible values for derby.storage.rowLocking Derby engine is not currently documented eventhough the property is
DRDAConnThread uses SQLState.NO_SUCH_DATABASE, which has no matching message
Add JMX extensions to Derby
Derby's "set schema" behavior is not compliant with SQL 2003 Foundation spec.
For existing databases JDBC metadata queries do not get updated properly between maintenance versions.
Change the internal implemetation of SYSCS_INPLACE_COMPRESS_TABLE to share existing alter table
Error 22001 can include 'XX-RESOLVE-XX' which should be removed
Calling SYSCS_INPLACE_COMPRESS_TABLE on SYS.SYSTABLES results in a lock timeout, with only a single connection active
xaHelper in ij creates global id that is not the same across platforms
Documentation should recommend using .newInstance() to instantiate JDBC driver
implement SQL generated columns
Provide some way to update index cardinality statistics (e.g. reimplement update statistics)
A connection request that has a default schema that is being created by another transaction will fail to connect
Compared with the previous release (10.4.2.0), Derby release 10.5.1.1 introduces the following incompatibilities. These merit your special attention.
Note for DERBY-4073: The method setSsl(int) has been removed from the client data source classes.
Note for DERBY-4050: Table growth may result from multi-threaded Clob updates before the fix for DERBY-4050.
Note for DERBY-4042: Format of file name arguments to the import procedures changed for files with single quotes (') in their names.
Note for DERBY-4008: After the change for DERBY-4008, applications may see a different error message when attempting to connect a 10.5 client to older revision servers (e.g. 10.4 and 10.3) with multibyte database names.
Note for DERBY-3977: A different exception is thrown by the embedded driver when trying to truncate a Clob with a too large length argument.
Note for DERBY-3701: An error message will be logged to derby.log if the Network Server tracing file cannot be created. Starting with version 10.5, the Network Server will attempt to create the trace directory if it does not exist. Any intervening directories in the given path will also be created if possible.
Note for DERBY-3652: Derby now follows the SQL Standard rules for matching functions and procedures to Java methods.
Note for DERBY-3420: The -ca command line option has been removed from ij.
Note for DERBY-3347: A bug that could cause unrecoverable database corruption has been fixed.
Note for DERBY-3327: The effect of setting the current default schema (SET SCHEMA schemaname) inside nested connection of a stored procedure or function has been changed to comply with SQL standard semantics.
Note for DERBY-3319: Exception is thrown when connection with uncommitted operations is closed.
Note for DERBY-2351: An ORDER BY clause of a DISTINCT query which specifies to order by a column which was not in the DISTINCT list is now rejected, because the intent of the query is ambiguous. Previously, Derby instead produced non-distinct results. Also, an ORDER BY clause which specifies a table-name-qualified column alias is now rejected as invalid, where previously it was accepted.
Note for DERBY-2085: Derby has improved the error message which is issued when an invalid column reference is found in a grouped query. Derby now issues message 42Y36 instead of 42Y30. Furthermore, the wording of message 42Y36 has been expanded to further explain the behavior.
Note for DERBY-1062: Applications will see a different error code and message when they attempt to call SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE on a table that does not exist.
Note for DERBY-48: In Derby, a user's initial default schema is named the same as the user name, or APP if a user is not provided at connect time. This schema is implicitly auto-created the first time a schema object is created in that schema.
Note for DERBY-4073
Summary of Change
The method setSsl(int) has been removed from the client data source classes.
Symptoms Seen by Applications Affected by Change
The application will either fail to compile, or experience a NoSuchMethodError. Only applications invoking the removed method are affected.
Incompatibilities with Previous Release
The method setSsl(int) can no longer be used to configure SSL with the client data sources.
Rationale for Change
Having two setSsl-methods caused choice problems for some applications using Derby data sources and also configuring the data source through introspection, since there were two methods called setS setSsl(int) and setSsl(String).
