Switch to standard view 
  Sybase logo
 
 
 



Targeted CR list for ASE 12.5.1 ESD #1

279264 Support DOL worktables requiring GETSORTED operation. During a query's execution when ASE internally needs to create a worktable, if it decides to create it with a 'Data Only Locked' locking scheme, then errors such as 614, 691, 692, 696, 804, and 831 may occur.
300435 When SET PLAN DUMP for Abstract Plans is enabled, the plans for statements inside a BEGIN-END block will be missing.
300931 dbcc settrunc may report memory allocation warnings when it is run in a database configured for Replication Agent.
308518 When creating a clustered index with sorted_data option, error 1509 Severity 20 State 1 may sometimes occur: "Row compare failure of RIDs (<pagenum>, <rownum>) and (<pagenum>, <rownum>), keycnt = <keycount>, compflags = <value>, rowformat = <value>, minlen = <value>.
317310 A 10330 error, "SELECT permission denied on object <object>, database <database>, owner dbo" may be incorrectly reported when a stored procedure executes a SELECT ... UNION on a table in a different database."
317331 ASE will stop accepting subsequent RPC calls (until it is restarted) after one fails with an error 1601 in the errorlog.
318945 Langinstall fails with 102 and 105 errors when installing thai utf8.
319256 On Windows platform, the configuration file is not copied to the new directory when upgrading ASE using a separate directory. The resulting ASE is then booted with the default configuration values.
319374 Null extended row may not be returned, if OR condition is specified for an index column in the inner table of an outer join and 3 or more tables are involved.
321599 In some cases the message "The requested update to the OAM for object <value>, dbid <value>, index <value>, oam page <value>, would cause a negative page count..." may be incorrectly reported in the error log when an error occurs whilst running a REORG REBUILD command.
322329 Certain queries involving a join between tables on columns with different datatypes and using subqueries could perform slower than expected in some scenarios because the subqueries are executed for all rows instead for only the ones that match the join condition.
322947 In rare circumstances, a 605 error "An attempt was made to fetch logical page '<value>' in database.... Page belongs to object with id '<value>', not to object '<value>'", a 69* error or an 806 error may be reported when a duplicate index key is inserted into a table having both a non unique clustered index and some overflow pages. This is a run-time error and the table is not corrupted since DBCC CHECKTABLE does not report any error.
323730 An ALTER TABLE operation with ADD, DROP or MODIFY COLUMN on a large Data Only Locked table having a placement index may cause subsequent DBCC CHECKTABLE to report 7928 errors: "Index <indexname> is not consistent; found <leafrow_num> leaf rows but <datarow_num> data rows. Drop and recreate the index". Subsequent REORG REBUILD on this table may result in data loss.
325280 The message "currect process infected with 11" is reported in the errorlog together with a stack trace which includes the module 'execudf' when bcp in to a table with a UDF default value.
325550 A 913 error, "Could not find row in sysdatabases with database id ... Run DBCC CHECKTABLE on sysdatabases" followed by a 14208 error "Unable to execute SQLJ function .... The function schema may have changed since the command was compiled." may be reported when executing a SQLJ function from a stored procedure or a view in a recently loaded database.
325943 Sub-optimal performance may result under certain conditions when executing a query involving a MAX() or MIN() aggregate and join operation between two or more tables because the optimizer chooses the best index to resolve the aggregate function. New command line traceflag provided to avoid aggregate optimization for queries involving joins.
326060 The error 10812 together with a stack trace which includes the modules 'trunctab' and 'cfg_notify_current_audit_table' may be reported when trying to change the current audit table using sp_configure, if the process gets into a deadlock or it receives an attention while trying to truncate the target audit table.
326152 Replication Agent may shutdown with a stacktrace, including function ra__pack_struct_tok(), followed by stack overflow, when trying to process an inconsistent log record.
326620 The message "process infected with 11" in the module 'dbt_curseq' together with a stack trace that includes the module 'lddb_chkcreator' may be reported in the error log when loading a database dump that was originally dumped on Adaptive Server 12.0.x.
