Parameters and Alerts for Database System Check 

Database checks are usually executed with the database in the OPEN state. In critical situations, the database system check can check certain parameters while the database is in the MOUNT state. These parameters are described in detail in the following tables:
Type DBA:
Parameter |
Severity |
Check during mount state |
ARCHIVE_STUCK |
W |
Yes |
Cause: The amount of archiving directory space for offline redo log files used is above the specified threshold level. The system will crash if the database system runs out of room to archive online redo log files.
The operand > of the threshold value checks the amount of used space in the archiving directory. The operand < of the threshold value checks the amount of free space in the archiving directory. Mount state: Checks whether the percentage of free space in the archiving directory is less than 10%. Action: Archive the offline redo log files using SAPDBA (BRARCHIVE). | ||
CONTROL_FILE_MISSING |
E |
Yes |
Cause: The system cannot access at least one of the control files (we recommend that three be accessible). If the database system is still running, at least one copy is still online.Action: Look for the parameter CONTROL_FILES , in the init<SID>.ora initialization file to determine which directories the control files are stored in. Make sure you can display all the files using the file manager (command ls ). Restore access to any non-accessible files. | ||
CONTROL_MIRROR |
E |
Yes |
Cause: You have entered less than two control files in the init<SID>.ora initialization file. For safe operation, we recommend a double mirror of the original file. This message is often generated because the control file is mirrored on the hardware level, but this fact is not indicated in the init<SID>.ora initialization file. For safe operation, we recommend that you either use Oracle mirroring (multiple init<SID>.ora entries) or both hardware and Oracle mirroring.Action: Edit the init<SID>.ora profile and add the additional control files under the parameter CONTROL_FILES . We recommend you have at least three copies of the file. Each copy of the file must be stored on a separate physical hard disk. After you edit the profile, restart the database to activate your changes.
If you only use hardware mirroring, you can switch this check off by setting the parameter check_control_mirror in the init<SID>.sap SAPDBA initialization file to N . This overwrites the standard configuration (transaction DB17 ). | ||
CRITICAL_SEGS |
W |
No |
Cause: If the number of displayed extents are assigned to the displayed table(s), this would cause a tablespace overflow error.Action: | ||
DF_OFFLINE |
E |
Yes |
Cause: The system cannot access the data file specified in the message, because this file is offline. All files in a SAP database should normally be online.
Files that are offline will not be part of database processing nor part of a database restore. Action: Close all applications and enter the following commands using the Oracle Server Manager:
Alternatively, use the SAPDBA function, see
| ||
FILE_MISMATCH |
E |
Yes |
Cause: The file type for a database file specified in the control file does not match the entry in the Data Dictionary of the Oracle database system. This error is usually caused by errors during processing of the control file or during restoration of the control file.Action: Execute CREATE CONTROLFILE using the Oracle Server Manager , specifying either the correct file or no file at all. | ||
FILE_MISSING |
E |
Yes |
Cause: At least one data file is missing (file cannot be accessed). This error occurs if a file has the status Recovery and the file either cannot be accessed or the file is too old.Action: Execute the following SAPDBA functions in the following order (
Database Check and then
| ||
FILE_TYPE_UNKNOWN |
E |
Yes |
Cause: The file type (data file, raw device, dir, link) is unknown.Action: Troubleshoot to determine the cause of the problem. If you are not sure how to proceed, contact SAP technical support. | ||
FS_FULL |
W |
No |
Cause: The amount of SAPDATA directory space used is above the specified threshold.
The operand > of the threshold value checks the amount of used space in all SAPDATA directories. The operand < of the threshold value checks the amount of free space in all SAPDATA directories. Action: Increase the amount of space in the SAPDATA directories. | ||
MANY_EXTENTS |
W |
No |
Cause: The tablespace(s) named in the alert message is (are) reaching the MAXEXTENTS limit or must be reorganized for efficient operation.Action: Reorganize the tablespace(s) using SAPDBA. You could also temporarily increase the MAXEXTENTS limit for the tablespace(s) using SAPDBA. | ||
MISSING_INDEXES |
E |
No |
Cause: UNIQUE indexes for SAP tables are missing. SAP tables that by default do not have an index are listed in a table of exceptions. The SAPDBA database check takes this table into account.Action: Create the missing indexes using Transaction SE11. (Database table ® Indexes) | ||
NOARCHIVELOG |
E |
Yes |
Cause: The Oracle database system is not archiving any online redo log files. SAPDBA checks if the ARCHIVELOG mode and the automatic archiving of online redo log files are active.Action: Using SAPDBA, restart the database in ARCHIVELOG mode. Choose Archive mode ® Toggle database log mode. SAPDBA shuts down the database, reconfigures the database mode, and then restarts the database. | ||
NO_OPT_STATS |
E |
No |
Cause: Statistics for the cost-based optimizer are missing. SAP tables that by default do not have statistics are listed in the exception table DBDIFF. The SAPDBA database check takes this table into account.Action (immediate): Call Transaction DB20. Under Standard operations, run NO_OPT_STATS . Action (longterm): Schedule the action Check optimizer statistics in the DBA planning calendar. | ||
REDOLOG_MIRROR |
W |
Yes |
Cause: The online redo log files are not mirrored. Mirroring is required for safe operation. We recommend activating Oracle mirroring even if you are already using hardware mirroring.Action: Using the Oracle Server Manager, activate mirroring of the online redo logs in your database, with the statement ALTER DATABASE ADD LOGFILE .
