---------------------------------------------------------------------- ELOQUENCE B.07.10 - patch PE71-0906050 ---------------------------------------------------------------------- This patch adds enhancements or fixes defects of the eloqdb6 server program as released with Eloquence B.07.10. This patch will be integrated in the Eloquence B.07.10 release. Eloquence B.07.10 must be installed before applying this patch. Severity: PE71-0906050: BUG FIX Superseded patches: PE71-0905130: BUG FIX PE71-0904270: BUG FIX PE71-0904240: BUG FIX PE71-0904060: BUG FIX PE71-0903030: BUG FIX PE71-0811250: BUG FIX PE71-0809150: BUG FIX, ENHANCEMENT PE71-0807210: ENHANCEMENT PE71-0807010: BUG FIX PE71-0805070: BUG FIX PE71-0803070: BUG FIX, ENHANCEMENT PE71-0802120: BUG FIX, ENHANCEMENT PE71-0801240: ENHANCEMENT, BUG FIX PE71-0709050: BUG FIX PE71-0708130: BUG FIX PE71-0708030: BUG FIX PE71-0707310: ENHANCEMENT, BUG FIX PE71-0707200: ENHANCEMENT, BUG FIX PE71-0705300: ENHANCEMENT PE71-0704110: BUG FIX, ENHANCEMENT PE71-0702080: BUG FIX PE71-0610130: BUG FIX PE71-0609270: BUG FIX PE71-0609120: BUG FIX PE71-0608110: ENHANCEMENT PE71-0607210: BUG FIX PE71-0607100: ENHANCEMENT PE71-0606290: BUG FIX PE71-0603280: BUG FIX PE71-0603100: ENHANCEMENT PE71-0602280: ENHANCEMENT, BUG FIX For additional information on the replication functionality please refer to the Eloquence web site: http://eloquence.marxmeier.com/support/B0710/doc/repl/ Patch PE71-0906050 ------------------ Platforms: All * Fixed a potential memory leak when a dbctl list output exceeded the size of the shared memory communications buffer (#3785). * Changed the Windows file version to "7.1.1.36" Notes / Related patches: - The following related patches (or superseding) should be installed with this patch: - PE71-0801241 - dblogreset utility - PE71-0809151 - dbrecover utility - PE71-0807030 - fwaudit utility - PE71-0905120 - dbfsck utility - To use replication functions, the following related patches (or superseding) need to be installed in addition: - PE71-0903300 - dbrepl - PE71-0705304 - chklic - PE71-0705305 - dbvolextend - PE71-0705306 - dbvolchange - PE71-0804170 - dbcfix Patch PE71-0905130 ------------------ Platforms: All * Fixed a problem which could cause the database server to temporarily hang while on-line backup mode was shutting down (#3782). If a checkpoint operation overlapped with stopping on-line backup mode, the checkpoint could block and in turn block any existing or new database sessions until on-line backup mode was shut down. * Changed the Windows file version to "7.1.1.35" Patch PE71-0904270 ------------------ Platforms: All * This patch corrects a problem introduced with fix #3777 delivered with the previous patch PE71-0904240. * Changed the Windows file version to "7.1.1.34" Patch PE71-0904240 ------------------ Platforms: All * The dbctl forwardlog restart command did not immediately update the volume generation count in the log volume (#3777). As a possible consequence, the database server could fail to start with a message like below if it was abnormally terminated after a dbctl forwardlog restart was issued: failed to open volume: volume ... has inconsistent generation count As a workaround it is possible to manually update the generation count in the log volume in this case. Please contact support if you are affected by this. This problem was introduced in patch PE71-0811250. * Changed the Windows file version to "7.1.1.33" Patch PE71-0904060 ------------------ Platforms: All * Fixed a problem where a database was not immediately available after a dbutil restructuring (#3751). A subsequent DBOPEN executed immediately after a restructuring dbutil process had finished could in some cases encounter a database status -2 (database in use). * Resuming a replication or recovery could in rare cases abort with a message like below (#3752): Assertion failed: Fwr_PageHashAdd() failed: key already present server panic: Aborting on internal failure, file volfwr.c, line 5367 This could happen in a corner case if a user transaction spans multiple forward-log files, where BTREE PAGE actions are present in different files at identical offsets. Under these conditions the problem could be triggered if these files had to be processed to locate the last status of the replication progress. * Resuming a replication or recovery could in rare cases abort with a message like below (#3755): recovery failed ... node ... page ... not in expected state Assertion failed: h->lower == data->lower server panic: Aborting on internal failure, file btree.c, line 2382 This could happen in a corner case if a user transaction spans multiple forward-log files, where the transaction was executed immediately after the last status of the replication progress was found and the replication was resumed. * In a rare corner case, a dbutil RENAME DATABASE operation could cause the database server to abort with a message like below (#3704): Assertion failed: SysCat_LinkAddrEquals ... server panic: Aborting on internal failure, file syscat.c, line 1922 * Changed the Windows file version to "7.1.1.32" Patch PE71-0903030 ------------------ Platforms: All * In rare cases, a malformed client request or an incomplete tcp/ip network packet could cause the eloqdb6 to abort with a message like below: Assertion failed: size != 0 Aborting on internal failure, file buffer.c, line 777 * If forward-logging was manually disabled and later enabled again it could happen that btree index actions were missing from the forward-log (#3694). If a user transaction was started while forward-logging was disabled but this transaction was then committed after enabling forward-logging, those btree index actions that were executed while forward-logging was disabled could be missing from the forward-log. As a consequence, a recovery or replication from this forward-log would fail, for example with a message like below: recovery failed in action ... page ... not in expected state server panic: Fatal problem detected in btree_FWR__verify_page Assertion failed: h->lower == data->lower * The "dbctl forwardlog status" command does no longer require dba privileges. Platforms: Windows * In the HTTP status display, a default VolumeFileSizeLimit = 0 is now correctly displayed as "unlimited (128 GB)". * Changed the Windows file version to "7.1.1.31" Patch PE71-0811250 ------------------ Platforms: All * A DBGET mode 6 or 16 on an index item could improperly return an end-of-chain condition when the last key in the index was deleted (#3683). * The database server could refuse to start due to an inconsistent volume generation count after a startup recovery was aborted while recovering from on-line backup mode (#3676). * Executing "dbctl backup start" or "dbctl forwardlog restart" could result in a corrupted transaction journal if the execution of the dbctl command encountered a server panic for any reason (#3677). * A missing newline in the "dbctl replication stop" output was