Application Changes Required
Use setSsl(String) instead of setSsl(int).
Note for DERBY-4050
Summary of Change
Table growth may result from multi-threaded Clob updates before the fix for DERBY-4050.
Symptoms Seen by Applications Affected by Change
Before the fix for DERBY-4050, applications may have seen growth in the size of the database, particularly the size of the dat files for Clob tables, when doing updates to the Clob tables from multiple threads. The fix for DERBY-4050 will prevent future growth but will not reclaim space consumed by the bug in the past. To reclaim the space, a full offline compress of the clob table is needed. See the Derby documentation for usage of
Incompatibilities with Previous Release
Rationale for Change
DERBY-4050 fixes a bug in space reclamation.
Application Changes Required
The compress table should only need to be run once if the fix for DERBY-4050 has been applied. No application changes should be required.
Note for DERBY-4042
Summary of Change
Format of file name arguments to the import procedures changed for files with single quotes (') in their names.
Symptoms Seen by Applications Affected by Change
In the previous release, applications that imported data from a file whose file name contained a single quote (') character would fail unless they replaced each single quote in the file name argument to the import procedure with two single quotes. Now this has changed and Derby does not understand the file name argument unless it has the same number of single quote characters as the actual file name. Applications that worked around the problems in earlier releases by adding extra single quotes, will now see exceptions like the following:
ERROR XIE04: Data file not found: Rock''n''roll.csv
Applications that do not import files whose names contain single quotes will not be affected by this change.
Incompatibilities with Previous Release
Previous releases required file name arguments to the import procedures to double each occurrence of a single quote in the file name, like this when importing a file called C:/Chip's/TERMS.dat:
PreparedStatement ps = conn.prepareStatement(
"CALL SYSCS_UTIL.SYSCS_IMPORT_TABLE(null,'MY_TABLE',?,null,null,null,0)");
ps.setString(1, "C:/Chip''s/TERMS.dat");
ps.execute();
In this release the above code would result in the following exception:
ERROR XIE04: Data file not found: C:/Chip''s/TERMS.dat
In this release, applications must use the actual name of the file, like this:
PreparedStatement ps = conn.prepareStatement(
"CALL SYSCS_UTIL.SYSCS_IMPORT_TABLE(null,'MY_TABLE',?,null,null,null,0)");
ps.setString(1, "C:/Chip's/TERMS.dat");
ps.execute();
Rationale for Change
It is more intuitive if the file name arguments match the actual file names. Also, in the previous releases there were some combinations of single quotes and other special characters (like double quotes) that it was not possible to get to work even with the workaround.
Application Changes Required
Applications that work around the issue in previous releases by adding extra single quotes to the file name arguments, must be changed so that they do not add extra single quote characters.
Note for DERBY-4008
Summary of Change
After the change for DERBY-4008, applications may see a different error message when attempting to connect a 10.5 client to older revision servers (e.g. 10.4 and 10.3) with multibyte database names.
Symptoms Seen by Applications Affected by Change
In previous releases when attempting to connect to network server using a multibyte database name the connect would fail with the the exception:
SQLSTATE:22005 java.sql.SQLDataException: Unicode string cannot convert to Ebcdic string
This message will not change for 10.5 clients connecting to 10.5 servers. After the change for DERBY-4008 the error message may change when connecting 10.5 clients to older revision server. Instead the message will be:
SQLSTATE:08006 java.sql.SQLNonTransientConnectionException: A network protocol error was encountered and the connection has been terminated: A PROTOCOL Data Stream Syntax Error was detected. Reason: 0x8,464. Plaintext connection attempt to an SSL enabled server?
with a chained SQLException:
SQLSTATE:22005 java.sql.SQLDataException: Unicode string cannot convert to Ebcdic string
The protocol error will also show on the server console.
Incompatibilities with Previous Release
The client error message when attempting to connect with multibyte database names will change with mixed revision client/server as described in the previous section.
