DBS Control parameters for Teradata Database 14.00

The following article provides a brief introduction to the DBSControl parameters for TD14.0

When performing a software upgrade or a migration, match the DBSControl settings on the new version
to the old version and keep them until system is stabilized. For new parameters, use the default. Only make changes as needed to address a problem on the system.

The Utilities: Volume 1 Manual (B035-1102-111A) describes all these fields in more detail. Please review as needed.

The Teradata Performance Management Manual (B035-1097-109A) has explanations for some of the performance fields along with suggested approaches for tuning. Please review as needed.
Note that making site specific tuning recommendations is a PS Service engagement; primarily because it entails evaluation and monitoring. Based on experience, DBS Control tuning rarely results in more than a 1-5% improvement in throughput and little or no impact  on response time; the better alternative is workload management which usually results in much more improvement in both areas.

_________________________________________________________________________________


DBS Control Record - General Fields:

1.  Version - This field is the version number of the DBS Control
    Record. It is incremented by 1 when it becomes necessary to
    migrate the DBS Control Record to a new format. This field
    cannot be modified by the DBSControl utility program.
    Recommended setting: ==> Cannot be modified

2.  SysInit - Upon successful completion, the SYSINIT utility program
    will set this field to TRUE. The default value is FALSE.
    When SysInit flag is TRUE, sysinit time (yyyy-mm-dd hh:mi)
    is displayed if available. This field cannot be modified by
    the DBSControl utility program.
    Recommended setting: ==> Cannot be modified

3.  DeadLockTimeOut - This field is used by the Dispatcher to
    determine the interval, in seconds, between deadlock time-
    out detection cycles.  The default value is 240.
    Default: ==> 240
    Reference Article: ==> SD1000CA92E
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> YES

4.  (Reserved for future use)

5.  HashFuncDBC - This field defines the hashing function that will
    be used by the DBS. The default value is 6 (MSCHASH_P90_UNI_5R).
    This field cannot be modified by the DBSControl utility program.
    Recommended setting: ==> Cannot be modified

6.  (Reserved for future use)

7.  (Reserved for future use)

8.  SessionMode - This field defines the system default transaction
    mode, case sensitivity, and character truncation rule for a session.
    The setting is either 0 (Teradata) or 1 (ANSI). The default is Teradata.
    The ANSI setting defaults SQL sessions to  ANSI transaction semantics,
    case sensitive data, and error reporting on truncation of character
    data. The Teradata setting defaults SQL sessions to Teradata
    transaction semantics, case insensitive data, and no error reporting
    on truncation of character data.
    Default: ==> 0 (Teradata)
    Reference Article: ==> KAP1AC242
    Recommended setting: ==> Customer preference
    Tpareset Required? ==> YES

9.  LockLogger - This Field defines the system default for the
    locking logger. This allows the DBA to log the delays caused by
    database locks to help in identifying lock conflicts. To enable
    this feature set the field to TRUE. To disable the feature set the
    field to FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP1AC24A
    Recommended setting: ==> FALSE (see Reference Article)
    Tpareset Required? ==> YES

10. RollbackPriority - This Field defines the system default for the
    rollback priority. This allows the DBA to either take the default
    priority of Rush or the user's logon priority. To enable this feature
    set the field to TRUE. To disable the feature set the field to FALSE.
    Note that this new feature will affect all users on the system.
    Default: ==> FALSE
    Reference Article: ==> S11000B2416
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> YES

11. MaxLoadTasks - Maximum combined number of FastLoad, MultiLoad,
    and possibly FastExport. This field works differently depending on
    the value of MaxLoadAWT.
    a) If the field MaxLoadAWT is zero (by default), this field is
       the concurrent job limit for all three utilities: FastLoad, MultiLoad,
       and FastExport.
    b) If the field MaxLoadAWT is changed to non-zero, this field is
       the concurrent job limit for only two utilities: FastLoad and MultiLoad.

    The valid range is 0 .. 30.  The default is 5.                    
    NOTE: This field can only be set to greater than 15 if the field
          MaxLoadAWT is changed from the default of zero.
    Default: ==> 5
    Reference Article: ==> KAP1AC256
    Recommended setting: ==> Customer tunable (see Reference Article)
    Tpareset Required? ==> NO

12. RollForwardLock - This field defines the system default for the
    RollForward Using Row Hash Locks option.  This allows the DBA to
    specify that row hash locks should be used to lock the target
    table rows during a RollForward.  To enable this feature set the
    field to TRUE.  To disable the feature set the field to FALSE.
    Note that this new feature will affect all users on the system.
    Default: ==> FALSE
    Reference Article: ==> S11000D704E
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> YES

13. MaxDecimal - Maximum number of Decimal Digits in the default
    max value used in expression typing. 0 is system default, because
    of initialization of SysDBSControl, and means use 15. Other valid
    values are presently 15, 18, and 38 .
    Default: ==> 15
    Reference Article: ==> S11000E5936
    Recommended setting: ==> Customer tunable (see Reference Article)
    Tpareset Required? ==> YES

14. CenturyBreak - This field defines which Teradata Dates are
    specific to the 21st Century.  For example, C = 25, strings
    inserted such as 00/01/01 and 24/01/01 will assume the years
    2000 and 2024 respectively.  A string inserted as 25/01/01
    will assume the year 1925.  Valid values range from 0 to 100.
    Default: ==> 0 (15)
    Reference Article: ==> SAN10006C0FA
    Recommended setting: ==> Customer tunable (see Reference Article)
    Tpareset Required? ==> NO

15. DateForm - This field defines whether IntegerDate (0)
    or ANSIDate (1) is used for a session.
    Default: ==> (0) IntegerDate
    Reference Article: ==> C10000C42F6
    Recommended setting: ==> Customer preference (see Reference Article)
    Tpareset Required? ==> No

16. System Time Zone Hour - This field defines the System TimeZone
    Hour offset from UTC.  Permitted range: -12 to 14.
    Default: ==> 0

        Reference Article: ==>N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

17. System Time Zone Minute - This field defines the System Time Zone
    Minute offset from UTC.  Permitted range: -59 to 59.
    Default: ==> 0
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

18. System Time Zone String - This field displays the System TimeZone String.

    It is set up using the tdlocaledef utility and any changes

    will have to be done through the same.
    Default: ==> 0
    Reference Article: ==> KAP3BD24E