326692 CIS : INSERT INTO <proxy table> command may insert additional EUCJIS character 0xA1C7 or SJIS character 0x8166.
326912 HA: Upon hitting 5156 error, dynamically increasing the number of devices on both primary and secondary does not resolve the issue.
326957 HA: HA failback operations may result in unexpected gaps for identity columns if any.
327503 Allow for CSMD configuration to get a memdump for error msg "ubo_object_from_slot:%s Invalid slot id %d".
327534 The message "current process infected with 11" in the module 'Address 0x00000000 ()' together with a stack trace which includes the modules 'send_proc_audit' and 'cast_const_to_varchar' may be reported in the error log if auditing is enabled and a stored procedure using Java parameters is executed.
327546 The error 923, "User <value> not allowed in database '<value>' - only the owner of this database can access it.", may be incorrectly raised when trying to online a database that is setup for "dbo use only", if the database owner in the dump has a different user id from the database owner of the database being onlined.
327744 Rows may be missing from the result set of a query similar to "SELECT DISTINCT <col1>, <col_unichar> ... ORDER BY <col_unichar>" when the ORDER BY clause contains a uni(var)char column and not all the columns in the SELECT DISTINCT clause are in the ORDER BY clause.
327860 The message "current process infected with 11" in the module 'decision' may be reported in the errorlog for queries involving joins and qualified table names.
328351 CIS : In rare circumstances, a task which issues the "set proxy" command in a HA environment will become hard affinity to the engine.
328709 Some rows may not be updated when an UPDATE statement has a FROM clause with a view containing a UNION ALL. In some cases the UPDATE may result in the message "current process infected with 11" in the module 'update()'. A DELETE using the same view in the FROM clause will result in the message "current process infected with 11" in the module 'norm3_setop()'.
329488 A 10711 error: "Cannot find a public method named '<method name>' in class '<class>'. Either the method does not exist, or it is being called with the incorrect argument types." may be raised when trying to execute a new method added to a java class in a database that has been just loaded.
329756 Historical Server crashes sometimes at the beginning of a view refresh
330159 Multiple sessions get blocked on Address locks when running Java applications. This blocking results in a poor performance for these connections. Optimizations were made to the JAVA VM to allow more concurrency.
330841 Starting with ASE 12.5.0.3, CPU utilization may go to 100% when "enable housekeeper GC" is set to 1, and the housekeeper queues have filled up to more than 80% of their maximum capacity because of an old open transaction.
330962 The JVM could incidentally trap with a Segmentation Violation (Infected with 11) during engine offline requests.
331083 A 511 error, "Attempt to update or insert row failed because resultant row of size <value> bytes is larger than the maximum size (<value>bytes) allowed for this table." maybe reported when executing an INSERT statement based on a SELECT using a view and a UNION clause. The error is raised when the target table has a column with a DEFAULT using a builtin function like suser_id().
331168 Increasing "size of process object heap" causes slow response time when executing single user java applications. Also at times Out of memory error is returned under a multi user scenario.
331545 After loading a database from an older server version (that is, one which requires upgrade to make it usable by the current server), "online database for standby_access" performs the upgrade even though the database is marked as being in standby mode. This prevents subsequent LOAD TRAN in that database.
331606 The message "Database '<db>' is now online with a lower server version number. Refer to the higher version release notes on the limitations of backward compatibility", may be incorrectly printed during the online of a database that was loaded from a dump which originated on an 11.9.x (or earlier) version of Adaptive Server.
331758 Under some circumstances incorrect results may be returned when executing a query using "GROUP BY" clause on columns defined with fixed char datatype and when the Adaptive Server is configured with a non binary sort order.
331777 Mirroring of master device on IBM may lead to 'uninitialized logical page' message
332268 When ASE runs out of LOCKS, 707 error "System error detected during attempt to free memory at address 0x<value>..." may be raised resulting in the session being killed. Alternatively an "infected by 11" error may occur. A stack trace will be reported in the error log showing modules 'proc_read', 'proc__fetchphdr' and 'proc_read__backout'.