Rationale for Change
This change was made in preparation for fixing DERBY-728 to allow multibyte characters in the database name. The change in error message with back revision servers was considered an acceptable change and necessary to fix DERBY-728.
Application Changes Required
To reveert to the previous error message, users can upgrade their server to the latest on the 10.3 or 10.4 branch, or upgrade their server to 10.5. Otherwise applications expecting the old error message should upon gettting the SQLState: 08006 message retrieve the chained exception with SQLExcepiton.getNextMessage() and then process the message as before.
Note for DERBY-3977
Summary of Change
A different exception is thrown by the embedded driver when trying to truncate a Clob with a too large length argument.
Symptoms Seen by Applications Affected by Change
In the previous release, the embedded driver would throw an exception with SQLState XJ076 when trying to truncate a Clob with a length argument bigger than the Clob length. Now, it will throw the same exception as the client driver, XJ079. The error messages also differ:
Before: XJ076: The position argument '18149' exceeds the size of the BLOB/CLOB.
Now: XJ079: The length specified '18149' exceeds the size of the BLOB/CLOB.
Incompatibilities with Previous Release
Applications catching a specific exception based on SQLState may behave differently. The incompatibility can only be seen if the application calls Clob.truncate with the embedded driver.
Rationale for Change
To make the embedded and the client driver consistent.
Application Changes Required
Look for SQLState XJ079 instead of XJ076 when Clob.truncate() is called.
Note for DERBY-3701
Summary of Change
An error message will be logged to derby.log if the Network Server tracing file cannot be created. Starting with version 10.5, the Network Server will attempt to create the trace directory if it does not exist. Any intervening directories in the given path will also be created if possible.
Symptoms Seen by Applications Affected by Change
Before the fix for DERBY-3110, if derby.drda.traceAll was set to true and the derby.drda.traceDirectory was set to a non-existent directory, no tracing would occur and no error would occur. After the fix for DERBY-3110, an error "java.lang.Exception: DRDA_UnableToAccept.S:Unable to accept connections" would occur and the client would hang and no tracing would occur. With this fix for version 10.5 and higher, the Network Server will attempt to create the trace directory if possible. For 10.4.2 (and the next release on the 10.3 branch), the Network Server will still not try to create the directory. For all these releases the Network Server will print an error on session connect if there is any problem creating the trace file, but the Network Server will not cause the session connection to fail. Users who have trace turned on and the trace directory set to a non-existent directory may now see exceptions in the derby.log on connect indicating that the trace file is not found or with 10.5 or higher they may see tracing occur where it did not before.
Incompatibilities with Previous Release
Tracing properties will not be ignored or cause the client to hang if the trace directory is set to a non-existent directory.
Rationale for Change
The tracing properties should not be summarily ignored or cause the client to hang if the trace directory does not exist.
Application Changes Required
Applications that counted on the derby.drda.traceAll property being ignored if derby.drda.traceDirectory was set to a non-existent directory, need to turn tracing off or they may now see many errors in the derby.log or large amounts of tracing.
Note for DERBY-3652
Summary of Change
Derby now follows the SQL Standard rules for matching functions and procedures to Java methods.
Symptoms Seen by Applications Affected by Change
In previous releases, Derby matched SQL routines to Java methods using an irregular set of rules which were hard to describe. At the same time, the user documentation falsely claimed that Derby followed the signature matching rules in the ANSI/ISO SQL Standard. Derby now conforms to the Standard behavior. In general, the Standard behavior is easier to understand and it is now possible to be confident that a function or procedure definition will match the desired Java method. In certain corner cases, however, methods which used to resolve will no longer resolve. The Standard resolution rules are described in the Derby Reference Guide.
Incompatibilities with Previous Release
Routines may resolve differently. Here are the differences listed in declining order of likelihood:
Ambiguity - Some SQL routines may match multiple Java methods now. This raises an error according to the Standard. This is most likely to affect the following matches:
Matching INTEGER to Java int and Integer.