    Recommended setting: ==> Customer tunable (see Reference Articles)
    Tpareset Required? ==> NO

19. (Reserved for future use) 
                   
20. RoundHalfwayMagUp - This field indicates how rounding should
    be performed when computing values of DECIMAL type.  If this
    field is FALSE, the system uses the rounding semantics
    traditional for Teradata, namely, a value exactly halfway
    between representable decimal values is rounded up or down
    so that the least significant digit is even.  If TRUE, the
    system uses rounding semantics appropriate for many business
    applications, namely, the magnitude of halfway values are
    rounded up.  Put differently, halfway values are rounded
    away from zero, so that positive halfway values are rounded
    up and negative halfway values are rounded down (toward
    negative infinity).
    Default: ==> FALSE
    Reference Article: ==> S11000D82EE
    Recommended setting: ==> Customer preference (see Reference Article)
    Tpareset Required? ==> YES

21. (Reserved for future use) 

22. Target Level Emulation - Used when a user wants to emulate a Target Machine
    a Target Level Machine.  
    TRUE means the user wants to run Target Level Emulation.      
    FALSE means the user does NOT.  The default is FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP1AC25A
    Recommended setting: ==> FALSE (see Reference Article)
    Tpareset Required? ==> NO

23. Export Width Table ID - The Export Width Table ID.
    This is the Export length adjustment required for inclusion  
    of SO and SI characters.  'Expected Defaults' is an expected
    width adjustment of 2.  'Compatability Defaults' is an  expected
    width adjustment of 0 and 'Maximum Defaults' is a expected    
    width adjustment of 1.
    For Expected Defaults, enter      : 0
    For Compatibility Defaults, enter : 1
    For Maximum Defaults, enter       : 2
    Default: ==> 0
    Reference Article: ==> S11000D8346
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> YES

24. DBQL Log Last Resp - This field indicates whether DBQL
    should generate a pseudo RESP step for each request that is logged
    to the DBC.DBQLogTbl (not requests logged to DBQLSummaryTbl).
    The pseudo RESP step is logged to the DBQLStepTbl after the last
    response is sent to the client/host and the request is complete.
    This allows an administrator to know the actual response time
    seen by the client/host.
    TRUE means the user wants the RESP step logged.      
    FALSE means the user does NOT.  The default is FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP1AC266
    Recommended setting: ==> Customer preference (see Reference Article)
    Tpareset Required? ==> NO
 
25. DBQL Options - indicates special options of DBQL,
    add values together to set applicable bit flags
    0: No special options  
    1: Long QueryId: No longer used, value does not matter and flag
        will not be reused for another purpose. TD12.0 and later
        always use a large queryid.
    2: DBQL_QS_OPT: Enable diagnostics for timestamp and
       summary queryseconds issues
       bit2 on:  DBQL_QS diagnostics enabled
       bit2 off: DBQL_QS diagnostics disabled
       Change to DBQL_QS_OPT bit2 is effective after write.
    Default: ==> 0
    Reference Article: ==> N/A
    Recommended setting: ==> Default (This field is obsolete as of TD12)
    Tpareset Required? ==> N/A

26. External Authentication - This field indicates whether external
    authentication is enabled. The valid values are 0, 1 and 2.
    0: Accept both external authentication and traditional logons.
    1: Reject external authentication and accept traditional logons.
    2: Accept external authentication and reject traditional logons.
    The default value is 0.
    Default: ==> 0 (ON)
    Reference Article: ==> KAP1AC26A
    Recommended setting: ==> Customer tunable
    Tpareset Required? ==> NO

27. IdColBatchSize - Indicates the size of the pool of
    numbers to be reserved for generating numbers for a
    batch of rows to be bulk-inserted into a table with
    an identity column. The valid range of value is
    1 .. 1000000. The default value is 100000.
    Default: ==> 100000
    Reference Article: ==> S11000D83E2
    Recommended setting: ==> Default (Customer tunable, see Reference Article)
    Tpareset Required? ==> NO

28. LockLogger Delay Filter - Indicate whether locking
    logger delay filter is ON. If it is ON, the delay filter
     time in field 29 will take effect. The default is OFF.
    Default: ==> FALSE (OFF)
    Reference Article: ==> SD1000CA97A
    Recommended setting: ==> FALSE (Enable as needed, see Reference Article)
    Tpareset Required? ==> NO

29. LockLogger Delay Filter Time - Indicates the delay filter
    time used by locking logger. Only blocked lock request with
    delay time greater than this value will be logger. The valid
    range is 0 .. 1000000 seconds, default is 0 second.
    Default: ==> 0 (seconds)
    Reference Article: ==> SD1000CA97A
    Recommended setting: ==> 1 to 10 seconds when enabled (see Reference Article)
    Tpareset Required? ==> No

30. ObjectUseCountCollectRate - Determines the amount of time in minutes
    after which the system collects use count data and updates the Data
    Dictionary columns AccessCount and LastAccessTimeStamp. The default
    value is 0, which disables the Database Object Use Count feature.
    The minimum recommended value is 10 minutes.
    Default: ==> 0 (Disabled)
    Reference Article: ==> KAP1AC26E
    Recommended setting: ==> If enabled, minimum recommended value is 10 minutes.
    Tpareset Required? ==> NO

31. LockLogSegmentSize - Indicates the size of the
    locking logger segment used by the lock manager in kilo bytes.
    The minimum value is 64 KB and maximum value is 1024 KB.
    Default: ==> 64KB
    Reference Article: ==>SD1000CA97A
    Recommended setting: ==> 64KB (see Reference Article)
    Tpareset Required? ==> YES

32. CostProfileId. - This value is used to select
    the cost profile from DBC.CostProfiles that will be used to
    create cost prediction method parameter values.  The profile
    with DBC.CostProfiles.ProfileId == CostProfileId will be chosen.
    If CostProfileId doesn't match any ProfileId value, then the
    system default profile (DBC.CostProfiles.ProfileId == 0) is used.
    If the system default profile cannot be read then the bootstrap
    profile is used.
    Default: ==> 0
    Reference Article: ==> IDA00108C7A
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