fixed (#3674). * Changed the Windows file version to "7.1.1.30" Patch PE71-0809150 ------------------ Platforms: All * In rare cases the eloqdb6 could abort with a message like below during a DBGET operation on an index item (#3540): Assertion failed: session->node_cnt == 0 Aborting on internal failure, file runutil.c, line 228 If a cache-miss is encountered during a DBDELETE commit phase a concurrently running DBGET could detect an inconsistency between indexes and data. The DBGET would then internally repeat the index access, which could cause the eloqdb6 to abort because the data set reference was not correctly reset before repeating the access. This problem was introduced with patch PE71-0803070. * The [Replication] IgnoreWrite configuration item was added (#3604). If set, opening a database in write mode on a replication slave server is granted but internally converted into a read-only open mode. This way, a program that opens a database in write mode but then only performs read operations may run on a replication slave server. * Fixed a problem where in rare cases the eloqdb6 could abort with a message like below when stopping on-line backup mode (#3637): Assertion failed: !extend_list Aborting on internal failure, file volredir.c, line 609 Restarting the eloqdb6 should resolve this problem. The eloqdb6 should then recover from on-line backup mode and resume normal operation. * Fixed a problem where in rare cases a forward recovery (dbrecover) or database replication could fail due to a wrong order of actions in the forward-log (#3634). When erasing or purging a database or restructuring a database with dbutil a checkpoint operation could result in a wrong order of operations in the forward-log. This could result in data corruption during forward recovery or replication. If this problem is encountered on a replicated slave server the slave server must be rebuilt from the master server volume files. When encountered during dbrecover, recovery must be restarted from the last backup after installing a corrected dbrecover binary (restarting dbrecover will not work). * Fixed a problem where a replication slave server could fail to resume replication after a crash of the master server (#3642). The replication could fail with a message like below returned by dbrepl: ERROR: replication failed, see slave server log for details dbrepl: Server replication call failed: REPL 1 [-810:1] A message like below was written to the slave server log: replication header sequence mismatch (...) The problem could occur if the dbrepl process was stopped after the master server crashed. The slave server could then fail to resume replication after the master was restarted because an internal state was not correctly maintained in this situation. The problem can be resolved by installing a corrected eloqdb6 binary on the slave server. The slave server should then correctly resume replication. * A message like below could be logged on a repliction slave server while a dbutil database restructuring is replicated (#3645): idb__repl_callback: entry not found (...) There is no problem involved with this message. * Changed the Windows file version to "7.1.1.29" Patch PE71-0807210 ------------------ Platforms: All * The dbrestore operation was enhanced to no longer block concurrent tasks while data is written to the volume files. * The dbrestore /nice option was enhanced to periodically synchronize the forward-log to disk after some data has been written. In case of high disk i/o load, this should reduce contention on the operating system buffer cache and performance of concurrent tasks. * Changed the Windows file version to "7.1.1.28" Patch PE71-0807010 ------------------ Platforms: All * A potential problem was fixed that could in rare cases result in an internal deadlock condition under high load when concurrently updating a btree index (#3599). When splitting or deleting a page in a btree index a lock to a related page was not obtained correctly and could cause a deadlock with a commit operation of a concurrent session. Platforms: Linux * A potential corruption of an internal memory area was fixed which could happen on multi-CPU Linux installations while starting the eloqdb6 server (#3075). This could cause the eloqdb6 to hang or crash during startup. Also, in some cases, if the eloqdb6 started correctly it could later abort with a panic message indicating a data corruption, for example: Assertion failed: Pool_TestFlag(Pool_BLOCKUSED, &header) Assertion failed: Node_GetItemPageReadOnly() failed: file is corrupt Assertion failed: flag_byte == FixRec_DELETED However, real data corruption did not occur. Restarting the eloqdb6 resolved the problem. * A similar corruption with similar symptoms and consequences could happen when running the eloqdb6 server with a D2 or D3 log flag, for example LogFlags=*2 or LogFlags=*1D2 (#3075). Platforms: Windows * Changed the Windows file version to "7.1.1.27" Patch PE71-0805070 ------------------ * Fixed a problem affecting a replicated slave server or recovery with dbrecover which in rare cases could cause a panic with a message like below (#3568): Assertion failed: h->lower == data->lower Aborting on internal failure, file btree.c, line 1866 The line number may differ. Besides the "lower" element, the failed assertion may also apply to the "prevpg" or "nextpg" or "upper" or "flags" elements. This could happen when replication was stopped and later resumed or when performing an incremental recovery with dbrecover. If replication or recovery is restarted it needs to continue at the exact point it left off previously. The last checkpoint is recorded in the volume file. However, any changes beyond the last checkpoint need to be verified if they were previously applied. If similar actions affecting specific btree changes are found the replication or recovery could fail to correctly locate the point-of-resume in the forward log. This could happen, for example, on multiple DBDELETE / DBPUT sequences affecting an index in a way that the same btree page was modified identically multiple times. The implementation was changed to maintain additional information in the volume files on the replicated slave server or during recovery to correctly identify the last change applied. If this problem is encountered on a replicated slave server the slave server must be rebuilt from the master server volume files. When encountered during dbrecover, recovery must be restarted from the last backup after installing a corrected dbrecover binary (restarting dbrecover will not work). * Fixed a potential page leak in the log volume when a replication was stopped and later resumed. * In some cases a replication master server did not correctly update the recovery status in the root volume file. This could have the effect that a subsequent dbrecover would then assume a previously incomplete recovery and unnecessarily require a previous forward log file to be present. For example, an