If you do not want to use Oracle mirroring, you can switch this check off by setting the parameter check_control_mirror in the init<SID>.sap SAPDBA initialization file to N . This overwrites the standard configuration (transaction DB17 ). | ||
REDOLOG_MISSING |
E |
Yes |
Cause: During the test to determine whether online redo logs (<n> groups, <m> members per group) can be accessed, R/3 determined that one online redo log cannot be accessed. This problem can have multiple causes.Action: First, check the file that cannot be accessed, and check the system for problems (network problems, server down, and so on). One possible action (ONLY for Oracle experts, as an error could result in an incomplete chain of offline redo logs which could prevent a full recovery of the database): If a second member of the group that cannot be accessed, exists, delete the non-accessible member. Using the Oracle Server Manager, execute ALTER SYSTEM SWITCH LOGFILE . Then, check if the status of the online redo logs in the table V$logfile is correct. | ||
TABLES_NOT_IN_TABLE_TABLESPACE |
E |
No |
Cause: At least one of the tables is stored in a tablespace with type INDEX, PURE INDEX, ROLLBACK, PURE ROLLBACK, TEMP or PURE TEMP. This can become dangerous during certain database restoration actions (for example, for an ALL_DATA backup strategy) as these tablespaces can then no longer be easily restored.
A tablespace is of the type INDEX, ROLLBACK or TEMP, if more than 75% of its segments are index, rollback, or temporary segments. A tablespace is of the type PURE INDEX, PURE ROLLBACK or PURE TEMP, if 100% of its segments are index, rollback, or temporary segments. Action: Use SAPDBA to reorganize (Reorganize single table or index) the affected tables. This will move the tables into a tablespace of type TABLE or PURE TABLE. | ||
TSP_BACKUP_MODE |
E |
Yes |
Cause: The tablespace(s) (or their data files) listed in the message is (are) currently in backup mode. This reduces R/3 System performance. The check takes into account whether a BRBACKUP online backup is running.Action: Make sure no database backup (backup by BRBACKUP or an external tool) is running. Use the function Check database (online quick check) in SAPDBA to cancel the backup mode. SAPDBA will display a confirmation prompt for you to confirm you want to cancel the backup mode. | ||
TSP_FULL |
W |
No |
Cause: The tablespace(s) listed in the message is (are) more than 90% full and must be extended in order to prevent a possible overflow.Action: Extend the tablespace using SAPDBA and Tablespace Administration. | ||
TSP_OFFLINE |
E |
No |
Cause: The tablespace(s) named in the alert message is (are) offline. It (They) cannot be accessed.Action: Bring the corresponding tablespaces online by entering the command ALTER TABLESPACE <tablespace> ONLINE in the Oracle Server Manager. | ||
Type DBO:
Parameter |
Severity |
Check during mount state |
DBO ALY |
E |
No |
Cause: Errors occurred in sapdba-analyze sessions | ||
DBO ALY DBSTATCO |
E |
No |
Cause: The age of the last successful sapdba-analyze session for the object DBSTATCO is older than the specified threshold value (default setting: >10 days) | ||
DBO NXT |
E |
No |
Cause: The age of the last successful sapdba-next session is older than the specified threshold value (default setting: >10 days) | ||
DBO OPT |
E |
No |
Cause: Errors occurred in sapdba-checkopt sessions | ||
DBO OPT PSAP% |
E |
No |
Cause: The age of the last successful sapdba-checkopt session for the object PSAP% is older than the specified threshold value (default setting: >10 days) | ||
DBO RSI |
E |
No |
Cause: There are errors in SAPDBA reorganizations of an individual table or an individual index. | ||
DBO RTC |
E |
No |
Cause: There are errors in SAPDBA reorganizations of a tablespace without data files. | ||
DBO RTD |
E |
No |
Cause: There are errors in SAPDBA reorganizations of a tablespace with data files. | ||
Action for all DBO alert messages:
Check the schedule in the DBA Planning Calendar (Transaction DB13 ) and the operation logs in the DB Operations Monitor (Transaction DB24 ).Type ORA:
Parameter |
Severity |
Check during mount state |
ORA -272 error writing archive log |
E |
Yes |