Failure - Some SQL routines may fail to match Java methods. This is most likely to affect the following matches:
SMALLINT no longer matches int.
INTEGER no longer matches long.
BIGINT no longer matches float.
REAL no longer matches double.
VARCHAR no longer matches Object.
NUMERIC no longer matches Object.
Change - Some SQL routines may resolve to different Java methods now.
Success - Some SQL routines may now resolve even though previously they did not resolve to Java methods. This is most likely to affect the following matches:
BLOB now matches Blob.
CLOB now matches Clob.
Rationale for Change
The previous behavior violated the SQL Standard and was very hard to explain. The new behavior is correct and easy to describe.
Application Changes Required
This release includes a lint tool, SignatureChecker. If your application uses SQL functions and/or procedures, you should run this tool against your databases in order to find routines which no longer match. To run the tool, make sure that your classpath contains the 10.5 jar files, including derbytools.jar.
On a J2SE platform, run the lint tool as follows (where CONNECTION_URL_TO_DATABASE is the connection URL you would use in order to obtain a connection via DriverManager.getConnection()):
java org.apache.derby.tools.SignatureChecker CONNECTION_URL_TO_DATABASE
Alternatively, you can invoke the tool via derbyrun.jar. E.g.:
java -jar derbyrun.jar SignatureChecker "jdbc:derby:myDB"
On a J2ME platform, run the lint tool as follows (where DATABASE_NAME is the database name you would set via EmbeddedSimpleDataSource.setDatabaseName()):
java org.apache.derby.tools.SignatureChecker DATABASE_NAME
The tool examines every routine registered in the database and prints out results like the following:
Found a matching method for: "APP"."DOINSERT"( )
Found a matching method for: "APP"."DOINSERTANDCOMMIT"( )
Found a matching method for: "APP"."APPENDFOOANDBAR"( VARCHAR )
Unresolvable routine: "APP"."IDONTEXIST"( VARCHAR , INTEGER ).
Detailed reason: No method was found that matched the method call z.iDontExist(java.lang.String, int),
tried all combinations of object and primitive types and any possible
type conversion for any parameters the method call may have.
The method might exist but it is not public and/or static, or the parameter types are not method invocation convertible.
Found a matching method for: "APP"."RUNDDL"( VARCHAR )
Unresolvable routine: "APP"."TABFUNCDOESNTEXIST"( VARCHAR , BIGINT ).
Detailed reason: No method was found that matched the method call
org.apache.derbyTesting.functionTests.tests.lang.TableFunctionTest.appendFooAndBar(java.lang.String, long),
tried all combinations of object and primitive types and any possible
type conversion for any parameters the method call may have.
The method might exist but it is not public and/or static, or the parameter types are not method invocation convertible.
In the example above, the lint tool found matches for all routines except for the functions app.iDontExit and app.tabFuncDoesntExist. If the lint tool cannot find a match for one of your functions or procedures, it tells you what signature it expected to find. You need to adjust your application in one of the following ways:
Method - Change the signature of your Java method to match the signature suggested by the lint tool.
Routine - Drop and recreate your function/procedure so that its arguments and return type match your Java method according to the Standard rules described in the Reference Guide.
Note for DERBY-3420
Summary of Change
The -ca command line option has been removed from ij.
Symptoms Seen by Applications Affected by Change
Applications which attempt to pass JDBC connection attributes using the -ca flag to ij will be rejected, with a usage message such as: Usage: java org.apache.derby.tools.ij -p propertyfile inputfile
Incompatibilities with Previous Release
Applications which attempt to pass JDBC connection attributes using the -ca flag to ij will fail, as that flag is no longer recognized by ij.
Rationale for Change
It used to be that you could not specify both client and embedded attributes in the connection URL. Since Derby now supports such combinations of attributes in the connection URL, the community decided that the -ca option doesn't really add value and so it has been removed.