on-line backup is performed. This starts a new forward log generation. If these volume files are later used in a recovery, dbrecover applies forward log files from the generation started with the backup. However, as the recovery status was not correctly updated, it would instead expect the previous generation (and just skip it). This could only happen if the eloqdb6 is configured as replication master server, not in standalone mode (default). * In rare cases DBFIND or DBGET on index items could return the status -804 (#3590). An internal index cursor invalidation could happen if an index traversal was interrupted by a concurrent commit or rollback on the same index. Under rare conditions this could result in an unexpected index cursor state and status -804 was returned to the application. * Fixed a problem with DBGET mode 5 or 6 on index items using packed decimal (P) or zoned decimal (Z) item types (#3574). This could have the effect that an improper end-of-chain condition could be returned to the application in some cases. With P or Z items, identical values may have a different binary representation. For example, the values 42 (unsigned) and +42 (positive) have the same numeric value but differ in their binary representation. Eloquence compares these items by their value, regardless of the binary representation. For example, when a P or Z item is used as search item, the values 42 (unsigned) and +42 (positive) use the same chain, they are considered identical. However, using DBGET mode 5 or 6 with a P or Z index item behaved differently. Although the underlying index correctly locates the entry by its value, a binary comparison was used on the result, causing an improper end-of-chain condition if a key value did not match the search argument in the binary comparison. * Changed the Windows file version to "7.1.1.26" Patch PE71-0803070 ------------------ * Fixed a problem affecting a replicated slave server or recovery with dbrecover which in rare cases could cause a panic with a message like below (#3552): Assertion failed: h->pgno == data->pgno Aborting on internal failure, file btree.c, line 1808 This problem was caused by a defect in the btree recovery code that could result in a corrupted index page if a btree root page was split for the first time and the page was previously not in the buffer cache. If this problem is encountered on a replicated slave server the slave server must be rebuilt from the master server volume files. When encountered during dbrecover, recovery must be restarted from the last backup after installing a corrected dbrecover binary (restarting dbrecover will not work). * The builtin dbstore function (dbctl dbstore) was enhanced to create the store archive using more restrictive permissions (#3544). The builtin dbstore function now restricts access to the account running the db server process. * Fixed a problem which could result in unexpected "protocol failure" and -700:-6 error messages in dbrepl. In some cases a replication slave server process prematurely closed the connection of the dbrepl utility when encountering a problem. This could have the effect that the actual error message was lost and a generic error message was output. * In rare cases a DBGET could return the status -803:258 (#3540). A potential race condition was identified that could result in DBGET returning status -803:258 in rare cases. If a cache-miss is encountered during a DBDELETE commit phase a concurrently running DBGET could detect an inconsistency between indexes and data and return an unexpected status code. * On a replication slave the forward-log file permissions were not correctly set and the [forwardlog] GroupReadAccess configuration option had no effect (#3548). * Changed the Windows file version to "7.1.1.25" Patch PE71-0802120 ------------------ * A problem in DBDELETE was fixed that could in some cases result in a noticeable delay (#3524). This delay could affect concurrent database sessions during write operations (DBPUT/DBUPDATE/DBDELETE) on the same data set. During DBDELETE, if the first or last record is deleted in a data set, the data set meta information was updated to reflect the new first or last record in the set. However, if there is a significant number of deleted records in the set that must be traversed to locate the new first or last record, this might take some time, subject to caching. The implementation was modified so that the first/last record meta information, as well as updating this information during DBDELETE, is no longer needed. This was achieved in a way that the data remains backwards-compatible with previous eloqdb6 versions. Please note: If a previous dbfsck utility is used it may report first/last record meta data inconsistencies. Please consider installing patch PE71-0802121 which provides a new dbfsck version where the first/last record check is relaxed, according to the way the new database server implementation handles the first/last record meta information. * Fixed a problem on creating a case insensitive index (#1073). A case insensitive index being created using the previous patch PE71-0801240 may have inconsistencies so that not all key values can be located in the index. As this correction affects new functionality introduced with the previous patch PE71-0801240, no existing database should be affected by this problem. However, if a case insensitive index was created with the previous version it should be re-created (deleted and then added again) using the dbutil program. * The STOP argument was added to the dbctl replication command on a replication slave (#3523): dbctl -u dba replication stop This disconnects the dbrepl process from the slave server. * The dbctl replication status output was modified (#3523). On a replication slave, it is now displayed in addition whether the replication is active (i.e., the dbrepl process is connected) or not. Please note: If you use external utilities that process the dbctl replication status output it may be necessary to adapt them to the new output format. * Fixed a problem where a configured StatFile was truncated during database server startup although StatFileFlags included the "a" (append) flag (#3536). * Changed the Windows file version to "7.1.1.24" Patch PE71-0801240 ------------------ Platforms: All * The database server was enhanced to support case insensitive indexes (#1073). When specified for an index, key comparison on strings is performed in a case insensitive manner. * Fixed a potential TurboIMAGE compatibility problem (#3502). A TurboIMAGE DBGET call that fails with a status code may return the current record number in the status array. The database server was modified to return the current record number in case a DBGET call fails with a status code. * Fixed a problem during incremental recovery or replication (#3515). An incremental dbrecover or replication could in rare cases fail with a message like below: Assertion failed: offset == data->offset Aborting on internal failure, file btree.c, line 2342 This was caused by an already modified btree page not being skipped during the sychronization phase of an incremental dbrecover or replication. In case this problem is encountered, the incremental dbrecover or replication will correctly continue after this patch is installed. * The internal SCAN_REC function was enhanced to gracefully handle a BOF status and to output a detailed log message in case of a failure. * Fixed a corner case problem in btree error handling (#3052). If a btree page split fails due to lack of disk space a cache page might not be properly released. This may result in a subsequent server panic with a message like below: buf_Sync: PIN LEAK detected. bhp=40329c58, node=#116 Assertion failed: !