Cause: Error writing to archive logAction: Check to determine whether the storage medium is functioning as well as whether enough storage space is available. | ||
ORA -376 file <name> cannot be read at this time |
E |
Yes |
Cause: The most likely cause is that the file is offline.Action: Use SAPDBA (function Check database) to check the status of the file. Bring the file online if necessary. Contact SAP technical support for additional help, as there are multiple potential causes for this problem. | ||
ORA -600 internal error code, arguments: [num], [?], [?], [?], [?], [?] |
E |
Yes |
Cause: This is a general internal error for Oracle program exceptions. It indicates that a process has encountered a low-level, unexpected condition. This message can be caused by: timeouts, incorrect files, incorrect data checks in memory, hardware or I/O error, or incorrectly recovered files.Action: Report this message to Oracle’s customer support service or to the SAP R/3 hotline. | ||
ORA –1113 file <name> needs media recovery |
E |
Yes |
Cause: The database system attempted to open a data file that requires media recovery. The system could not access the file due to a physical hard disk drive error.Action: You must recover the database. However, choosing a recovery strategy requires a careful analysis of the failure. We recommend you contact SAP technical support for help with this problem. | ||
ORA –1115 IO error reading block from file <name> block <#> num |
E |
Yes |
Cause: The disk where the file is stored is probably offline.Action: Bring the disk online. Then, execute the operation again. If you are unable to restore access to the file or are unsure how to proceed, we recommend you contact SAP technical support for help with this problem. | ||
ORA –1122 database file <name> failed verification check |
E |
No |
Cause: The information in the datafile is not consistent with the information in the control file.Action: Make sure you are using the correct data file and control file for the database. | ||
ORA –1135 file name accessed for DML query is offline |
E |
Yes |
Cause: A query failed because it attempted to access a data file belonging to an offline tablespace. An offline tablespace must be brought online to access its data.Action: Check the Oracle alert file in the directory displayed in the detail view of the database system check (Transaction DB16). Other Oracle messages may provide additional information about the problem. If you are unsure how to proceed, contact SAP technical support. | ||
ORA –1555 snapshot too old rollback segment number %s with <name> too small |
W |
No |
Cause: Rollback segments required for consistent reading were overwritten by other database users.Action: Change the configuration of the rollback segments. Only execute this operation when the database load is low. For more information, see the SAPNet notes. | ||
ORA –1562 failed to extend rollback segment number <n> |
W |
Yes |
Cause: The rollback segment named in the alert message could not be extended, as the tablespace PSAPROLL is full.Action: Use SAPDBA (Tablespace extension), to extend the tablespace PSAPROLL . | ||
ORA –1578 Oracle data block corrupted (file <name>, block <number> |
E |
Yes |
Cause: The database system attempted to read a data file that contains corrupted data.Action: Recover the database. Choosing a recovery strategy requires a careful analysis of the failure. We recommend you contact SAP technical support for help with this problem. | ||
ORA –3113 end-of-file on communication channel |
E |
No |
Cause: An unexpected end-of-file was processed on the communication channel. The problem could not be handled by the Oracle Net8, two task, software. This message could occur if the shadow two-task process associated with a Net8 connection has terminated abnormally, or if there is a physical failure (network or server offline).Action: If this message occurs during a connection attempt, check the setup files for the appropriate Oracle Net8 driver, and confirm Net8 software is correctly installed on the server. If the message occurs after a connection is established, and the error is not due to a physical failure, check if a trace file was generated on the server at the time of the failure. | ||
Checkpoint not complete |
E |
No |
