Cause: Both define_array and bind_array have been called on this cursor. This was illegal. It was not possible for a cursor to contain both array binds and array defines. The semantics of this setting were invalid. Array defines were used to move data from select queries into PL/SQL tables and array binds to bind PL/SQL tables to non-select queries. Cause: The system failed to match the referred owner. There are two cases that this may occur. (1) This user owns data in the transportable set, but this user is not specified in the TTS_OWNERS list of the import command line option, assuming that TTS_OWNERS is specified. (2) This user is specified in the TTS_OWNER list, but this user does not own any data in the transportable set. Cause: The system failed to match the referred tablespace. There are 2 cases that this may happen. (1) This tablespace is in the transportable set, but it is not specified in the TABLESPACES list of the import command line option, assuming that TABLESPACES is specified. (2) This tablespace is in the TABLESPACES list, but it is not in the transportable set. Cause: A parameter was specified as an argument to procedure CREATE_PLAN_DIRECTIVE or UPDATE_PLAN_DIRECTIVE of package DBMS_RESOURCE_MANAGER that is not allowed when the GROUP_OR_SUBPLAN is a plan. Only the following parameters are allowed when GROUP_OR_SUBPLAN is a plan: - MGMT_P1, MGMT_P2, MGMT_P3, MGMT_P4 MGMT_P5, MGMT_P6, MGMT_P7, MGMT_P8 - UTILIZATION_LIMIT Action: It is possible that an attempt was made to create the Java class from a damaged class file, in which case the CREATE should be reattempted with a correct class file. It is also possible that the message is the result of using "-" in the resolver so that the verifier could not check the correctness of some code. In that case, the class needs to be created with a resolver. Action: It is possible that the Java class was created from a damaged class file, in which case the CREATE should be reattempted with a correct class file. It is also possible that the message is the result of using "-" in the resolver so that the verifier could not check the correctness of some code. In that case, the class needs to be created with a resolver. Cause: An error occurs during a conversion between datetime/interval and string data type due to one of the following reasons: - The buffer is too short to hold the result. - The format string is bad. Perform one of the following actions: - Remove the use of the PL/SQL function. - Mark the PL/SQL function as DETERMINISTIC. - Disable the REWRITE option on the materialized view. The function should be marked DETERMINISTIC only if it always returns the same result value for any given set of input argument values, regardless of any database state or session state. Do not mark the function as DETERMINISTIC if it has any meaningful side-effects. Action: Reorganize dimension levels and hierarchies into a single dimension such that no column is used to define levels in different dimensions. There is no limit on the number of levels or hierarchies you can place in a dimension. A column can be used to define any number of levels provided all such levels are in the same dimension and provided no two levels contain identical sets of columns. Action: In order for the materialized view to work correctly, any fine grain access control procedure in effect for the query must return a null policy when the materialized view is being created or refreshed. This may be done by ensuring that the usernames for the creator, owner, and invoker of refresh procedures for the materialized view all receive a null policy by the user-written fine grain access control procedures. Cause: A JOIN KEY clause was specified that did not conform to certain requirements. A JOIN KEY clause connecting a child level with its non-immediate ancestor level is allowed only when that child level and its ancestor level satisfy the following conditions: 1. The immediate parent of the child level must be a skip level. 2. The child level cannot be a skip level. 3. The ancestor level must be a non-skip level. 4. The ancestor level must be the nearest non-skip level to the child level in the hierarchy. Cause: The dimension statement failed because the column(s) in the JOIN KEY clause permitted NULL values. The JOIN KEY columns are related to the columns of the dimension's skip levels in one or both of the following ways: 1. A skip level is defined over one or more of the JOIN KEY columns. 2. The attribute clauses of a skip level determine one or more of the JOIN KEY columns. Cause: All identifiers used in the DBMS_OLAP package must have at most 30 characters, and must start with a letter that is followed by any combination of letters, numerals, and the signs '$', '_', and '#'. Other characters cannot be used in identifiers. The rules for DBMS_OLAP identifiers are the same as for PL/SQL identifiers. Cause: The specified filter is invalid. It contains at least one invalid filter item. If a filter item has a string list, it becomes illegal when the string list cannot be successfully parsed. If the filter item contains a range definition, and the lower bound of the range is greater than the higher bound, the item also becomes invalid. Cause: The OVER keyword was expected but not found. It may be due to one of the following reasons. 1. The OVER clause for the window function was missing. 2. There were unexpected identifiers before the OVER keyword. Note that for FIRST_VALUE, LAST_VALUE, NTH_VALUE, LEAD, and LAG, the IGNORE/RESPECT NULLS and FROM FIRST/LAST can only be specified either in the argument or before the OVER keyword. 