(bhp->flags & BUF_PINNED) Aborting on internal failure, file mpool.c, line 926 * Fixed a problem in the volume file extension procedure that could result in growing a volume file infinitely until the disk space is exhausted (#3481). This could happen if the volume file size is limited below the current size by changing the VolumeFileSizeLimit config item to a value below the size of existing volume files. Platforms: HP-UX and Linux * The database server was enhanced to support a new configuration option to enable read access on forward-log files for the group (GID) specified in the database server configuration file (#3475). By default the database server creates any forward-log files with restrictive permissions that only allow the configured user (and the superuser) to access the forward-log files. The new [forwardlog] GroupReadAccess configuration option may be used to specify read access for the configured group to the forward-log files. # [forwardlog] # GroupReadAccess = 0|1 If set to a nonzero value forward-log files are created with a permission that allows group read access (configured with the [Server] GID option). If set to zero forward log files are created with a permission to restrict access to the owner (configured with the [Server] UID option). The default value is 0 to permit owner access only. Platforms: Windows * Changed the Windows file version to "7.1.1.23" Patch PE71-0709050 ------------------ Platforms: All * Fixed a problem with DBFIND modes 6 and 7 with arguments using wildcards (#3429). These DBFIND modes are used internally by the TurboIMAGE compatibility library (image3k library) to implement TPI and index access. This problem was introduced in the previous patch PE71-0708130. Platforms: Windows * Changed the Windows file version to "7.1.1.22" Patch PE71-0708130 ------------------ Platforms: All * Fixed a problem that could result in a crash of the server process during replication or forward-recovery of a database restructuring with a log message like below (#3444): Assertion failed: meta->ulist_cache_used <= (int)node->node.ulist.num_pages * A TurboIMAGE/SuperDex compatibility problem was solved (#3429). If the wildcard tokens (?, #, @) were used within a DBFIND search argument, using the "greater than (or equal)" or "less than (or equal)" conditions caused the server to return a status 53 (bad argument). A wildcard search argument could not be used to evaluate a greater/less comparison. The implementation was modified to support "greater/less than (or equal)" conditions on search arguments if leading literal characters are present in the search argument. In this case, the leading literal part is used to evaluate a greater/less comparison. * A TurboIMAGE/SuperDex compatibility problem was solved (#3000). After a DBFIND on an index using a "greater than" or "less than" condition, the DBGET modes 15 or 16 behaved as "greater than or equal" or "less than or equal", respectively. * Changed the Windows file version to "7.1.1.21" Patch PE71-0708030 ------------------ Platforms: All * Fixed a possible internal deadlock condition when purging huge databases (#3426). This problem was introduced as a side effect by changing the checkpoint locking behavior in PE71-0707200. * Changed the Windows file version to "7.1.1.20" Patch PE71-0707310 ------------------ Platforms: All * Fixed a problem resulting in dbutil database restructuring to fail with an error message like below (#3412): *** Database in use [-2] FATAL: Fatal problem during schema upload - can't continue This problem was introduced as a side effect with patch PE71-0707200 (#3111). * Enhanced the dbstore/dbrestore operations to support canceling the operation (#3421). If the dbctl session is terminated the dbstore/dbrestore operation will now terminate as well. A message as below is logged by the server: Session was terminated - database not stored Session was terminated - database not restored * The dbrestore operation was enhanced to support the /nice command line option (#3421). This results in better performance for concurrent processes while a dbrestore is ongoing, at the cost of slowing down the dbrestore. The dbstore operation also supports the /nice option but it makes little difference in practice. * Database restructuring was changed to include additional progress information in the log file. For each completed step a message is logged. In addition a progress message is logged each 10 minutes (subject to the LogFlags setting). Messages like below are output to the server log: restructuring data set 'ARTIKEL': 49760 records processed rebuilding indexes for data set 'ARTIKEL': 76024 records processed relinking detail data set 'BUCHUNG' paths: 1094500 records processed * Changed the Windows file version to "7.1.1.19" Patch PE71-0707200 ------------------ Platforms: All * Added the option to limit database transaction sizes (#3388). Two size limits are implemented: A configurable "softlimit" and an internal "hardlimit". The minimum of either value defines the max. size an uncommitted transaction may have. The internal "hardlimit" is determined by the half of the configured log space and subtracting the configured checkpoint size: configured log space / 2 - configured checkpt size The softlimit is configurable with the new TransactionSizeLimit config item. By default it is set to half the size of the internal hardlimit. For example, assuming a size limit of 1 GB for the log volume and a checkpt size of 50 MB the hardlimit would be 450 MB and the default softlimit would be 225 MB. Once the size of an uncommitted transaction reaches or exceeds the limit a status -801:28 is returned. The only valid options at this point are to commit or rollback the transaction. If the status -801:28 is returned by the DBCOMMIT call the only valid option is to rollback the transaction. A message like below is logged to the server log: Transaction size limit exceeded, size: xxx pages, limit: xxx pages * The new [Config] TransactionSizeLimit config item may be used to configure a size limit for database transactions. It is defined as below: This configuration item may be used to limit the max. size of a database transaction in