33. DBQLFlushRate - Determines the frequency for writing DBQL
    cache entries to DBQL dictionary tables.  Default is 600
    seconds (10 minutes).  Valid range is 1 to 3600 seconds.
    The recommended value is 600 seconds or more.
    Default: ==> 600 seconds
    Reference Article: ==> C10000F849A
    Recommended setting: ==> 600 seconds or more
    Tpareset Required? ==> NO

34. Memory Limit Per Transaction - Specifies the maximum amount
    (number of pages) of in-memory temporary storage that can be used
    by the RSG to store the records for one transaction.
    If the transaction exhausts this amount, then it is moved to
    a disk file (specified by general field number 37).
    The default value is 2 pages. The minimum value is 0 page
    and maximum value is 127 pages.
    Default: ==> 2 pages
    Reference Article: ==> S11000D82E6
    Recommended setting: ==> Default (Customer tunable, see Reference Article)
    Tpareset Required? ==> YES

35. Client Reset Timeout - Specifies how long the RSG should wait
    for an intermediary to reconnect after a communication failure,
    an intermediary reset, or a server reset before taking the  
    needed actions.  The default value is 300 seconds.
    The minimum value is 0 sec and maximum value is 65535 sec.
    Default: ==> 300 seconds
    Reference Article: ==> S11000D82E6
    Recommended setting: ==> Default (Customer tunable, see Reference Article)
    Tpareset Required? ==> YES

36. Temporary Storage Page Size - Specifies the size (in KB) of a memory
    page that is used to store data rows of a replicated
    transaction. The default value is 4 KB.
    The minimum value is 1 KB and maximum value is 1024 KB.
    Default: ==> 4KB
    Reference Article: ==> S11000D82E6
    Recommended setting: ==> Default (Customer tunable, see Reference Article)
    Tpareset Required? ==> YES

37. Spill File Path - Specifies a directory that will be used  
    by the RSG for spill files.  
    Default: ==> MPRAS = /var/tdrsg; Linux = /opt/teradata/tdat/temp/tdrsg
    Reference Article: ==> S11000D82E6
    Recommended setting: ==> Default (Customer preference, any valid path)
    Tpareset Required? ==> YES

38. MDS Is Enabled -  Assumed that the MDS packages are installed,
    the MDS will be running if this flag is TRUE. The default is FALSE.
    Default: ==> FALSE
    Reference Article: ==> S11000D82E6
    Recommended setting: ==> Customer preference (see Reference Article)
    Tpareset Required? ==> YES

39. ChecktableTableLockRetryLimit (Default = 0 is retry forever).
    When a table is locked, checktable will retry until it gets the table.
    If the retry limit is set, checktable will retry within the specified limit.
    Retry limit range = 0 to 32767 minutes
    Default: ==> 0
    Reference Article: ==> KAP1AC272
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

40. EnableCostProfileTLE - This boolean determines whether new Optimizer
    Cost Profile System (OCES) diagnostics are enabled in combination with
    Target Level Emulation (TLE). This is initialized to FALSE. Attempts to
    use combined OCES and TLE diagnostic syntax will get a syntax error.
    When set to TRUE, combined OCES/TLE diagnostic syntax is enabled. When
    enabled all TLE diagnostics are also processed by the new OCES logic.
    If EnableSetCostProfile == 0, then this field will be set FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP1ACD8E
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

41. EnableSetCostProfile - This controls usage of DIAGNOSTIC SET PROFILE...
    statements to dynamically change cost profiles used for query planning.
    Meaningful values are 0, 1, and 2.  If this == 0, then the statements
    are disabled and will produce syntax error messages.  If this == 1, then
    profiles can be activated only at SESSION or REQUEST level.  If this == 2
    then profiles can be activated at any level.  The default value is 0, and
    the legal range is 0..2.
    Default: ==> 0
    Reference Article: ==> KAP1ACD96
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

42. (Reserved for future use)
       
43. DisableUDTImplCastForSysFuncOp - This field disables/enables
    implicit cast/conversion of UDT expressions passed to system
    operators/functions. Conversions are from UDTs to compatible
    predefined types. A value of TRUE disables implicit conversions.
    A value of FALSE enables. The default is FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP1ACD9A
    Recommended setting: ==> Default (Customer preference)
    Tpareset Required? ==> NO

44. CurHashBucketSize - This field defines the number of
    bits currently used for the hash bucket size.
    The valid values are 16 or 20, which correspond to the
    number of hash buckets in the system as follows:
    16: The system is configured for 16-bit hash buckets.
        The system has 2^16 = 65,536 (64K) hash buckets.
    20: The system is configured for 20-bit hash buckets.
        The system has 2^20 = 1,048,576 (1M) hash buckets.
    The default value is 20.
    Recommended setting: ==> This field is informational only. It cannot be changed.
   
45. NewHashBucketSize - This field defines the number of
    bits which will be used for the hash bucket size in
    the system after the next reconfig or sysinit.
    The valid values are 16 or 20, which correspond to the
    number of hash buckets in the system as follows:
    16: The system will use 16-bit hash buckets.
        The system will have 2^16 = 65,536 (64K) hash buckets.
    20: The system will use 20-bit hash buckets.
        The system will have 2^20 = 1,048,576 (1M) hash buckets.
    The default value is 20.
    Default: ==> 20 for new installations; 16 for upgrades from pre-TD12
    Reference Article: ==> KAP1ACD9E
    Recommended setting: ==> 20
    Tpareset Required? ==> Changes take effect after next sysinit or reconfig

46. MaxLoadAWT - This field defines the maximum number of
    AMP worker tasks (AWT) that concurrent FastLoad and MultiLoad
    can use altogether. The valid range is 0 .. (60% of maxampworkertasks).

    If the value is zero, this AWT limit check is not performed.
    The default is zero. This field can only be set to zero if the
    MaxLoadTasks is less than or equal to 15.        

    This field also affects the behavior of the MaxLoadTasks field.
    If it is zero (by default), the MaxLoadTasks field is the concurrent
    job limit for all three utilities: FastLoad, MultiLoad, and FastExport.
    If it is non-zero, the MaxLoadTasks field is the concurrent job limit
    for only two utilities: FastLoad and MultiLoad.
    Default: ==> 0
    Reference Article: ==> SD1000CA7A2
    Recommended setting: ==> Customer tunable (see Reference Article)
    Tpareset Required? ==> NO

47. MaxRowHashBlocksPercent - This field defines the maximum percentage
    of an AMP Lock Table's control blocks that a transaction can utilize
    for its rowhash locks. The transaction would be aborted if its rowhash
    locks exceed this threshold on any AMP. The default value for
    MaxRowHashBlocksPercent is 50. A value of 0 indicates that the system
    will use the default value.
    Default: ==> 50%
    Reference Article: ==> KAP1ACDA2
    Recommended setting: ==> Customer tunable (see Reference Article)
    Tpareset Required? ==> NO

48. MonSesCPUNormalization - This field disables/enables
    normalization of CPU values in Monitor Session response.
    A value of FALSE disables normalization of CPU values.
    A value of TRUE enables the normalization. The default is FALSE.