Application Changes Required
Applications should include connection attributes in the connection URL. For example: connect 'jdbc:derby:myDB;territory=no_NO;collation=TERRITORY_BASED;create=true';
Note for DERBY-3347
Summary of Change
A bug that could cause unrecoverable database corruption has been fixed.
Symptoms Seen by Applications Affected by Change
A bug that could cause database corruption was introduced in the 10.3 codeline and affects the following releases:
Apache Derby 10.3.1.4
Apache Derby 10.3.2.1
Users who are hit by this bug may experience exceptions at various times during execution of SQL statements, booting or shutdown of a database, or during checkpointing. It may result in a number of different error messages, including any of these:
ERROR XSDB3: Container information cannot change once written: was 0, now 80
ERROR XSDG1: Page Page(1039,Container(0, 5856)) could not be written to disk, please check if disk is full.
ERROR XSDG2: Invalid checksum on Page Page(0,Container(0, 1313))
ERROR XSDG3: Meta-data for Container org.apache.derby.impl.store.raw.data.RAFContainer4@1afb0c7 could not be accessed
ERROR XSLA1: Log Record has been sent to the stream, but it cannot be applied to the store (Object null). This may cause recovery problems also.
Incompatibilities with Previous Release
Rationale for Change
Database corruption is bad.
Application Changes Required
No changes are required. However, since the database corruption may go unnoticed for a while, users may want to check the consistency of their databases after upgrading Derby. The process is described on the following wiki page:
. If a corruption is detected, restoring the database from backup is the only reliable way to recover.
Note for DERBY-3327
Summary of Change
The effect of setting the current default schema (SET SCHEMA schemaname) inside nested connection of a stored procedure or function has been changed to comply with SQL standard semantics.
Previously, setting the schema in a nested connection would have an effect on the current default schema also in the SQL connection of the caller. The SQL standard requires that the value of the default current schema at the time of the call be reestablished when the call is completed. This is now implemented.
If the current schema is dropped by a stored procedure or function, the current schema as well as the saved values of the callers will be reset to the initial default schema of the root connection.
Symptoms Seen by Applications Affected by Change
Rationale for Change
SQL standard compliance.
Application Changes Required
Note for DERBY-3319
Summary of Change
Exception is thrown when connection with uncommitted operations is closed.
Symptoms Seen by Applications Affected by Change
In the previous release, applications could close a connection obtained from Derby's implementations of javax.sql.DataSource or javax.sql.ConnectionPoolDataSource, even if the connection had uncommitted operations. Now, Derby raises an SQLException when an attempt to close a connection with an active transaction is made. The exception will have SQLState 25001, and its message will say the following:
java.sql.SQLException: Cannot close a connection while a transaction is still active.
Incompatibilities with Previous Release
Applications that close connections with active transactions now fail.
Rationale for Change
The previous behaviour was not consistent with the behaviour of connection objects obtained from java.sql.DriverManager, which already raise an exception in such a situation. The previous behaviour could also cause resource leaks because there is no way to free the resources held by an active transaction once its connection object has been closed.
Application Changes Required
Users must call commit() or rollback(), or use auto-commit, before attempting to close a connection with uncommitted operations.
Note for DERBY-2351
Summary of Change
An ORDER BY clause of a DISTINCT query which specifies to order by a column which was not in the DISTINCT list is now rejected, because the intent of the query is ambiguous. Previously, Derby instead produced non-distinct results. Also, an ORDER BY clause which specifies a table-name-qualified column alias is now rejected as invalid, where previously it was accepted.
Symptoms Seen by Applications Affected by Change
New rules for DISTINCT and ORDER BY
Applications which specify certain combinations of SELECT DISTINCT with ORDER BY will now receive an error message, whereas formerly such applications received non-distinct results.