MB. If set to zero, the transaction size is not limited. If set to -1 (the default), the size limit is set to a default value which depends on the configured log volume space. The default value is -1. * Fixed a potential starvation problem of the checkpoint thread. When a large number of transactions are committed concurrently the internal checkpoint thread could, in theory, be blocked for a long time. The internal locking protocol was changed to escalate the checkpoint lock after some timeout. * Fixed a problem that could result in a transient index problem on a replication slave when replication was active. A missing lock could potentially result in an inconsistent index lookup. * Fixed a server hang if the DBLOCK-COMPAT database property was set for a database and [Config] LockConflictingItems was enabled in the server configuration. This could result in an endless loop and hang the server process. * Fixed a problem where a database could be purged after it was renamed although it was still opened by another session (#3111). * Fixed a problem that could result in a crash of the server process if the server log flags were set to output debug and replication was active. * The item format flags in the node schema audit record was enhanced to indicate the role of an item as below: Bit 16 (0x10000) is set if the item is a search item. Bit 18 (0x40000) is set if the item is a unique key. Currently, this indicates it is a master search item. Bit 19 (0x80000) is set if the item is a sort item. * On the replication slave server, a DBCLOSE could wrongly resume the replication after a database is closed (#3383). * If forward-logging is disabled, the server now immediately stops writing to the forward-log (#3089). * Disabling forward-logging on the replication slave server (eg. through dbctl) could result in a server abort due to an internal inconsistency. * Changed the Windows file version to "7.1.1.18" Patch PE71-0705300 ------------------ Platforms: All * This patch adds the option to Eloquence B.07.10 to replicate database transactions to other database environments. For documentation please refer to the Eloquence web site: http://www.marxmeier.com/eloquence/support/B0710/doc/repl/ * The DBLOCK-COMPAT database property may be used to modify the locking policy for a database. This may be used to specify an IMAGE compatible lock behavior for a database. When set to 1, a DBLOCK is required for writing to the database. When set to 0 (or undefined) writing to a database does not require a previous DBLOCK (but no competing lock may be granted). This is the default behavior of Eloquence. This db property may be changed with dbutil, either interactively or using a script as below: database "SAMPLE"; create property "DBLOCK-COMPAT" value "1"; * A forward-log is now retained across a crash of the database server. If the server process terminates abnormally, a subsequent server start or use of the dblogreset utility performs a startup recovery. Previous releases disabled an existing forward-log in such a situation so that it became necessary to create a new backup. The eloqdb6 startup recovery and the dblogreset utility have been enhanced to reliably continue the forward-log after a server crash. Please note that audit information are not retained for any recovered transactions. The fwaudit utility outputs a warning note if it discovers this situation. * Changed the Windows file version to "7.1.1.17" Patch PE71-0704110 ------------------ Platforms: All * In rare cases, a DBPUT could fail with status -809:976 (#3255). This is caused by a race condition while the same automatic master entry is added/deleted concurrently. * A problem in the client request handling was fixed that could in some cases result in a slight prioritization of some connections over others (#3325). * DBGET mode 4 does not update index position (#3333) After a DBFIND on an index, a DBGET mode 4 could not be used to position the current record inside the result. The index position was not updated by a dbget mode 4. This was inconsistent with the behavior of image paths and not compatible with TurboIMAGE/SuperDex behavior. * The server was modified to solve a compatibility issue with TurboIMAGE applications (#3247). TurboIMAGE documents that a DBFIND resets the current record number for a data set. The Eloquence DBFIND call previously did not affect the current record number. This changed DBFIND behavior is enabled if both, this patch and the corresponsing client library patch PE71-0704100 (or superseding) is installed. Otherwise the behavior is unchanged. * The server was enhanced to support additional options for logging server or individual session performance information. The following config items are supported: [server] StatFile StatFile specifies a file name to be used for logging the server utilization. If enabled, this file is updated once a minute. As the file is re-opened each time it is updated it may be moved/deleted freely. This config item was introduced with the B.07.10 release. Please refer to the B.07.10 release notes for details. User visible change: The config file must specify an absolute file name. This is consistent with the corresponding dbctl command. [server] StatFileFlags StatFileFlags specifies options that influence the StatFile format. By default (StatFileFlags not set) the file content is replaced each time it is updated. Also, the content is formatted with multiple lines, each containing a descriptive text and the actual value, separated by a colon. The following flags are supported: s (single line) causes the values to be formatted into a single line. Values are separated by a space and no descriptive text is present. a (append) causes additional values to be appended to the file instead of replacing the previous content. t (local timezone) causes the timestamp to include the offset of the local timezone from UTC. If not present, the timestamp value denotes UTC. This flag allows to use the timestamp value with DSI (MeasureWare) on HP-UX without requiring a conversion. These flags may be combined (eg. StatFileFlags = sat). Example output (single line format): 1172193450 6 110 0 441 0 0 1 [server] SessionStatFile If specified, SessionStatFile is used for logging session utilization information. Depending on the SessionStatMode setting, information is logged when a session ends or after the next database call after the specified interval expires. This file is opened on the first event and kept open until a new value is specified with dbctl SessionStatFile or the SessionStatMode is changed through dbctl. The config file must specify an absolute path name. This is consistent with the corresponding dbctl command. The information logged to SessionStatFile is substantially similar to session details provided in the HTTP status and may be used to evaluate performance/behavior of an application after it has completed. Every entry in SessionStatFile consists of a single line, fields