Cause: After a log switch, data confirmed by COMMIT is written from the database buffer to the datafiles (synchronization). Oracle writes the alert message to the alert log, if the write session is still active at the next log switch.Action: If this alert message occurs often, there is a database performance problem. To correct the problem, enlarge the online redo log files. For more information, see the Oracle documentation. | ||
Type PROF:
Parameter |
Severity |
Check during mount state |
CONTROL_FILE_RECORD_KEEP_TIME |
E |
No |
The amount of time (days) reusable areas in the control file are locked. Default setting: < 30 | ||
CURSOR_SPACE_FOR_TIME |
W |
No |
The setting, CURSOR_SPACE_FOR_TIME = FALSE, requires a considerably less amount of space in the Oracle shared pool. Default setting: <> FALSE | ||
DBWR_IO_SLAVES |
A |
No |
Number of IO slaves used by the DBWR process. Default setting: <> 0 | ||
DB_BLOCK_BUFFERS |
W |
No |
Number of blocks in the database buffer cache. Default setting: < 8960 | ||
DB_BLOCK_CHECKPOINT_BATCH |
W |
No |
Number of buffers the DBWR process writes per write session during a checkpoint. Default setting: < 8 | ||
DB_BLOCK_SIZE |
W |
No |
Size (bytes) of the Oracle database blocks. Default setting: <> 8192 | ||
DB_FILES |
W |
No |
Number of data files Default setting: < 254 | ||
DB_FILE_MULTIBLOCK_READ_COUNT |
W |
No |
Number of blocks read during a complete tablescan I/O operation. Default setting: > 8 | ||
DISK_ASYNCH_IO |
W |
No |
The setting, DISK_ASYNCH_IO = TRUE, means that I/O operations are asynchronous for data files, control files and log files. Default setting: <> TRUE | ||
HASH_JOIN_ENABLED |
W |
No |
The setting, HASH_JOIN_ENABLED = FALSE, means that the optimizer does not use any hash joins when calculating the access plan. Default setting: <> FALSE | ||
LOG_ARCHIVE_START |
E |
No |
The setting LOG_ARCHIVE_START = TRUE means that automatic archiving is activated when the Oracle instance is started. Default setting: <> TRUE | ||
LOG_BUFFER |
W |
No |
Size (KB) of the redo log buffer in the SGA. Default setting: >< 1100,300 | ||
LOG_CHECKPOINT_INTERVAL |
W |
No |
Number of the redo log blocks that trigger a new checkpoint. The default setting is extremely high, to ensure that a checkpoint is only triggered by a redo log switch. Default setting: < 3000000000 | ||
LOG_CHECKPOINT_TIMEOUT |
W |
No |
Time (seconds) between two checkpoints. Default setting: > 0 | ||
LOG_SMALL_ENTRY_MAX_SIZE |
W |
No |
Maximum size (bytes) of a copy in the redo log buffer that can be created using the redo allocation latch (without redo buffer copy latch). Default setting: <> 120 | ||
OPEN_CURSORS |
W |
No |
Maximum number of open cursors per user process. Default setting: >< 2000,800 | ||
OPTIMIZER_FEATURES_ENABLED |
W |
No |
Oracle release indicating which features of the optimizer are active. Default setting: <> 8.0.4 | ||
OPTIMIZER_INDEX_COST_ADJ |
W |
No |
Ratio (%) that influences the analysis of index scans Default setting: <> 10 | ||
OPTIMIZER_MODE |
W |
No |
Controls how the optimizer behaves (for example, rule-based = RULE, or cost-based = CHOOSE). Default setting: <> CHOOSE | ||
ROW_CACHE_CURSORS |
W |
No |
Number of recursively-buffered cursors used by the row cache manager. Default setting: >< 1500,270 | ||
SHARED_POOL_SIZE |
W |
No |
Size (MB) of the shared pool in the SGA. Default setting: < 50 | ||
SORT_AREA_RETAINED_SIZE |
E |
No |
Size (KB) of the sort area in the PGA after a sort. This amount is the sort result. Default setting: >< 540,0 | ||
SORT_AREA_SIZE |
E |
No |
Size (KB) of the sort area in the PGA. Default setting: >< 10240,2048 | ||
TIMED_STATISTICS |
E |
No |
The setting, TIMED_STATISTICS = TRUE, creates time-related statistics. Default setting: <> TRUE | ||
TRANSACTION_AUDITING |
E |
No |
The setting, TRANSACTION_AUDITING = FALSE, means no special REDO record (with session and user information) is created by the transaction layer. Default setting: <> FALSE | ||
Action for all PROF alert messages:
Adjust the Oracle init<SID>.ora initialization file according to the PROF parameters types of the database system check (Transaction DB17).

SAPDBA stores comprehensive information about every database operation in a log. To display this log, use the