As an example, take the following:
create table person (name varchar(10), age int);insert into person values ('John', 10);insert into person values ('John', 30);insert into person values ('Mary', 20);
SELECT DISTINCT name FROM person ORDER BY
The query above is now rejected, with the error message:
If the AGE column is included in the DISTINCT list in the above query, there is no ambiguity
New column alias rules
Applications which specify a column alias for a column in the SELECT statement, and which specify an ORDER BY clause which specifies that column alias qualified by the table name, will now receive an error indicating that the ORDER BY clause is invalid.
As an example, take the following:
create table t1 (i int, j int); select t1.id as idcolumn1, t1.id as idcolumn2 from t1 order by t1.idcolumn1, t1.idcolumn2;
This query is now rejected, as there is no column named 'idcolumn1' in table 't1'. The error message is:
Valid forms of the query above are:
select t1.id as idcolumn1, t1.id as idcolumn2 from t1 order by idcolumn1, idcolumn2;
select t1.id as idcolumn1, t1.id as idcolumn2 from t1 order by t1.id, t1.
Rationale for Change
When the query ambiguously specifies both DISTINCT and ORDER BY, Derby was unsure whether to return the rows properly ordered, but non-distinct, or to return a distinct set of rows, but in an unknown order. Since no clear resolution of the ambiguity could be found, we chose instead to reject the query.
The rules for resolving column references in ORDER BY clauses have been enhanced to consider column aliases and column names more fully. Derby now uses different resolution rules depending on whether the ORDER BY column reference is table.column, or just column:
if the table name is provided, we match against the underlying table name, and don't consider any aliases
if the table name is NOT provided, we first match against the alias name, if present, and if no alias name matches then we match against the underlying source column name.
Application Changes Required
A query which specifies ordering by a non-distinct column should instead include the ORDER BY column in the DISTINCT list, to resolve the ambiguity about which values of that column should be used to distinctly identify the resulting rows.
A query which specifies table-name.alias-name should be rewritten to specify either simply alias-name, or table-name.column-name.
Note for DERBY-2085
Summary of Change
Derby has improved the error message which is issued when an invalid column reference is found in a grouped query. Derby now issues message 42Y36 instead of 42Y30. Furthermore, the wording of message 42Y36 has been expanded to further explain the behavior.
Consider the table
CREATE TABLE t (i INT, j INT)
and consider the following query:
SELECT i FROM t GROUP BY i ORDER BY j
In Derby releases 10.2, 10.3, and 10.4, this query gives the following error message:
ERROR 42Y30: The SELECT list of a grouped query contains at least one invalid expression. If a SELECT list has a GROUP BY, the list may only contain valid grouping expressions and valid aggregate expressions.
This is misleading since there is no invalid expression in the SELECT list. It is the ORDER BY clause that is wrong. In Derby 10.5, this query will now give the following error message, which is a revised version of the message issued by 10.1 (that is, Derby 10.5 behaves more like Derby 10.1 in this respect):
ERROR 42Y36: Column reference 'J' is invalid, or is part of an invalid expression. For a SELECT list with a GROUP BY, the columns and expressions being selected may only contain valid grouping expressions and valid aggregate expressions.
Symptoms Seen by Applications Affected by Change
An application which issued a grouped select statement with an invalid column reference in the order by clause may now receive a different error message than it did in 10.4.
Rationale for Change
We feel that message 42Y36 is more helpful to the user in explaining the problem with the invalid query.
Note for DERBY-1062
Summary of Change
Applications will see a different error code and message when they attempt to call SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE on a table that does not exist.
Symptoms Seen by Applications Affected by Change
Applications will see a different error code and message when they attempt to call SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE on a table that does not exist. The new error code and message will be ERROR 42Y55: 'ALTER TABLE' cannot be performed on 'MissingTableName' because it does not exist. The old error code and message used to be ERROR 42X05: Table/View 'MissingTableName' does not exist.