are separated by a vertical bar (|) character. The following information is provided in SessionStatFile: timestamp - the timestamp (UTC) the entry was added TID - The id of the database thread Type - Type of Entry (E or U character), E specifies the entry was logged when the thread had completed (application disconnected from the database), U specifies the entry was logged after the interval specified in SessionStatMode had expired OSuser - Operating system account used by the application DBuser - database login (most recently) used by the application ConnTime - Connect time (in ISO format YYYY-MM-DD HH:MM:SS) ConnSec - Number of seconds elapsed since connecting Stat - Three numerical values for each monitored database activity (a subset of database activity typically called from applications). The values specify two counters and the wall time for the sum of all calls of a category. The first counter specifies the number of database calls (from the client library, may be different from application calls if client side caching is used), the second counter may specify a count related to the particular call (see below). The wall time is specified in usec (1 mio per second). The follwing database activities are monitored IO_READ - Disk reads accounted to the session. This includes both reading activity as well as any disk reads required for writing activity. The count field specifies the number of IO requests, the count2 field specifies the number of pages (8K units). DBFIND - DBFIND calls. DBGET - DBGET calls (single record). DBGETB - Multi-record DBGET calls. These are used internally by the client library if client side caching is used. The count2 field specifies the number of records obtained. DBPUT - DBPUT calls DBUPDATE - DBUPDATE calls DBDELETE - DBDELETE calls DBLOCK - DBLOCK calls, count2 specifies the number of unconditional DBLOCK calls that could not be granted immediately but were blocked by a competing lock. DBUNLOCK - DBUNLOCK calls TXBEGIN - Begin Transaction TXCOMMIT - Commit Transaction TXROLLBACK - Transaction Rollback IP - IP Address and port number (separated by a colon) used to connect to the database AppEnv - Information collected from the application environment, such as process ID, operating system specific user id, (subset of the) command line, EQ_AUDIT_INFO content Please Note: The content of the SessionStatFile is subject to change without notice. Example output (single line): 1172196823|9|E|mike|public|2007-02-22 16:54:39|4|11|11|751| 0|0|0|0|2|163|45|6|10591|0|0|0|0|0|0|0|0|0|0|1|8879|0|1|6| 0|0|0|0|0|0|0|0|0|127.0.0.1:64169|uid=102 pid=4812 pname=query3k [server] SessionStatMode SessionStatMode is a numeric value that specifies when an entry is logged to the SessionStatFile. The following values are supported: 0 - (zero) The SessionStatFile is disabled 1 - (one) A log entry is written to the SessionStatFile when a session ends. Any other value is understood to specify an interval (in seconds) after which an entry is logged to the SessionStatFile in addition to the entry that is logged after the session ends. The specified value must be at least 60 seconds. * The server was enhanced to support additional dbctl commands to allow dynamically changing the logging of performance information. The following additional dbctl commands were added dbctl statfile [FileName|DISABLED] Obtain the current value of [server] StatFile or specify a new value. If used without additional file name argument this returns the current value for [server] StatFile. When a file name is present, it specifies a new value for [server] StatFile. If DISABLED is specified, the [server] StatFile is unconfigured and no longer used. Otherwise an absolute file name must be specified. The file may not exist. When a file name is specified dba capabilities are required. For example: dbctl statfile dbctl -u dba statfile /tmp/server.stat dbctl statfileflags [flags] Obtain the current value of [server] StatFileFlags or specify a new value. If used without additional flags argument this returns the current value for [server] StatFileFlags. When a flags argument is present, it specifies a new value for [server] StatFile. An empty argument may be used to reset the flags. When a flags argument is specified dba capabilities are required. For example: dbctl statfileflags dbctl -u dba statfileflags sat dbctl -u dba statfileflags "" dbctl sessionstatfile [FileName|DISABLED] Obtain the current value of [server] SessionStatFile or specify a new value. If used without additional file name argument this returns the current value for [server] SessionStatFile. When a file name is present, it specifies a new value for [server] SessionStatFile. If DISABLED is specified, the [server] SessionStatFile is unconfigured and no longer used. Otherwise an absolute file name must be specified. The file may not exist. When a file name is specified dba capabilities are required. For example: dbctl sessionstatfile dbctl -u dba sessionstatfile /tmp/session.stat dbctl sessionstatmode [mode] Obtain the current value of [server] SessionStatMode or specify a new value. If used without additional mode argument this returns the current value for [server] SessionStatMode. When a mode argument is present, it specifies a new value for [server] SessionStatMode. When a mode argument is specified dba capabilities are required. For example: dbctl sessionstatmode dbctl -u dba sessionstatmode 1 * The template config file for the database server was updated. This version adds the new config options provided with this server release and corrects some mistakes in the inline comment. Platforms: HP-UX * On HP-UX PA-RISC the listen queue backlog size was increased (#3324) Due to a build problem a smaller listen queue backlog value was used on HP-UX PA-RISC. If a large number of new connections is started concurrently this could, in some cases, cause new connections to fail temporarily (ETIMEDOUT error returned by connect). Platforms: Windows * Changed the Windows file version to "7.1.0.16" Patch PE71-0702080 ------------------ Platforms: All * A cache-miss (in the eloqdb6 buffer cache) might result in temporarily blocking concurrent access to the same set until the disk access was completed. If some structural information was not found in the buffer cache the server might fetch the data from disk with an internal lock held. This could impact performance. The server was modified to no longer hold this internal lock during disk access. In addition, internal lock handling was improved for database read operations. * A DBGET call might overly impact performance of concurrent applications when client-side caching is used (#3315). When client side caching is enabled (the default), a DBGET might request multiple records from the server in one request. The number of records requested from the server depends on previous application database use. In some situations the performance of concurrent applications might be impacted when