1<!--$Id: 4.4.16.html,v 1.24 2007/05/17 18:17:18 bostic Exp $--> 2<html> 3<head> 4<title>The Berkeley DB Package: DB 4.4.16 Change Log</title> 5<meta name="description" content="Berkeley DB: A database programmatic toolkit."> 6<meta name="keywords" content="embedded,database,programmatic,toolkit,b+tree,btree,hash,hashing,transaction,transactions,locking,logging,access method,access methods"> 7</head> 8<body bgcolor=white> 9 10<h3 align=center>Berkeley DB 4.4.16 Change Log</h3> 11 12<h3>Database or Log File On-Disk Format Changes:</h3> 13<ol> 14 15<li>The on-disk log format has changed. 16 17</ol> 18<h3>New Features:</h3> 19<ol> 20 21<li>Add support to compact an existing Btree database. [#6750] 22<li>Add support for named in-memory databases. [#9927] 23<li>Add support for database environment recovery serialization. This 24simplifies multiprocess application architectures. Add DB_REGISTER flag 25to DB_ENV->open(). [#11511] 26<li>Add utility for performing hot backups of a database environment. 27[#11536] 28<li>Add replication configuration API. [#12110] 29<li>Add replication support to return error instead of waiting for 30client sync to complete. [#12110] 31<li>Add replication support for delayed client synchronization. [#12110] 32<li>Add replication support for client-to-client synchronization. 33[#12110] 34<li>Add replication support for bulk transfer. [#12110] 35<li>Add new flags DB_DSYNC_DB and DB_DSYNC_LOG [12941] 36<li>Add DbEnv.log_printf, a new DbEnv method which logs printf style 37formatted strings into the Berkeley DB database environment log. 38[#13241] 39 40</ol> 41<h3>Database Environment Changes:</h3> 42<ol> 43 44<li>Add a feature to support arbitrary alignment of mutexes in order to 45minimize cache line collisions. [#9580] 46<li>Change cache regions on 64-bit machines to allow regions larger than 474GB. [#10668] 48<li>Fix a bug where a loop could occur if the application or system 49failed during modification of the linked list of shared regions. 50[#11532] 51<li>Fix mutex alignment on Linux/PA-RISC, add test-and-set mutexes for 52MIPS and x86_64. [#11575] 53<li>Fix a bug where private database environments (DB_PRIVATE) on 64-bit 54machines would core dump because of 64-bit address truncation. [#11983] 55<li>Fix a bug where freed memory is accessed when DB_PRIVATE 56environments are closed. This can happen on systems where the operating 57system holds mutex resources that must be freed when the mutex is 58destroyed. [#12591] 59<li>Fix a bug where the DbEnv.stat_print method could self-deadlock and 60hang. The DbEnv.stat_print method no longer displays statistics for any 61of the database environments databases. [#12039] 62<li>Fix a bug where Berkeley DB could create fragmented 63filesystem-backed shared region files. [#12125] 64<li>Fix a bug where Berkeley DB stat calls could report a cache size of 650 after the statistics were cleared. [#12307] 66<li>Threads of control joining database environments are now configured 67for all of the subsystems (lock, log, cache, or transaction) for which 68the environment was originally configured, it is now an error to attempt 69configuration of additional subsystems after an environment is created. 70[#12422] 71<li>Fix a bug where negative percentages could be displayed in 72statistics output. [#12673] 73<li>Fix a bug that could cause a panic if the cache is filled with 74non-logging updated pages. [#12763] 75<li>Fix a bug that could cause an unreported deadlock if the application 76was using the DB_DIRTY_READ flag and the record was an off page 77duplicate record. [#12893] 78<li>Fix a bug where a handle lock could be incorrectly retained during 79a delete or rename operation. [#12906] 80 81</ol> 82<h3>Concurrent Data Store Changes:</h3> 83<ol> 84 85<li>Lock upgrades and downgrades are now accounted for separately from 86lock requests and releases. [#11155] 87<li>Fix a bug where a second process joining a Concurrent Data Store 88environment, with the DB_CDB_ALLDB flag set, would fail. This would 89happen if the first thread were not entirely finished with 90initialization. [#12277] 91 92</ol> 93<h3>General Access Method Changes:</h3> 94<ol> 95 96<li>Fix a bug where filesystem operations are improperly synchronized. 