    Recommended procedure to change this flag to avoid inconsistent TDWM exception
    behavior:
    1) Disable TDWM Category 3 if enabled. Adjust CPU related exception thresholds.
    2) Set Monitor Session Rate to zero.
    3) Change MonSesCPUNormalization flag.
    4) Set Monitor Session Rate to previous non-zero value.
    5) Enable TDWM Category 3 if it was previously enabled.
    Default: ==> FALSE
    Reference Article: ==> KAP1ACDA6
    Recommended setting: ==> Default (Customer preference - see Reference Article)
    Tpareset Required? ==> NO

49. TempLargePageSize - This field specifies another method of RSG
    storage memory pool in addition to the existing temporary storage page
    size. If a transaction contains any record that exceeds the small page
    size, then RSG will allocate memory from the large page size. If that
    is also not sufficient enough to hold a transaction, then the entire
    transaction is spilled to disk. The default value for
    TempLargePageSize is 64K Bytes. A value of 0 indicates that the system
    will use the default value.
    Default: ==> 64K
    Reference Article: ==> KAP1ACDAA
    Recommended setting: ==> Default (Customer tunable - see Reference Article)
    Tpareset Required? ==> YES

50. RepCacheSegSize - This field specifies the size of a cache segment
    (in kilobytes) that is allocated in each AMP to store EVLobjects,
    which consist of executable code and related metadata that is used
    by the replication system to convert row images into the external
    format. The default value for RepCacheSegSize is 512K Bytes. A value
    of 0 indicates that the system will use the default value.
    Default: ==> 512KB
    Reference Article: ==> KAP1ACDAE
    Recommended setting: ==> Default (Customer tunable - see Reference Article)
    Tpareset Required? ==> YES

51. MaxDownRegions - Sets the maximum allowable number of down regions for a
    subtable.  The valid range is 0 to 12.  When this value is set to
    zero, a subtable with fatal file system error will be forced to mark down
    on all the AMPs, if it is possible.  Same action applies when this value is
    exceeded.  Other than those the subtable will mark with down region.
    The maximum value is 12 and the default value is 6.
    Default: ==> 6
    Reference Article: ==> KAP1ACDB2
    Recommended setting: ==> Default (Customer tunable - see Reference Article)
    Tpareset Required? ==> YES

52. MPS_IncludePEOnlyNodes - This flag is used to include PE-only nodes in
    Monitor Physical Summary PM/API response calculation.
    The default is FALSE, i.e., PE-only nodes, nodes configured with zero AMPs, are
    excluded from Monitor Physical Summary response calculation by default.
    Default: ==> FALSE
    Reference Article: ==> KAP1ACDB6
    Recommended setting: ==> Default (Customer preference - see Reference Article)
    Tpareset Required? ==> NO

53. PrimaryIndexDefault - This field determines whether a primary index is
    automatically created for tables whose definitions do not explicitly
    include the PRIMARY INDEX, NO PRIMARY INDEX, PRIMARY KEY, or UNIQUE clauses.
    When this value is set to P or D, the first column of such tables is
    automatically designated a nonunique primary index (NUPI). When this
    value is set to N, these tables are created without primary indexes, as
    NoPI tables.
    Default: ==> D
    Reference Article: ==> KAP1ACDDE
    Recommended setting: ==> Default (Customer preference - see Reference Article)
    Tpareset Required? ==> NO

54. AccessLockForUncomRead - This field, when enabled, imposes          
    access lock on source table(s) of DML statement when              
    transaction isolation level is Read Uncommitted (RU).              
    This flag only affects DMLs that involve data modification.        
    The default value is FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP1ACDE2
    Recommended setting: ==> Default (Customer preference - see Reference Article)
    Tpareset Required? ==> NO

55. EnabNonTempoOp - This field is to enable
    nontemporal operations such as NONTEMPORAL DELETE,
    NONTEMPORAL UPDATE, NONTEMPORAL INSERT, and
    NONTEMPORAL ALTER TABLE. A nontemporal operation is
    a DML activity that either causes rows to be deleted
    or the TransactionTime column values to be set
    differently. To perform such operations, the site
    must first set the flag to TRUE and then submit the
    SQL statements. The default value for this flag is
    FALSE (0). This flag can be enabled only if

    the system is licensed to use Temporal Tables.

    Default: ==> FALSE
    Reference Article: ==> KAP1B42BA
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

     
56. InclNTforGrntOrRevokAll - This field is to include
    the nontemporal privilege in GRANT ALL, REVOKE ALL
    statements. If this field is false then GRANT ALL
    or REVOKE ALL will not include the NONTEMPORAL    
    privilege. The default value is false.
    This flag can be enabled only if

    the system is licensed to use Temporal Tables.

    Default: ==> FALSE
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

     
57. TimeDateWZControl - This field, when enabled, the
    time and time zone values returned by the built-in
    functions CURRENT_TIME, CURRENT_TIMESTAMP, and
    CURRENT_DATE reflect the session time and session
    time zone. When this field is disabled, the returned
    time value reflects the local time of the Teradata
    Database server, however the returned time zone value
    reflects the session time zone. This field will be disabled

    by default after upgrade or sysinit, if the system timezone

    settings are zero and enabled with 2 for non-zero settings.

    Valid values for this field are 1 (disabled),

    2 (enabled with UTC i.e. all datetime data

    is stored in UTC on disk), 3 (enabled with LOCAL i.e.

    datetime data without time zone is stored in system LOCAL

    time on disk) and 0 (returns the field to disabled).

    To perform operations on temporal tables, this field

    must be enabled.