a large number of records was requested in one call. The server was changed to improve concurrency in this case. * Aborting on-line backup mode recovery could result in data corruption (#3279). The server process (or the dblogreset utility) performs a recovery from on-line backup on startup if the server was previously shut down while in on-line backup. This recovery transfers any changes temporarily stored in the log volume(s) during the on-line backup mode to the data volume(s). Depending on the amount of data that was saved in the log volume(s) and disk performance this could take some time. If this recovery was interrupted (or aborted for some reason) in some cases this recovery could not be re-run afterwards because the generation count of the data volume would differ from the log volume(s) generation count. In this case the data volume is corrupted as the changes were only partially transferred. * Forward-log files are now created with restrictive file permissions. Only the user account configured to run the database server may access these files (#3304). * The dbctl logfile command no longer accepts existing files (#3303). * The [ForwardLog] EnableAudit and AuditOnly configuration values were not displayed in the HTTP status (#3311) Platforms: HP-UX and Linux * If the server process is started from the root user, ownership of a log file created during the server startup is changed to the user that is configured to run the database (#2368). * The eloqdb6 IO threads did not correctly change the log file after a dbctl logfile command. The previous log file was not closed (#3294). Platforms: Windows * Changed the Windows file version to "7.1.0.14" Patch PE71-0610130 ------------------ Platforms: All * A rare defect was fixed that could result in a server abort while stopping on-line backup (#3217). A message like below is written to the server log file: Assertion failed: page_addr < vol->curr_size file volume.c, line 6183 This problem could happen if during on-line backup the data volume was virtually extended and the volume list of available blocks had to be extended as well (this is required around every 500 MB). When on-line backup was later stopped, the stored free-list information was then applied in a wrong way. This problem was introduced with changes to the recovery code in the previous patch PE71-0609120. * Using dbrestore with a pipe (for example with gzip) could in rare cases cause the server process to hang (#2987). This was caused by a log message that was output asynchronously when the pipe was closed and could in rare cases result in internal memory corruption. * Changed the Windows file version to "7.1.0.13" Patch PE71-0609270 ------------------ Platforms: All * Fixed a rare race condition that could result in a server abort when extending a volume file (#2525). A message like below is written to the server log file: Assertion failed: page != 0 It is expected that this problem can only happen in on-line backup mode. As a workround the volume extension size for the log volume could be enlarged above the default 1 MB. * Fixed a corner case in the lock scheduler that could result in a server panic on a DBUNLOCK (#3204). A message like below is written to the server log file: Assertion failed: lock_granted_tail && lq->l_prev This problem was introduced with the revised lock scheduler in patch PE71-0607100. A specific sequence of DBLOCK and DBUNLOCK calls could result in a situation that violates an internal assumption. * Changed the Windows file version to "7.1.0.12" Patch PE71-0609120 ------------------ Platforms: All * Fixed a problem in btree management that could result in an internal error when a page split operation failed due to insufficient disk space (#3052). * Corrected an internal problem where a zero record number in the free-record list could in rare cases result in an internal failure during the checkpoint (#3131). * A problem was solved that could result in an inconsistency of the internal volume generation number between data and log volumes (#3129). This problem could result in an error during eloqdb6 startup as below: failed to open volume: volume #1 has inconsistent generation count This could happen under a rare condition when forward-logging was configured but disabled subsequently and afterwards the eloqdb6 process terminated unexpectedly. It could also happen if the forward log was configured and the eloqdb6 process did abort twice in a row. In both cases, the log volume header was not updated and therefore contained a wrong volume generation. * Fixed problem with the volume generation number that could get out of sync if forward-logging was restarted during on-line backup and the eloqdb6 was then stopped while still in on-line backup mode (#3130). * Fixed a potential corruption of the server catalog that could happen in some cases if the server was aborted after changing the server catalog (schema, dbcreate, dbutil, dbrestore) but before running an internal checkpoint (#3163). In this case the records allocated for the server catalog could not be marked in use by the crash recovery and become corrupted subsequently. This problem could only affect eloqdb6 patch levels PE71-0606290 and newer. * Session statistic was enhanced to separate regular DBGET calls from multiple record fetches (using client side caching). * Changed the Windows file version to "7.1.0.11" Patch PE71-0608110 ------------------ Platforms: Linux * This patch solves a compatibility issue with the Linux glibc2.4. Patch PE71-0607210 ------------------ Platforms: HP-UX, Linux * In some situation the queue entry of an completed IO request was not released (#3118). This could after some time result in the server message tio_enqueue: TIO queue is full Patch PE71-0607100 ------------------ Platforms: All * The lock scheduler was revised to enhance scalability and performance with a large number of competing locks. * Changed the Windows file version to "7.1.0.10" Patch PE71-0606290 ------------------ Platforms: All * Improved scalability of dbunlock with a large number of blocked competing lock requests. * Fixed a problem with forward log files that could cause an internal failure with dbrecover as below (#3072). L0: Fwr_TranslatePage(page=0x3229) failed: page not found Assertion failed: page not found, file volfwr.c, line 1504 The problem was triggered by an inconsistency in the forward log that could happen in some cases if dbrestore was used while the eloqdb6 server was active. This problem could no longer happen as a side effect of installing patch PE71-0603280. This patch provides a fix for the root cause. * The forward log format that was slightly changed. This requires installing updated utilities that access the forward log. * Changed the Windows file version to "7.1.0.9" Platforms: HP-UX * Work around HP-UX problem with memory windows (#3050). When using memory windows on HP-UX with a configured buffer size of 1 GB the server process could abort