97[#10564] 98<li>Add support for database files larger than 2GB on Windows. [#11839] 99<li>Rename DB_DEGREE_2 (and all related flags) to DB_READ_COMMITTED. 100Rename DB_DIRTY (and all related flags) to DB_READ_UNCOMMITTED. [#11776] 101<li>Fix a bug where wrapping of sequences was incorrect when the cache 102size is smaller than the range of the maximum value minus the minimum 103value. [#11997] 104<li>Fix a bug that could result in a hot backup having a page missing 105from a database file if a file truncation was in progress during the 106backup but was then aborted. [#12017] 107<li>Fix a bug where a long filename could cause one too few bytes to 108be allocated when opening a file. [#12085] 109<li>Fix a bug in secondary cursor code if a write lock is not granted. 110[#12109] 111<li>Fix a bug in secondary cursors where the current record would change 112on error. [#12141] 113<li>Fix a bug in Db->truncate where the method was not checking to see 114if the handle was opened read-only. [#12179] 115<li>Fix a bug in sequences so that they are now platform independent, 116taking into account little-endian and big-endian architectures. They 117will be automatically upgraded in 4.4. [#12202] 118<li>Fix a bug with non-wrapping sequences when initial value was 119INT64_MIN. [#12390] 120<li>Add a retry for operating system operations that return EIO (IO 121Error) to better support NFS mounted filesystems. [#12426] 122<li>Fix sequence wrapping at INT64 limits. [#12520] 123<li>Fix a bug where errors during DB->associate could leave secondaries 124half associated. [#13173] 125<li>Fix a bug so that we no longer will update in CDS and DS if the file 126size limit will be exceeded. [#13222] 127 128</ol> 129<h3>Btree Access Method Changes:</h3> 130<ol> 131 132<li>Remove maxkey configuration. [#8904] 133<li>Fix a memory leak in operations on large Btrees. [#12000] 134 135</ol> 136<h3>Hash Access Method Changes:</h3> 137<ol> 138 139<li>Fix a bug where access to HASH or encrypted database pages might be 140blocked during a checkpoint. [#11031] 141<li>Fix a bug where recovery would fail when a database has a hash page 142on the free list and that hash page was freed without using transactions 143and later allocated and aborted within a transaction. [#11214] 144<li>Fix a bug in hash duplicates where if the caller left garbage in the 145partial length field, we were using it. Fix a bug where a replacement 146of a hash item that should have gone on an overflow page, did not. 147[#11966] 148<li>Fix a bug where free space was miscalculated when adding the first 149duplicate to an existing item and the existing item plus the new item 150does not fit on a page. [#12270] 151<li>Fix a bug where allocations of hash buckets are not recovered 152correctly. [#12846] 153 154</ol> 155<h3>Queue Access Method Changes:</h3> 156<ol> 157 158<li>Improve performance of deletes from a QUEUE database that does not 159have a secondary index. [#11538] 160<li>Fix a bug where updates that do not use transactions, but do enable 161locking, failed to release locks. [#11669] 162<li>Fix a bug where a transaction might not be rolled forward if the 163site was performing hot backups and an application aborted a prepared 164but not committed transaction. [#12181] 165<li>Fix a bug with queue extents not being reclaimed. [#12249] 166<li>Fix a bug where a record being inserted before the head of the queue 167could appear missing if DB_CONSUME is not specified. [#12919] 168<li>Fix a bug that might cause recovery to move the head or tail of the 169queue to exclude a record that was deleted but whose transaction did not 170commit. [#13256] 171<li>Fix a bug that could cause recovery to move the head or tail pointer 172beyond a record that was aborted but was rolled backward by recovery. 