332378 A UNION VIEW used in an INSERT query may result in a 622 error where the database for the target table for the INSERT is different than the database for the base table of the view.
332429 Historical Server memory corruption, potentially causing crashes
332447 When loading a transaction dump into a database that is bigger than 38000 pages, if the source and target databases have a different layout, the segment map will be set to SYSTEM + DEFAULT (segmap = 3) if the source database has segregated log. If the source database has mixed log and data, the segment map will be set to SYSTEM + DEFAULT + LOGSEGMENT (segmap = 7).
332536 The branch of a distributed transaction will continue to persist after the transaction manager has committed the transaction, if the branch has incorrectly nested transactions.
333077 Queries containing GROUP BY with an expression using local variable may return duplicate rows. The workaround for this problem is to remove the local variable from the GROUP BY list or to add a DISTINCT clause to the query.
333216 The error 643 together with a stack trace that includes the module 'tmp_check_nulls', or the error 233 reporting an empty column name, will be incorrectly raised if a table is created inside a stored procedure including the 'ON segment' option.
333614 Server fails to boot and will raise 1569 error when changing character set from utf-8 with one of the external sortorders (big5bin:194, cp932bin:192, cyrdict:140, eucjisbn:192, euckscbn:161, gb2312bn:137, gbpinyinbn:163, rusdict:165, sjisbin:179 and turdict:155).
333873 SQL Derived tables cannot have unnamed columns. The columns have to be always named either using the "AS" clause or the derived column list. Else, the user will get 11073 error. select * from (select sum(advance) from titles) d1 will get 11073 error. Rewrite as: select * from (select sum(advance) from titles) d1(a1) or: select * from (select sum(advance) as a1 from titles) d1
334069 Re-loading the Job Scheduler sequencer code by multiple stop / start operations can cause corruption within the JS Task sequencer memory table resulting in Job Scheduler Task violations such as Segmentation violations or Bus errors. When this happens the only way to restart the Job Scheduler is by rebooting ASE.
334358 In rare circumstances, a distributed transaction may be left in an "In Abort Tran" state after one thread of the transaction manager attempts to issue a rollback request for the transaction when another thread was still busy using the transaction. This would typically happen when some sort of error had been encountered by the transaction (for example, an out of locks error).
334729 On Solaris, HP, AIX and Linux platforms, Job Scheduler template files installTemplateProcs.bat, installTemplateXml.bat and installTemplateXml are missing.
334831 If a row level access rule is attached to a table involved in an outer-join query it may generate incorrect results.
335343 CIS : Conversion/truncation errors may occur, When an Engine is brought online after it was offlined and while inserting into proxy database involving Unichar columns.
335657 The message "current process infected with 10" together with stack trace which includes the module 'PDFreeDataChain' may be reported when an error is found at SSL initialization or while adding a certificate.
335774 If there are two derived tables at the same level and only one of them has a derived column list, then it may result in error 11061.
338157 ASE 12.5.1 does NOT support XPath functions. However, an XPath query with function operator gives incorrect error message 'insufficient memory' instead of 'XPATH: Parenthesized expression not supported'.
338523 In rare circumstances, running a Java native call produces a hard fault in the ASE Java Subsystem. The message "current process infected with 11" in the module 'ubffree' together with a stack trace which includes GarbageCollection modules may be reported in the error log on Unix platforms.
339731 Executing a stored procedure whose name has been stored into a T-SQL variable of CHAR datatype may fail and raise an unexpected 2812 error stating the stored procedure was not found whereas it actually exists in the database.
340261 After upgrading from Mac OS X 10.2 to 10.3 all attempts to login to ASE fail due to invalid passwords.
340554 Connectivity include files oscompat.h, oserror.h, ospublic.h, srv.h, and srverror.h are missing from $SYBASE/OCS-12_5/include directory.
340914 Incorrect version of installsecurity script being distributed with ASE 12.5.1 on Windows platform. Correct script to use is distributed as instsecu script.

 



Back to Top
© Copyright 2010, Sybase Inc.