    Default: ==> 0 (after upgrade), 2 (after sysinit)
    Reference Article: ==> KAP3BF672
    Recommended setting: ==> Default
    Tpareset Required? ==> NO
 
   
58. (Reserved for future use)

59. SysinitRelease - This field displays the Major TDBMS Base Release on
    which a Sysinit was last performed.
    Recommended setting: ==> Cannot be modified

60. DefaultCaseSpec - This field defaults character                    
    comparisons to case specific in Teradata mode.                    
    It also defaults the char column to case specific                  
    when a table is created                                            
    The default value is FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP1A8522
    Recommended setting: ==> Customer preference (see Reference Article)
    Tpareset Required? ==> YES
                                       
   
61. PMPC_TimeoutSecs - This field specifies timeout interval in seconds
    for certain supported PM API commands. The default value is  60 seconds.
    Valid range is 0 - 3600. Zero value disables timeout, i.e., the response for
    PM API request is returned only upon the completion of its processing.
    Default: ==> 60 seconds
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

   
62. ExportOrderBySegmentLimit - This field specifies the maximum number
    of segments can be allocated for FastExport's redistribution
    when data is to be sorted.
    The valid range is 500 to 2000. The default value is 500.
    Default: ==> 500
    Reference Article: ==> N/A
    Recommended setting: ==> Increase the value of this setting if a large, complex FastExport job generates error 2587
    Tpareset Required? ==> NO

   
63. MLoadDiscardDupRowUPI - When IGNORE DUPLICATE ROWS is specified,
    this flag indicates whether the DBS MultiLoad protocol should
    silently discard rows with UPI violations or log these rows to
    the application error (or UV) table.
    The valid values are TRUE and FALSE. The default is FALSE.
    FALSE:   When IGNORE DUPLICATE ROWS is used, MultiLoad will log rows
             with UPI violations to the application error (or UV) table.
    TRUE:    When IGNORE DUPLICATE ROWS is used, MultiLoad will
             silently discard rows with UPI violations. These rows
             are not logged to application error (or UV) table.
    Default: ==> FALSE
    Reference Article: ==> NTA 2237
    Recommended setting: ==> Customer preference
    Tpareset Required? ==> NO

   
64. DBQL CPU/IO Collection - This option modifies the manner in        
    which DBQL collects CPU/IO statistics. It ranges from the          
    V2R6 value of one (1) to its maximum allowed value. Higher values  
    provide increasingly refined algorithms for collecting CPU/IO      
    usage for steps and requests logged by DBQL. Each level enhances    
    or replaces the levels below it.                                    
    The default value of zero (0) (i.e. a value has not been set)      
    means use the default value for the release which is initially      
    classic. A future release will make the default value 2.            
    0 - Use default for this release                                    
    1 - Use classic V2R6 algorithm 1 for step adjustments              
    2 - Use AMP algorithm 2 for improved collection
    Default: ==> 0
    Reference Article: ==> KAP1ACEAA
    Recommended setting: ==> Default
    Tpareset Required? ==> NO
                     
   
65.  NoDot0Backdown - This field is used to prevent a backdown from

     14.10 when a feature is used while running on 14.10 that is

     not supported on 14.00 and cannot be backed out.

     This field is FALSE by default on 14.00 and cannot be changed

     This field must be set to enable certain 14.10 features. Once

     set this flag cannot be undone except by a 14.00 sysinit.

    Default: ==> FALSE (after upgrade), TRUE (after sysinit)
    Reference Article: ==> N/A
    Recommended setting: ==> Customer preference
    Tpareset Required? ==> NO

66. SnapBypassAggrCache - If TRUE snapshot dumps will be 

     suppressed  if resource exhaustion is detected when reserving

     memory for aggregate processing.                             

     WARNING: This might mask any real resource leak issue in the system.

    Default: ==> TRUE

         Reference Article: ==> DR169629 as of TDBMS_14.0.7.3 the default is changed to TRUE
    Reference Article: ==> DR149835
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

67. (Reserved for future use)
   

68. (Reserved for future use)

69  PMPC_SessionRateThreshold - This field specifies a threshold in    

    seconds for allowing Monitor Session cache refresh on demand by    

    external applications when TDWM Workload category is enabled.      

    This value is not used if TDWM Workload category is disabled and   

    Monitor Session cache refresh is always based on demand and        

    session rate.                                                      

    Valid range is 0 - 600.                                            

    If the value is zero or exception interval is higher than this     

    threshold value cache refresh on demand is enabled, otherwise      

    monitor session cache is refreshed at exception interval rate.

    Default: ==> 60
    Reference Article: ==> DR146269
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

70  RoundNumberAsDec - If TRUE indicates that the RoundHalfwayMagUp   

    flag will be honored for NUMBER data just like for DECIMAL data.  

    if FALSE (default), NUMBER data will be Rounded away from 0.

    Default: ==> FALSE
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> YES

71. DMLStatementShipping - This flag enables or disables replicating    
    SQL in lieu of change data for DML statements, such as INSERT,      
    UPDATE, DELETE and MERGE. Valid values are:                        
                0 - Use default (Disabled)
                1 - Enabled                                              
                2 - Disabled  
    Default: ==> 0
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

DBS Control Record - File System Fields:

1.  FreeSpacePercent - This field is used by the DBS and the File
    System to determine the percentage of free space to leave on
    cylinders during bulk load operations. The valid range of values
    is 0 .. 75. The default value is 0 (percent).
    Default: ==> 0%
    Reference Article: ==> KAP1ACDEE
    Recommended setting: ==> Default (Customer tunable - see Reference Article)
    Tpareset Required? ==> NO

2.  MiniCylPackLowCylProd - This field is used by the File System to
    determine the number of free cylinders below which it will start
    to perform the "MiniCylPacks" operation in anticipation of their
    need. Setting this field to 0 indicates that "MiniCylPacks" will
    only be performed when no free cylinders exist. The default value
    is 10.
    Default: ==> 10
    Reference Article: ==> KAP1ACDF6
    Recommended setting: ==> Set this value to no more than 20 free cylinders
    Tpareset Required? ==> NO

3.  PermDBSize - This field is used by the File System to determine the
    maximum size of a Permanent Table's multi-row Data Blocks in 512-byte
    sectors. The valid range of values is 42 .. 255. The default value is
    127 (sectors).
    Default: ==> 127 Sectors
    Reference Article: ==> KAP1ACDFE
    Recommended setting: ==> Default (Customer tunable - see Reference Article)
    Tpareset Required? ==> NO