173[#13318] 174 175</ol> 176<h3>Recno Access Method Changes</h3> 177 178None. 179 180<h3>C++-specific API Changes:</h3> 181<ol> 182 183<li>Fix a bug so that a DbMemoryException will be raised during a 184DB_BUFFER_SMALL error. [#13273] 185 186</ol> 187<h3>Java-specific API Changes:</h3> 188<ol> 189 190<li>Add VersionMismatchException to map the DB_VERSION_MISMATCH error. 191[#11429] 192<li>Fix a bug in Environment.getConfiguration() method in non-crypto 193builds. [#11752] 194<li>Fix a bug that caused a NullPointerException when using the 195MultipleDataEntry default constructor. [#11753] 196<li>Fix handling of replication errors. [#11822] 197<li>Remove EnvironmentConfig.setReadOnly() method. [#11882] 198<li>Fix a bug where prefix strings in the error handler may be 199corrupted. [#11967] 200<li>Fix a bug so that nested exceptions will appear in stack traces. 201[#11992] 202<li>Fix a bug on LogSequenceNumber objects in the Java API. [#12223] 203<li>Fix a bug when no files are returned from a call to 204DB_ENV->log_archive. [#12383] 205<li>Fix a bug when multiple verbose flags are set. [#12383] 206<li>Fix a bug so that an OutOfMemoryError is thrown when allocation 207fails in the JNI layer. [#13434] 208 209</ol> 210<h3>Java collections and bind API Changes:</h3> 211<ol> 212 213<li>Binding performance has been improved by using System.arraycopy in 214the FastOutputStream and FastInputStream utility classes. [#12002] 215<li>The objectToEntry method is now implemented in all TupleBinding 216subclasses (IntegerBinding, etc) so that tuple bindings are fully 217nestable. An example of this usage is a custom binding that dynamically 218discovers the data types of each of the properties of a Java bean class. 219For each property, it calls TupleBinding.getPrimitiveBinding using the 220property's type (class). When the custom binding's objectToEntry method 221is called, it in turn calls the objectToEntry method of the nested 222bindings for each property. [#12124] 223<li>The getCause method for IOExceptionWrapper and 224RuntimeExceptionWrapper is now defined so that nested exceptions appear 225in stack traces for exceptions thrown by the collections API. [#11992] 226<li>TupleBinding.getPrimitiveBinding can now be passed a primitive type 227class as well as a primitive wrapper class. The return value for 228Integer.TYPE and Integer.class, for example, will be the same binding. 229[#12035] 230<li>Improvements have been made to prevent the buffer used in serial and 231tuple bindings from growing inefficiently, and to provide more 232alternatives for the application to specify the desired size. For 233details see com.sleepycat.bind.serial.SerialBase and 234com.sleepycat.bind.tuple.TupleBase. [#12398] 235<li>Add StoredContainer.getCursorConfig, deprecate isDirtyRead. 236Deprecate StoredCollections.dirtyReadMap (dirtyReadSet, etc) which is 237replaced by configuredMap (configuredSet, etc). Deprecated 238StoredContainer.isDirtyReadAllowed with no replacement (please use 239DatabaseConfig.getDirtyRead). Also note that 240StoredCollections.configuredMap (configuredSet, etc) can be used to 241configure read committed and write lock containers, as well as read 242uncommitted containers, since all CursorConfig properties are supported. 243[#11776] 244<li>Add the protected method SerialBinding.getClassLoader so that 245subclasses may return a specific or dynamically determined class loader. 246Useful for applications which use multiple class loaders, including 247applications that serialize Groovy-defined classes. [#12764] [#12749] 248 249</ol> 250<h3>Tcl-specific API Changes:</h3> 251<ol> 252 253<li>Fix a bug that could cause a memory leak in the replication test 254code. [#13436] 255 256</ol> 257<h3>RPC-specific Client/Server Changes:</h3> 258<ol> 259 260<li>Fix double-free in RPC server when handling an out-of-memory error. 261[#11852] 262 263</ol> 264<h3>Replication Changes:</h3> 265<ol> 266 267<li>Fix race condition (introduced in 4.3) in rep_start function. 268[#11030] 269<li>Changed internal initialization to no longer store records. [#11090] 270<li>Add support for replication bulk transfer. [#11099] 271<li>Berkeley DB now calls check_doreq function for MASTER_REQ messages. 