3. A ranking function was given, and it was intended to be used to aggregate, but the argument of the function was missing. Cause: If the window specification is specified using RANGE option and there are multiple ORDER BY expressions, then the aggregation group cannot contain any expression (It can only have CURRENT ROW, UNBOUNDED PRECEDING, or UNBOUNDED FOLLOWING). First end point (bound) cannot be UNBOUNDED FOLLOWING and second end point cannot be UNBOUNDED PRECEDING. If the first end point is CURRENT ROW, then second end point can only be CURRENT ROW or expr /UNBOUNDED FOLLOWING. If the first end point is expr FOLLOWING, then second end point can only be expr /UNBOUNDED FOLLOWING. Cause: An attempt was made to modify an object more than once in a transaction. This error is usually caused by a DDL statement that fires a system trigger that tries to modify the same object. It can also happen when an attempt is made to perform more than one DDL operation on a queue table in the same transaction without issuing a commit between operations. Action: If the function is deterministic, mark it DETERMINISTIC. If it is not deterministic (it depends on package state, database state, current time, or anything other than the function inputs) then do not create the index. The values returned by a deterministic function should not change even when the function is rewritten or recompiled. Cause: A connection had to be routed through a local socket rather than a dispatcher. When this occurs, shared servers are less beneficial because the session that owns the socket cannot relinquish the process until the socket is closed. The most likely cause is that no dispatcher is configured for protocol=tcp. Cause: The execution of the session was stopped by the debugger. This can occur because of an explicit request sent by the debugger, or because the debugger disconnected without requesting the ORACLE server to let the session continue to run after the disconnection. It can also occur if the ORACLE server needs to disconnect the debugger at a time when the person controlling the debugger might not expect, such as upon a reset of the related package state or a change of the login user. To prevent continuing the execution using debugger-introduced modifications that might have been intended to be temporary, the ORACLE server needs to fully terminate the session. Cause: An attempt was made to grant the SELECT or READ privilege when the WITH HIERARCHY OPTION was specified on an object that is not one of the following: - Object table of non-final type - Object view of non-final type Cause: The user is trying to drop a type from being stored in a substitutable column or table. This error is raised due to one of the following reasons : - the column is enabled to store instances of some subtype of the type being dropped. - the column is enabled to store instances of only one type Cause: The user tried to create a column or table of an object type that contains a supertype attribute. This is not supported because it leads to infinite recursion in our current storage model. Note that creating a column of a type implies that we create columns corresponding to all subtype attributes as well. Cause: Either one of the following is the cause: - Checkout is requested for a resource that isn't under version control". Only version-controlled resource can be checked out. - Checkout is requested for a row of a non-versioned table. Cause: An attempt was made to perform a simple username and password-based authentication to the directory server using a non-existent or an invalid scheduler credential. This error could also occur if the credential existed as a synonym and there was insufficient privileges on the base credential pointed to by the synonym. Action: Reissue the command using a scheduler credential object that exists or create a new credential object and then reissue the command. Do not qualify the credential name with a schema name. The credential name should be owned by the current user invoking the DBMS_LDAP method. If the credential is owned by a different user, create a synonym pointing to the credential, request the database administrator to grant EXECUTE privilege on the credential and then reissue the command. Action: Check the values of subscription_name and publication_id. Check any other subscribe calls to see if they subscribe to columns that are shared among more than one publication on the same source table. Do not attempt to subscribe to the same publication more than once in the same subscription. Use the publication_id variant of the SUBSCRIBE call if needed to specify precise publications. Action: Check the input parameters on the call to SUBSCRIBE. Validate that the proposed source table has been published by checking the USER_PUBLICATIONS view for that source table. Contact the publisher if user privileges are required to access the publication. Retry the command with correct security or publication information. Cause: Failure during detach from a LogMiner session during advance of an asynchronous change set. This is an internal error. This exception is raised when a previous exception occurred during the internal protocol with LogMiner, after which Change Data Capture attempted to detach the LogMiner session as part of recovery. The detach session also failed Cause: There was an error originating from this synchronous change table. One possible cause is that schema redefinition has occurred on the source table and one or more columns in the change table are now a different type than corresponding source columns. Another possible cause is that there is a problem accessing the synchronous change table. Action: It is likely that this table existed before the import operation and import did not overwrite it. Determine whether the imported change table should supercede the original table. If so, drop the original table and retry the import. If not, the change table must have a different name in order to be imported. Cause: The source table was missing or was not set up correctly for an imported Change Data Capture change