4.  JournalDBSize - This field is used by the File System to determine
    the maximum size of Permanent Journal Table multi-row Data Blocks
    in 512-byte sectors. The valid range of values is 42 .. 255.
    The default value is 254 (sectors).
    Default: ==> 254
    Reference Article: ==> KAP1B43F2
    Recommended setting: ==> Default (Customer tunable - see Reference Article)
    Tpareset Required? ==> NO

5.  DefragLowCylProd - This field is used by the File System to determine
    the number of free cylinders below which it will start to perform
    the "Cylinder defragmentation" operation. Setting this field to 0
    disables "Cylinder defragmentation". The default value is 100.
    Default: ==> 100 free cylinders
    Reference Article: ==> S11000D84AA
    Recommended setting: ==> Default (Customer tunable - see Reference Article)
    Tpareset Required? ==> NO

6.  PermDBAllocUnit - This field is used by the File System to determine
    the allocation unit for Permanent Table's multi-row Data Blocks in
    units of 512-byte sectors. If a Permanent Table Data Block contains
    multiple rows, the size of the Data Block will be a multiple of the
    PermDBAllocUnit. The valid range of values is 1 .. 127. The default
    value is 1 (sector).
    Default: ==> 1 sector
    Reference Article: ==> KAP1ACE06
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

7.  CylindersSavedforPERM -  This field is used to save X number
    of cylinders for Perm Data only.  If the number of Free cylinders
    falls  below this value, any allocation of cylinders for Spool
    Data will result in ERRFILDISKFULL. The default value is 10.
    Valid range is 1 .. 524287.
    Default: ==> 10 cylinders

         Reference Article: ==> KAP245F12
    Reference Article: ==> S11000B041A
    Recommended setting: ==> Default (Customer tunable)
    Tpareset Required? ==> NO

8.  DisableWALforDBs - This field is used to force the writing of data
    blocks directly to disk rather than writing WAL records for them.
    If DisableWALforDBs is FALSE, user table data blocks are committed
    to the WAL Log and cached in memory, or written to disk, depending
    on the specific operation. If DisableWALforDBs is TRUE, user table
    data blocks are always written to disk. The default value is FALSE.
    Default: ==> FALSE
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

9.  DisableWAL - This field is used to force the writing of data blocks
    and cylinder indexes directly to disk rather than writing WAL records.
    Data blocks inserted by multiload always are written to the disk.
    If DisableWAL is FALSE, blocks are committed to the WAL Log and
    cached in memory, or written to disk, depending on the specific
    operation. If DisableWAL is TRUE, all blocks are always written
    directly to disk. The default value is FALSE.
    Default: ==> FALSE
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

10. WAL Buffers - This field is used by the File System to determine the
    number of WAL append buffers to allocate. More buffers increase the
    chance that when a task needs to append a WAL record, there is an
    available buffer to hold the row. Buffers might be unavailable because
    writes are pending, or they are full and a write has not occurred.
    Valid range is 5-40. The default value is 20.
    Default: ==> 20
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> YES

11. MaxSyncWALWrites - This field is used by the File System to determine
    the maximum number of outstanding WAL Log writes to allow before tasks
    requiring synchronous writes are delayed to achieve better buffering.
    If the number of outstanding writes is less than or equal to this
    value, requests for synchronous operations will append the record
    to the current buffer, force a new buffer, and issue a synchronous
    write on the current buffer. If the number of outstanding writes
    is greater than this value, the record will be appended but the
    write will be delayed until one of the outstanding writes completes.
    Valid range is 1-40. The default value is 5.
    Default: ==> 5
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> YES

12. SmallDepotCylsPerPdisk - This field is used by the File System to determine
    the number of depot cylinders to allocate that contain slots of size 128KB.
    This number of cylinders is reserved on each pdisk. Valid range is 1-10.
    The default value is 2.
    Default: ==> 2
    Reference Article: ==> KAP1ACE0E
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> YES

13. LargeDepotCylsPerPdisk - This field is used by the filesystem to determine
    the number of depot cylinders to allocate that contain slots of size 512KB.
    This number of cylinders is reserved on each pdisk. Valid range is 0-10.
    Setting this field to 0 disables large depot usage. The default value is 1.
    Default: ==> 1
    Reference Article: ==> KAP1ACE12
    Recommended setting: ==> Default
    Tpareset Required? ==> YES

14. WAL Checkpoint Interval - This field is used by the File System to
    determine the amount of time that elapses between WAL checkpoints.
    Valid range is 1-240. The default value is 60 (seconds).
    Default: ==> 60 seconds
    Reference Article: ==> KAP1ACE1A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

15. Free Cylinder Cache Size  - This field is used by the File System to determine
    how many cylinders are to be managed in File System cache for use as
    spool cylinders.
    NOTE - A value of 1 will disable use of the cache.
    Valid range is 1 - 1000.  Default is 100.
    Default: ==> 100
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> YES

16. Bkgrnd Age Cycle Interval - This field is used by the File System to
    determine the amount of time that elapses between background cycles to
    write a subset of modified segments in the cache to disk. This background
    activity serves to reduce the size of the WAL log and promotes improved  
    disk space utilization in WAL modes. Valid range is 1-240. The default    
    value is 60 (seconds).
    Default: ==> 60 seconds
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

17. DisableAutoCylPack - Indicates whether AutoCylPack is Disabled/Enabled.
    When set to FALSE, AutoCylPack is enabled.
    When set to TRUE, AutoCylPack is disabled.
    Default is FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP23B3F6
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


18. AutoCylPackColdData - Indicates whether AutoCylPack works on cylinders
    storing COLD data or not. Data that is accessed infrequently is classified
    as COLD data.
    When set to TRUE, AutoCylPack works on COLD cylinders.
    When set to FALSE, AutoCylPack does not work on COLD cylinders.
    Default is FALSE.
    This field is effective only if you have purchased Teradata Virtual Storage.
    Default: ==> FALSE
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


19. AutoCylPackFSP - Indicates the default target Free Space Percent (FSP) for
    AutoCylPack. FSP is the proportion of space that should not be occupied by data
    on cylinders.
    For each affected table, AutoCylPack honors FSP settings in the following
    order of precedence.
    1. FSP specified in the CREATE or ALTER table statement.
    2. FSP specified during most recent Ferret PACKDISK command.
    3. FSP specified in this field, i.e. AutoCylPackFSP.
    Valid range is 10 - 75. Default is 10.
    Default: ==> 10
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