272[#11207] 273<li>Fix a bug where transactions could be counted incorrectly during 274txn_recover. [#11257] 275<li>Add DB_REP_IGNORE flag so that old messages (especially PERM 276messages) can be ignored by applications. [#11585] 277<li>Fix a bug where op_timestamp was not initialized. [#11795] 278<li>Fix a bug in db_refresh where a client would write a log record on 279closing a file. [#11892] 280<li>Fix backward arguments in C++ rep_elect API. [#11906] 281<li>Fix a bug where a race condition could happen between downgrading a 282master and a database update operation. [#11955] 283<li>Fix a bug on VERIFY_REQ. We now honor wait recs/rcvd. [#12097] 284<li>Fix a bug in rebroadcast of verify_req by initializing lp->wait_recs 285when finding a new master. [#12097] 286<li>Fix a bug by adding lockout checking to __env_rep_enter since 287rename/remove now call it. [#12192] 288<li>Fix a bug so that we now skip __db_chk_meta if we are a rep client. 289[#12316] 290<li>Fix a replication failure on Windows. [#12331] 291<li>Remove master discovery phase from rep_elect as a performance 292improvement to speed up elections. [#12551] 293<li>Fix a bug to avoid multiple data streams when issuing al ALL_REQ. 294[#12595] 295<li>Fix a bug to request the remaining gap again if the gap record is 296dropped after we receive the singleton. [#12974] 297<li>Fix a bug in internal initialization when master changes in the 298middle of initializing. [#13074] 299<li>Fix a bug in replication/archiving with internal init. [#13110] 300<li>Fix pp handling of db_truncate. [#13115] 301<li>Fix a bug where rep_timestamp could be updated when it should not 302be updated. [#13331] 303<li>Fix a bug with bulk transfer when toggling during updates. [#13339] 304<li>Change EINVAL error return to DB_REP_JOIN_FAILURE. [#12110] 305<li>Add C++ exception for DB_REP_HANDLE_DEAD. [#13361] 306<li>Fix a bug where starting an election concurrently with processing a 307NEWMASTER message could cause the send function to be called with an 308invalid eid. [#13403] 309 310</ol> 311<h3>XA Resource Manager Changes:</h3> 312 313None. 314 315<h3>Locking Subsystem Changes:</h3> 316 317None. 318 319<h3>Logging Subsystem Changes:</h3> 320<ol> 321 322<li>Add set_log_filemode for applications that need to set an absolute 323file mode on log files. [#8747] 324<li>Fix a bug that caused Not Found to be returned if a log file exists 325but is not readable. [#11185] 326<li>Removed checksum of records with an in-memory log buffer. [#11280] 327<li>Fix a bug so that the DB_LOG_INMEMORY flag can no longer be set 328after calling DB_ENV->open. [#11436] 329<li>Fix a bug introduced after release 4.0 where two simultaneous 330checkpoints could cause ckp_lsn values to be out of order. [#12094] 331<li>Fix a bug when in debug mode and using the DEBUG_ROP which will now 332log read operations in __dbc_logging. [#12303] 333<li>Fix a bug where failing to write a log record on a file close would 334result in a core dump later. [#12460] 335<li>Fix a bug where automatic log file removal, or the return of log 336files using an absolute path, could fail silently if the applications 337current working directory could not be reached using the systems getcwd 338library call. [#12505] 339<li>Avoid locking the log region if we are not going to flush the log. 340This can improve performance for some write-intensive application 341workloads. [#13090] 342<li>Fix a bug with a possible segment fault when memp_stat_print is 343called on a temporary database. [#13315] 344<li>Fix a bug where log_stat_print could deadlock with threads during a 345checkpoint. [#13315] 346 347</ol> 348<h3>Memory Pool Subsystem Changes:</h3> 349<ol> 350 351<li>Fix a bug where modified database pages might not be flushed if 352recovery were run and all pages from a database were found in the system 353cache and up to date, followed by a system crash. [#11654] 354 355</ol> 356<h3>Transaction Subsystem Changes:</h3> 357<ol> 358 359<li>Add new DbTxn class methods allowing applications to set/get a 360descriptive name associated with a transaction. The