Incompatibilities with Previous Release
The SQL exception code in the previous release was 42X05 and error message was "Table/View 'MissingTableName' does not exist."
Rationale for Change
To avoid duplication of code, the error handling is now done by a generic routine in ALTER TABLE rather than a routine specific to SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE. This change to use ALTER TABLE code to implement SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE was made in 10.5 release and that is the reason behind the change in error code behavior.
Application Changes Required
If the application is looking for 42X05, it should be changed to look for 42Y55.
Note for DERBY-48
Summary of Change
In Derby, a user's initial default schema is named the same as the user name, or APP if a user is not provided at connect time. This schema is implicitly auto-created the first time a schema object is created in that schema.
Previously, this auto-creation would be performed as part of the user transaction. This would sometimes lead to locking issues as described in this issue. With this change, the auto-creation is now performed and committed immediately in a separate sub-transaction.
Symptoms Seen by Applications Affected by Change
The initial default schema will be present in cases where it previously would not yet have been created: If the user transaction that creates a schema object leading to auto-creation of the initial default schema rolls back for some reason after having created the schema, up till now the auto-creation of the initial default schema would be rolled back as well. Since it is now created and committed in a sub-transaction, the schema creation will not be rolled back: the default schema will persist.
Incompatibilities with Previous Release
Most applications should not be impacted by this change, but there are some corner cases as described below:
If the application tests for the existence of the initial default schema by querying Derby system tables, the results could now be different than in earlier releases, if the test is made after a rollback as described in the previous section.
Since the initial default schema will now potentially exist in cases where it would previously not exist, schema operations may be impacted, e.g. where before a DROP SCHEMA &default schema name& RESTRICT would fail due to it not yet existing, it could now work (if empty), depending on when the drop attempt is made.
Rationale for Change
Implicit schema creation is now performed in its own transaction to avoid deadlocks with other connections accessing the same schema.
Doing this is a separate transaction avoids holding dictionary locks longer than necessary, cf.
and thus reduces the chance for deadlocks.
Application Changes Required
Verify that the application code does not rely on the initial default schema being absent after a rollback.
Build Environment
Derby release 10.5.1.1 was built using the following environment:
Branch - Source code came from the 10.5 branch.
Machine - Microsoft Windows XP Professional Version 2002 Service Pack 2.
Ant - Apache Ant version 1.7.0 compiled on December 13 2006.
JDK 1.4 - Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2) Classic VM (build 1.4.2, J2RE 1.4.2 IBM Windows 32 build cn142- (SR7)
Java 6 - Java(TM) SE Runtime Environment (build pwi81106_07(SR3)) IBM J9 VM (build 2.4, J2RE 1.6.0 IBM J9 2.4 Windows XP x86-32 jvmwi5_25433
Compiler - The 1.6 compiler was used to compile all classes, but java 1.4.2 classes were used to compile appropriate 1.4.2 support
JSR 169 - J2ME support was built using IBM's j9 jvm from WEME6.1
Verifying releases
It is essential that you verify the integrity of the downloaded files using the PGP and MD5 signatures. MD5 verification ensures the file was not corrupted during the download process. PGP verification ensures that the file came from a certain person.
The PGP signatures can be verified using
or . First download the Apache Derby
as well as the asc signature file for the particular distribution. It is important that you get these files from the ultimate trusted source - the main ASF distribution site, rather than from a mirror. Then verify the signatures using ...
% pgpk -a KEYS
% pgpv db-derby-X.Y.tar.gz.asc
% pgp -ka KEYS
% pgp db-derby-X.Y.tar.gz.asc
% gpg --import KEYS
% gpg --verify db-derby-X.Y.tar.gz.asc
To verify the MD5 signature on the files, you need to use a program called md5 or md5sum, which is included in many unix distributions. It is also available as part of . Windows users can get binary md5 programs from , , or .
We strongly recommend you verify your downloads with both PGP and MD5.
Send feedback about the website to: