IBM WEBSPHERE CONTENT DISCOVERY SERVER, VERSION 8.3 FIX PACK 6: README (c) Copyright International Business Machines Corporation 2005,2008. All rights reserved. US Government Users Restricted Rights - Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. PREREQUISITE You must install IBM(R) WebSphere(R) Content Discovery Server Version 8.3 before installing this fix pack. INSTALLING THE FIX PACK 1) Stop any processes that are running under the WebSphere Content Discovery Server Version 8.3 installation directory. 2) Back up all files in the WebSphere Content Discovery Server Version 8.3 installation directory. 3) Download the appropriate file for your operating system: o WCDS_8.3_patch_6_aix.tar.gz (for IBM AIX(R)) o WCDS_8.3_patch_6_linux_RH.tar.gz (for Red Hat Linux(R)) o WCDS_8.3_patch_6_linux_SUSE.tar.gz (for SUSE Linux) o WCDS_8.3_patch_6_sunos.tar.gz (for Sun Solaris) o WCDS_8.3_patch_6_win32.zip (for Microsoft(R) Windows(R)) 4) If you are using Java(TM) JavaServer Pages(TM) (JSP) or Microsoft Active Server Pages (ASP) layouts, download the file WCDS_8.3_patch_6_layouts.zip. 5) Extract the files to the WebSphere Content Discovery Server Version 8.3 installation directory. Note: You might need to grant write permissions on the AIX machine for the following directories: web, aix, platform/common 6) Run WebSphere Content Discovery Server. The fix pack is installed automatically. 7) To verify that the installation was successful, run the iphrase version command in the WebSphere Content Discovery Server Version 8.3 installation directory. The version should be 8.3.8013. UPDATING THE INTERACTIVE ANALYTICS REPORTING SOLUTION If you are using the Interactive Analytics reporting solution, follow the instruction below. Note: If you followed these instructions when you installed a previous fix pack, you do not need to follow them again. 1) Stop the MicroStrategy(R) Server. 2) Locate the MDB (Microsoft Access Database) file that is referenced by the DiscoveryServer83_Metadata DSN (either Oracle_DiscoveryServer_Metadata.mdb or MS_SQL_DiscoveryServer_Metadata.mdb, depending on the database that is used for reports). 3) Replace the existing MDB file with the MDB file of the same name that is contained in the corresponding zip file in the fix pack. 4) Run the MicroStrategy configuration wizard to restart the MicroStrategy server. For instructions, see: http://publib.boulder.ibm.com/infocenter/discover/v8r4/topic/com.ibm.discovery.ds.install.doc/t_IG_Configuring_MicroStrategy.htm 5) From the command line, change the directory to the WebSphere Content Discovery Server installation directory. 6) Run the following command: bin/iphrase reportAdmin -addDateViews -p deployment///default.prp PROBLEMS FIXED This patch contains the following fixes: Fix Pack 6 ---------- * 22042 - Admin should not allow -startServer alone * 22393 - For database content sources, Management Console metadata edit shows all None values * 23493 - XML connector - automatic binding generation disabled * 23547 - Live Build creates journal entries greater than SERVER_MAX_REQUEST_BYTES * 23585 - Change csv output to use double quotes instead of single quotes (APAR IC52752) * 23620 - Wrong Management Console password gives mysterious error (APAR IC52866) * 23653 - Internet Explorer 7 does not work properly with 8.3 Management Console (APAR IC53025) * 23916 - Business rule fires twice (APAR IC53886) * 24241 - Traceback when query contains term with multiple wildcard characters (APAR IC54701) * 24245 - List index out of range traceback for some queries (APAR IC54701) * 24248 - Phrase matches are not correctly scored if found in multiple indices * 24249 - DataEngine EXCEPTION_ACCESS_VIOLATION for synonym searches (APAR IC54462) * 24250 - Term score not correct for multi-term phrase constraints (APAR IC54689) * 24252 - Match terms with punctuation ("won't") do not highlight properly (APAR IC54691) Fix Pack 5 ---------- * 23211 - Business Rule Trigger "Queries containing feature constraints" (APAR IC51254) * 23247 - OR'd constraints not functioning properly (APAR IC51335) * 23261 - Server failover mechanism isn't functioning (APAR IC51373) * 23268 - Memory usage spike during build with acronym generation (APAR IC51385) * 23360 - Unable to upgrade XML connectors (APAR IC51728) * 23369 - -exportCsv does not populate rowsData.csv or datelookup.csv (APAR IC51773) * 23482 - Multi-thread spider fails under certain conditions (APAR IC52323) * 23418 - IR relevance score computation error * 22575 - RENDER_TAB_FEATURE not working properly * 22469 - Unable to start the Management Console on AIX * 22627 - No results voiceover for quoted and unquoted search strings (APAR IC50982) Fix Pack 4 ---------- * Management Console server tries to use default.prp setting for numworker (APAR IC51311) * Server becomes unresponsive (APAR IC51590) * Server failover mechanism is not functioning (APAR IC51373) Fix Pack 3 ---------- * Hooks cannot access/change cgi arguments in soap requests * BR trigger "results containing direct answer" does not work * Unhandled exception when turning on metaData.txt logging * Acronym detection does not allow spaces in iphrase_home path name * SOAP Exception when trying to add business rules * Displayed text differs for quoted and unquoted search strings (skins and LE) (APAR IC50982) * Ir_index_print -checkIndex' produces memory leak * Notes connector incompatible with Unicode & Python 2.4 * Business rules trigger on initial query but not on refinements * Error in regex for detecting non-ASCII encoding in http/mime headers * Unnecessary check for IR column type causes numerous unwanted error messages * Navigation tabs cannot be rendered (APAR IC50951) * Upgrade fails with cast error * Direct Nav business rule fails due to missing action.portlet * Boolean search drops term from breadcrumb (APAR IC51243) * MC.NET creates empty text triggers, causing business rule to always fire (APAR IC51203) * Time aggregates produce incorrect results in Content Util reports (APAR IC51255) * Summarized 'None' values cause crash in SOAP rendering (APAR IC51264) * Text features having 'ir_extract_type: snippet' cause traceback (APAR IC51266) * BR trigger "User clicking on a leaf category" causes server error * Need option to turn off incremental build tracking (APAR IC51326) * Exception in modifyServerHTTPRequest are not logged (APAR IC51338) * Build failure not reported by admin (due to DB retry?) (APAR IC51339) * JDBC reconnects, but original error is lost (APAR IC51346) * DB reconnects after build build abort (Ctrl+C) (APAR IC51351) * "Generate acronyms" doesn't work with structured content sources (APAR IC51360) * Exact phrase match queries not working (APAR IC51365) * Cannot specify tool property variables * Federation: Crash in DataEngine on front end during tally compute (APAR IC51393) * Word splitting does not product a 'respell' message. (APAR IC51092) Fix Pack 2 ---------- * modifyServerHTTPRequest / modifyServerHTTPResult should be called for all URLs * No results should not return all results * Business rules are not scalable enough * Solaris,AIX:traceback while running spider wizard. * Single live build update gen takes minutes to apply, 100% CPU * pre pruned constraint clears too many docs with multiselect tables * Upgrade to 8.3 should fix "lib" lines in hooks.py * Improve performance of prePruned logic * Searches with dashes and quotes * rstrip traceback during report exportdb * admin publish via ftp on Solaris is failing * lib/database used by swat-library should check db connection before each use * spider hangs after TB * Underscores in words do not produce results. * NCR Server Crashing - Java HotSpot in DataEngine.so * End of JDBC build dies with maximum recursion depth exceeded * build fails to index iframe content when src=XXX URL is relative * Library initialization fails to call hook instance of MetadataEnhancer * ir prepruned needs to be extended to handle additional use cases * RENDER_TAB_FEATURE not working properly Fix Pack 1 ---------- * An incremental build consumes three times more RAM than a clean build. * The live build server does not stop immediately when an abortlivebuild command is issued. * An error is returned when joining tables in an IBM DB2 hierarchical database. * ODBC and JDBC content sources fail to pass select max (id) to the modifyDBSetup Extension API. * A “malloc of 0 bytes” error message is returned upon starting the server if no IR indices are specified for searching. * Acronym extraction can save XML files with invalid characters. * Alphanumeric tokenization needs to be more configurable. * Negative memory values can be returned when the memory usage exceeds 2 gigabytes on Windows machines. * An error is returned when saving the Metadata Editing page. * The Management Console allows connections to a server when the -edit argument is not specified. * An “iteration over non-sequence” error message is returned when a feature is selected in the Edit Collection Features window. * OR constraints in business rules are not supported. * The tally_sort_order attribute is ignored for taxonomy features. TRADEMARKS The following terms are trademarks of International Business Machines Corporation in the United States, other countries, or both: IBM AIX WebSphere Java and all Java-based trademarks are trademarks of Sun Microsystems, Inc. in the United States, other countries, or both. Linux is a registered trademark of Linus Torvalds in the United States, other countries, or both. Microsoft, Windows, and Windows NT are trademarks of Microsoft Corporation in the United States, other countries, or both. Other company, product, or service names may be trademarks or service marks of others.