descriptive name 361is also displayed by the db_stat utility. [#0382] 362<li>Fix a bug where aborting a transaction with a large number of nested 363transactions could take a long time. [#10972] 364<li>Add support to allow the TXN_WRITE_NOSYNC flag to be specified on 365the transaction handle. [#11151] 366<li>Fix a bug that could cause a page to be on the free list twice if 367it was originally put on the free list by a non-transactional update and 368then reallocated in a transaction that aborts. [#11159] 369<li>Remove the requirement for the DB_AUTO_COMMIT flag to make database 370operations transactional. Specifying the database environment as 371transactional or opening the database handle transactionally is 372sufficient. [#11302] 373<li>Fix a bug so that environments created from errant programs that 374called dbp->close while transactions were still active can now be 375recovered. [#11384] 376<li>Fix a bug that caused free pages at the end of a file to be 377truncated during recovery rather than placed on the free page list. 378[#11643] 379<li>Fix a bug that caused a page to have the wrong type if the truncate 380of a BREE or RECNO database needed to be rolled forward. [#11670] 381<li>Fix a bug when manually undoing a subdb create, dont try to free a 382root page that has not been allocated. [#11925] 383<li>Add a check on database open to see if log files were incorrectly 384removed by system administration mistakes. [#12178] 385<li>Fix a bug when calling DB->pget and then specifying the 386DB_READ_COMMITTED (DB_DEGREE_2) on a cursor. If followed by a 387DBC->c_pget, the primary database would incorrectly remain locked. 388[#12410] 389<li>Fix a bug where the abort of a transaction in which a sub database 390was opened with the DB_TXN_NOT_DURABLE flag could fail. [#12420] 391<li>Fix a bug that could cause an abort transaction that allocated new 392pages to a file that were not flushed to disk prior to the abort 393transaction to report out of disk space. [#12743] 394<li>Fix a bug that could prevent multiple creates and destroys of the 395same file to be recovered correctly. [#13026] 396<li>Fix a bug when recovery previously handled a section of the log that 397did not contain any transactions. [#13139] 398<li>Fix a bug that could result in the loss of durability in 399Transactional Environments on Mac OS X. [#13149] 400<li>Fix a bug that could cause the improper reuse of a transaction id 401when recovery restores prepared transactions. [#13256] 402 403</ol> 404<h3>Utility Changes:</h3> 405<ol> 406 407<li>Add utility for performing hot backups of a database environment. 408[#11536] 409<li>Change the Verify utility to now identify any nodes that have 410incorrect record counts. [#11934] 411<li>Fix a bug in the 1.85 compatibility code supporting per-application 412Btree comparison and prefix compression functions. The functions would 413not work on big-endian 64-bit hardware. [#13316] 414 415</ol> 416<h3>Configuration, Documentation, Portability and Build Changes:</h3> 417<ol> 418 419<li>Change the ex_tpcb sample application to no longer displays 420intermediate results. It displays results at the end of the run. 421[#11259] 422<li>Change the Visual Studio projects on Windows so that each is in an 423intermediate directory. [#11441] 424<li>Fix errors in test subdb011. [#11799] 425<li>Fix a bug that could cause applications using gcc on Power PC 426platforms to hang. [#12233] 427<li>Fix a bug where installation will fail if a true program cannot be 428found. [#12278] 429<li>Fix a bug that prevented C++ applications from configuring XA 430[#12300]. 431<li>Fix a race condition in the Windows mutex implementation found on 4328-way Itanium systems. [#12417] 433<li>Add pthread mutex support for IBM OS/390 platform (z/OS or MVS). 434[#12639] 435<li>Fix a bug where the Tcl API did not configure on OS X 10.4. [#12699] 436<li>Fix portability issues with queue or recno primary databases. 437[#12872] 438<li>Fix a bug where utility attempted to send replication message. 439[#13446] 440 441</ol> 442 443</body> 444</html> 445