20. AutoCylPackStyle - Indicates the data migration style that AutoCylPack
    is willing to perform in the process of achieving target Free Space Percent.
    If AutoCylPackStyle is:
    1. Data will be moved to achieve an overall increase in available space
       per cylinder.
    2. Data will be moved to achieve either an overall increase or decrease
       in available space per cylinder.
    3. Data will be moved to achieve an overall decrease in available space
       per cylinder.
    Valid range is 1 - 3. Default is 2.
    Default: ==> 2
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


21. AutoCylPackThresh - Indicates the variance between actual Free Space Percent
    (FSP) on a cylinder and the target FSP for the cylinder above which AutoCylPack
    will process the cylinder.
    If the variance between existing FSP of a cylinder and the target
    FSP is greater than AutoCylPackThresh, AutoCylPack will process the cylinder.
    Otherwise, it will not.
    Valid range is 5 - 30. Default is 5.
    Default: ==> 5%
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


22. AutoCylPackFreeCylThresh - Indicates the number of free cylinders
    below which AutoCylPack will not operate. The value of this field should
    be greater than DBSControl field MiniCylPackLowCylProd field.
    This is to prevent AutoCylPack and MiniCylPack working against each other.
    Valid range is 20 - 100. Default is 20.
    Default: ==> 20
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


23. AutoCylPackInterval - Indicates the amount of time
    AutoCylPack sleeps between checks of whether any cylinders need processing.
    AutoCylPack sleeps if it completes a pass of considering all the cylinders
    represented in the MI, if it runs into extreme lock contention on CIs/DBs,
    or if the system stays busy for a long time (see help for AutoCylPackIOThresh).
    Valid range is 60 - 600. Default is 60 seconds.
    Default: ==> 60 seconds
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


24. AutoCylPackIOThresh - Indicates the maximum number of outstanding I/Os
    on a node above which AutoCylPack will sleep for a short interval
    before checking the outstanding I/O count again.
    It limits the performance impact of AutoCylPack on the foreground work load.
    If the system continues to stay busy, AutoCylPack will sleep according to the
    AutoCylPackInterval defined.
    Valid range is 1 - 1000. Default is 1.
    Default: ==> 1
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


25. (Reserved for future use)

26. (Reserved for future use)

27. MergeBlockRatio - This field is used by the File System when merging datablocks
    together to determine the maximum size of a newly merged datablock. The value
    in this field is interpreted as a percentage of the maximum multi-row datablock
    size defined for the table to which the blocks belong. If no maximum multi-row
    size is defined as a table-level attribute, the value in this field is interpreted
    as a percentage of the maximum multi-row datablock size as defined in the
    PermDBSize field. For spool, temporary, and WAL data, this value is not used and
    merges are disabled. For all other data, the value is only in effect when 13.10
    features are enabled. Valid range is 0 - 100. The system defined default value
    is 60 (percent). If 0 is entered, the value reverts to the current system defined
    default value.
    Default: ==> 60%
    Reference Article: ==> KAP1B43F6
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


28. DisableMergeBlocks - This field is used by the File System to determine whether
    datablocks are eligible to be merged together. If the value is TRUE, then no  
    datablocks in any tables will ever be merged. I.e., the MergeBlockRatio field is
    ignored, and any table-level MergeBlockRatio attribute setting is also ignored.
    If the value is FALSE and 13.10 features are enabled, then the value in the    
    MergeBlockRatio field and/or table-level MergeBlockRatio attributes are examined
    to determine the eligibility of blocks to be merged together. The default value
    is FALSE.
    Default: ==> FALSE
    Reference Article: ==> KAP1B32FA
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

100. MIRebuildPreloadHighWaterMark - This field is used by the File System to

     determine how many CIs to initially preload before acquiring the first set

     of CIs to begin building the MI.  It is used in conjunction with the

     MIRebuildPreloadLowWaterMark field when rebuilding the MI.  The MI rebuild

     preload high and low water marks are constrained by the following rules:

         - If the MI rebuild preload high and low water mark settings are both

           0, then the default water mark settings are used.

         - If the preload high water mark is 0, then the default high water

           mark is used.

         - If the preload high water mark is not 0, then the low water mark

           setting as specified in the DBS control GDO is used as is. In other

           words, if the low water mark setting is 0 in this case, the default

           low water mark setting is NOT used.  This allows preload water mark

           settings such as HWM = 10, LWM = 0.

         - If the MI rebuild preload water mark settings are non zero and

           equal, then MI rebuild preloads are disabled, i.e.  HWM = 0,

           LWM = 0.

     MI rebuild preload water mark tunables are used in the following way:

         Initially, CIs are preloaded up to the high water mark level before

         any CIs are acquired.  After preloading the CIs, CIs are acquired one

         at a time until the number of CIs preloaded but not yet acquired is at

         the preload low water mark.  And then CIs are once again preloaded up

         to the preload high water mark, and then CIs are acquired until the

         preloaded CI count drops down to the preload low water mark.  Repeat

         until all CIs are processed.  For example, with a MI rebuild preload

         high water mark of 192 and low water mark of 168, initially, 192 CIs

         are preloaded.  Then 24 CIs are acquired and processed, 24 CIs

         preloaded, 24 CIs acquired and processed, 24 CIs, preloaded, and so

         on.  In other words, the CI preload cycle looks like 192, 24, 24, ...

     The default value of MIRebuildPreloadHighWaterMark is 192.

     Default: ==> 0 (192)
     Reference Article: ==> DR147950
     Recommended setting: ==> Default
     Tpareset Required? ==> NO

101. MIRebuildPreloadLowWaterMark - This field is used by the File System to

     determine how many CIs to preload after acquiring a set of CIs when

     building the MI.  It is used in conjunction with the

     MIRebuildPreloadLowWaterMark field when rebuilding the MI.  The default

     value of MIRebuildPreloadLowWaterMark is 168.

     Default: ==> 0 (168)
     Reference Article: ==> DR147950
     Recommended setting: ==> Default
     Tpareset Required? ==> NO

102. HeatMapChecksumCount - The sampling percentage of each disk sector in the

     HeatMapStructure. Default is to sample 2% of the disk block. Valid range

     is 0 - 100 %.