table. A synchronous change table requires the source table to exist and have the Change Data Capture trigger defined on it. An asynchronous change table requires table rules to be defined for the source table. Action: Drop the imported change table because it is invalid. Verify that the schema containing the source table was included in the original export. If needed, perform the export again, including the schema of the source table. If change table is asynchronous, ensure that STREAMS_CONFIGURATION=y is specified for the import. Action: It is likely that this view existed before the import operation and import did not overwrite it. Determine whether the imported subscriber view should supercede the original view. If so, drop the original view and retry the import. If not, the subscriber view must have a different name in order to be imported. Cause: All Change Data Capture (CDC) identifiers must have at most 30 characters, and must start with a letter that is followed by any combination of letters, numerals, and the signs '$', '_', and '#'. Other characters cannot be used in identifiers. The rules for CDC identifiers are the same as for PL/SQL identifiers. Cause: The necessary privileges are not available for operations such as: restarting a job on behalf of another owner, using a device as a member of the dump file set, or ommiting a directory object associated with any of the various output files. Refer to any following error messages for additional information. Action: Verify that the mode of the job specified objects to be moved. For command line clients, verify that the INCLUDE, EXCLUDE and CONTENT parameters were correctly set. For DBMS_DATAPUMP API users, verify that the metadata filters, data filters, and parameters that were supplied on the job were correctly set. Action: If this error occurs from a Data Pump client (e.g. expdp or impdp), it means that the master process for the operation had an unhandled exception. Check the log file for the operation and the trace file for the master process for any additional information. Try the operation again. If the error occurs again, contact Oracle Customer Support and report the error. Action: If this error occurs from a Data Pump client (e.g. expdp or impdp), it means that the master process for the operation died unexpectedly. Check the log file for the operation and the trace file for the master process for any additional information. Try the operation again. If the error occurs again, contact Oracle Customer Support and report the error. Action: Make sure that the number of workers value is greater than 0 and less than the maximum value for the platform (normally 32767). If this error occurs from a Data Pump client (e.g. expdp or impdp) and you specified the parallel parameter, try the operation again with a smaller value for the parallel parameter. If the error occurs again, contact Oracle Customer Support and report the error. Action: The query defining the materialized view must conform to a star or snowflake schema, the fact-table and materialized view must both be partitioned, and the partition-key of the materialized view must be functionally-dependent on the partition-key of the fact-table. Run CAN_SYNCREF_TABLE to obtain more diagnostics and troubleshoot the problem further. Action: Do not use the following ALTER TABLE statements on a table with a staging log: * ALTER TABLE EXCHANGE PARTITION * ALTER TABLE DROP PARTITION * ALTER TABLE TRUNCATE PARTITION You must use the REGISTER_PARTITION_OPERATION procedure in the DBMS_SYNC_REFRESH package to perform these operations. Cause: An attempt was made to perform a PMOP (partition maintenance operation) or otherwise modify a table that was locked by synchronous refresh. The synchronous refresh operation locks tables related to the refresh group at PREPARE_REFRESH time and unlocks them after the EXECUTE_REFRESH is done to prevent wrong results. Cause: An error was detected by the refresh statistics subsystem at the position indicated by the first argument in the error message. The secondary error is indicated by the second argument in the error message. This error message is recorded in the trace file and alert log but is not thrown and the refresh operation is not aborted because of this error. Cause: An error was detected while shrinking the materialized view (MV) log during refresh at the position indicated by the first argument in the error message. The secondary error is indicated by the second argument in the error message. This error message is recorded in the trace file and alert log but is not thrown and the refresh operation is not aborted because of this error. Cause: Improper use of WITH clause because one of the following two reasons: 1. nesting of WITH clause within WITH clause not supported yet 2. For a set query, WITH clause can't be specified for a branch. 3. WITH clause cannot be specified within parenthesis. Cause: The recursive component of the UNION ALL in a recursive WITH clause element used an operation that was currently not supported. The following should not be used in the recursive branch of the UNION ALL operation: GROUP BY, DISTINCT, MODEL, grouping sets, CONNECT BY, window functions, HAVING, aggregate functions. Cause: The recursive component of a recursive WITH clause element contained one of the following: - A LEFT OUTER join where the right side was the recursive query name. - A RIGHT OUTER join where the left side was the recursive query name. - A FULL OUTER join involving the recursive query name. Cause: The current varchar behavior is different. e.g. instr('abcd', 'cd', -2, 1) returns 3, whereas instr(to_clob('abcd'), 'cd', -2, 1) returns 0, (i.e. no match), because the reverse search starts from offset -2, which points to 'c' and moving backward, i.e. 'd' is ignored. This is symmetric to instr('dcba','dc',2,1), which returns 0.