with a SEGV with a message like below: ** Cought signal 11 -- traceback follows: signo = 11 addr = bffffff0 As a workaround, a configured buffer size of exactly 1 GB is internally reduced by one page. Patch PE71-0603280 ------------------ Platforms: All * Fixed a problem with forward log files that could cause an internal failure with dbrecover (#3059). Internal failure: temporary file for transaction # not found Assertion failed: Internal failure during forward-recovery This problem was introduced with patch PE71-0602280 that enables use of a more efficient format to record index changes in the forward log files. In a specific case this could result in an inconsistent transaction status recorded in the forward log file that is catched by a consistency check in dbrecover. This problem can possibly affect installations that previously installed patch PE71-0602280 or PE71-0603100 and use additional indexes. A patched dbrecover is available on request to workaround this problem if necessary. * Fixed a problem with forward logging not enabled after a server abort without a server restart (#3056). A message like below is written is the server log file: Note: forward-logging has been disabled because the server was not shutdown cleanly After a server abort any recently committed transactions are re-applied. This temporarily disables forward logging. A temporary flag was not reset after completing recovery causing forward logging to remain disabled until the server process is restarted. Patch PE71-0603100 ------------------ Platforms: All * Improve bimport performance on master sets (#3045). The bimport performance for master sets is often dominated by creating the associated index. This may also apply to detail sets when an additional index is used. The eloqdb6 server was changed to bypass the index transaction logic during bimport to improve database load time. Patch PE71-0602280 ------------------ Platforms: All * An internal deadlock condition was fixed that could result in on-line backup requests (dbctl backup start|stop|status) to hang (#3014). A server shutdown would not succeed in this case requiring to kill the server process. An on-line backup request needs to obtain an internal lock to ensure transactional consistency. If an on-line backup request gets blocked initially due to a concurrent operation (for example a dbstore operation is currently running) and another on-line backup request is issued a deadlock condition occurred that caused both requests to hang. * The server process was enhanced to use a more efficient format to record index and meta data changes. This enhancement results in a substantial reduction in disk space for the forward-log file when index entries are changed frequently. This change required a modification of the forward-log file format. While older forward-log files are still supported, related utility programs need to be updated to handle the new format. This requires installation of additional patches on dbrecover, dblogreset and fwaudit utilities (see related patches). * A rare defect with forward-recovery was fixed that could result in a corrupted volume set (#2999). This causes an internal consistency check to fail and eloqdb6 to abort on startup with an error message like below: Assertion failed: page == new_vol->flist_tail server panic: Aborting on internal failure, file volume.c, line 3988 This problem could happen if the data volume was extended and the volume list of available blocks had to be extended as well (this is required around every 500 MB). The recovery then missed to update the flist_tail field in the volume header, causing a panic the next time the server was restarted. The dbrecover utility was fixed to correctly update the flist_tail field. In addition the eloqdb6 behavior was modified to detect and correct this problem on startup. A notice is written to the server event log file in this case. * On HP-UX and Linux some log messages regarding eloqdb6 startup and shutdown were raised in priority so they show up in the server event log file even if LogFlags=*0 (default) is used. The following messages are affected: - "Eloquence database server active" when the eloqdb6 server process started successfully. - A new message outputs the current server patch level to the log. - "Eloquence data base server terminated" when the eloqdb6 server finished shutdown. - The message when a new forward-log segment was created. On all platforms the message indicating a previous unclean shutdown was recovered is now written to the server log. Platforms: Windows * On the Windows platform, dbctl bimport refused an absolute path beginning with a drive letter (#2997). * Changed the Windows file version to "7.1.0.7" Installation: ------------- Please download the patch archive that corresponds with the installed release. The patch files follow the conventions below: PE71-0906050-hpux-ia64.tar.gz ^ ^ ^ | | Architecture / OS specific build | Operating system Patch ID HP-UX: In order to install this patch, you need to unpack it with gzip and tar. Gzip is included with HP-UX. Installation requires root privileges. cd /opt/eloquence6 gzip -dc /path/to/PE71-0906050-hpux.tar.gz | tar xf - Files: bin/eloqdb6 newconfig/config/eloqdb6.cfg share/doc/PE71-0906050-README Linux: In order to install this patch, you need to unpack it with tar. Installation requires root privileges. cd /opt/eloquence6 tar xzf /path/to/PE71-0906050-linux.tar.gz Files: bin/eloqdb6 newconfig/config/eloqdb6.cfg share/doc/PE71-0906050-README Windows XP/2000/NT: This patch should *only* be installed if you previously installed the Eloquence server components on your system. Installation requires administrative capabilities. Two options are available for patch installation. The patch is available as self extracting archive for automatic installation and as a zip archive for manual installation. Both patches are equivalent. Installation requires administrative capabilities. For automatic installation of this patch, please download the patch file PE71-0906050-win32.exe. Before installation, please consider closing all applications, then execute the patch installation program. Installation does not require a reboot unless the patched files were active. For a manual installation of the patch, please download the patch file PE71-0906050-win32.zip and unpack its contents. Then perform the following steps: * Please make sure the eloqdb6 service has been stopped previously (in the Service Control Manager or with NET STOP eloqdb6). * Please copy the eloqdb6.exe file into the WINDOWS SYSTEM DIRECTORY (for example C:\Windows\System32). * Please copy the eloqdb6.cfg.sam file into the etc subdirectory of your Eloquence installation (for example C:\Programs\Eloquence\etc). * Please copy the PE71-0906050-README.txt file into the share\doc subdirectory of your Eloquence installation (for example C:\Programs\Eloquence\share\doc). Files: eloqdb6.exe eloqdb6.cfg.sam PE71-0906050-README.txt