     Default: ==> 2
     Reference Article: ==> DR152969
     Recommended setting: ==> Default
     Tpareset Required? ==> NO

DBS Control Record - Compression Fields:

Note on Compression: 
Block Level Compression (BLC) enables data compression at the data block (DB)
level of the Teradata Database file system. Compression reduces the amount of
storage required for a given amount of data. BLC increases CPU utilization for
dynamic compression and decompression operations, which can result in reduced
performance. However, in some situations, BLC can improve database performance
due to reduced I/O bandwidth.

Compression can be enabled or disabled entirely, or for various categories of
tables. Only primary, fallback, and CLOB subtables, and join and hash indexes
are subject to compression. Secondary indexes are never compressed.

The PermDBSize and JournalDBSize settings in DBS Control apply to the sizes of
uncompressed DBs.

Note: Compression is subject to the BlockLevelCompression setting, and to the
criteria specified by the MinPercentCompReduction and MinDBSectsToCompress DBS
Control settings. DBs not meeting these minimum criteria will not be compressed.


 1. BlockLevelCompression - This field controls whether block level compression is
    enabled globally or not.  Valid values are ON, to enable compression, or OFF, to
    disable compression.  This field can only be set to ON if the No 13.0 Backdown
    flag has been set, otherwise it has a default setting of OFF.
    If this field is ON, compression is enabled or disabled for individual
    table types as specified by their corresponding tunables.
    If this field is OFF, it supersedes all table type specific settings.
    Default: ==> OFF
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


2.  This field is reserved for future use.

3.  CompressSpoolDBs - This field is used by the File System to determine

    how to handle Spool DBs with respect to compression.  This setting and

    all other compression attributes specified for the base table of this

    table type also applies to sort and index maintenance work subtables

    and CLOB subtables corresponding to this table type.

    There are three values that are used, ALWAYS, NEVER and IFNOTCACHED.

    The default value is NEVER.

    ALWAYS means that the data blocks of a new Spool table will be compressed.

    NEVER means that the data blocks of a new Spool table will not be compressed.

    IFNOTCACHED means that the data blocks of a new Spool table will be compressed

    if the DBs are not being cached.

    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

 4.  This field is reserved for future use.

 5.  CompressMloadWorkDBs - This field is used by the File System to
     determine how to handle Multiload sort and index maintenance work
     tables with respect to compression.
     There are three values that are used, UNLESSQBNO, NEVER and ONLYIFQBYES.
     The default value is ONLYIFQBYES.
     ONLYIFQBYES means that the data blocks of a new table will not be
     compressed unless the Query Band option specifies to compress them.
     UNLESSQBNO means that the data blocks of a new table will be compressed
     unless the Query Band option specifies not to compress them.
     NEVER means that the data blocks of a new table will not be compressed
     regardless of what Query Band options are specified.
    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


 6.  CompressPJDBs - This field is used by the File System to determine
     how to handle Permanent Journal DBs with respect to compression.  This
     setting and all other compression attributes specified for the base table
     of this table type also applies to sort and index maintenance work tables
     corresponding to this table type.
     There are two values that are used, ALWAYS and NEVER.  The default is NEVER.
     ALWAYS means that the data blocks of a new table will be compressed.
     NEVER means that the data blocks of a new table will not be compressed.
    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


 7.  MinDBSectsToCompress - A parameter to be used by the compression.
     algorithm to determine if enough sectors exist in the Data Block
     to warrant compression.  If not enough sectors exist in the Data Block,
     the Data Block will not be saved in the compressed format.
     The default value is 16.  The allowable range is 2 to 255.
    Default: ==> 16
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


 8.  MinPercentCompressReduction - A parameter to be used by the compression
     algorithm to determine if compression of the Data Block has reduced the
     size a sufficient amount.  If the size reduction is not adequate, the
     Data Block will not be saved in the compressed format.
     The allowable range is 0 to 99 with a default of 20
    Default: ==> 20%
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


 9.  CompressionAlgorithm - A parameter indicating which algorithm to use for

     compression/decompression.  'ZLIB', 'ELZS_H' and 'ELZS_S' are the algorithms

     supported.  The default is 'ZLIB'.  If ELZS_H is selected and the

     board is absent, ELZS_S will automatically be used.  ELZS_S is not a

     recommended selection unless the hardware is suspect.  Before choosing ELZS_H

     or ELZS_S, please ensure all nodes have the requisite support installed.  If

     some node lacks support, unanticipated behavior can be expected.

         o Data qualifying for compression will not be compressed.

         o Data already compressed previously via ELZS algorithms (perhaps present

           from a restore from a system that had support) will cause the task

           accessing the data to be aborted when trying to decompress the data.

    Default: ==> ZLIB
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

10.  CompressLevel - A parameter used by the selected compression algorithm.

     For ZLIB, valid values are from 1 thru 9.
     A value of 1 indicates preference for maximum compression speed.
     Each higher values sacrifices more speed for better compression ratio.
     The highest value of 9 optimizes for minimal compression size.
     The default value is 6, a balance between speed and size.
    Default: ==> 6
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO


11. This field is reserved for future use.

12.  UncompressReservedSpace - Specifies a percentage value indicating

     the minimum amount of free space that should be left in a database

     when performing an uncompress operation.  If there is not sufficient

     space the uncompress operation will be terminated and the table left

     in a partially compressed state (but still usable).

     Allowable range is 0 to 90(percent) with a default of 5.

    Default: ==> 5%
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

13.  This field is reserved for future use.

14.  CompressPermPrimaryDBs - This field is used by the File System to determine

     how to handle primary subtable datablocks in Permanent storage of tables

     using manual compression.  This field does not affect tables using automatic

     compression.  System data, dictionary data, table headers,

     LOBs (except CLOBs), Snapshot Logs, Restore Logs, and all primary row

     secondary indexes are always excluded from compression.  This setting

     and all other compression attributes specified for the base table of

     this table type also apply to sort worktables and index maintenance

     worktables corresponding to this table type.

     ONLYIFQBYES means that the datablocks of a new subtable will NOT be

     compressed unless a Query Band option is provided that specifies to

     compress them.

     UNLESSQBNO means that the datablocks of a new subtable will be compressed

     unless a Query Band option is provided that specifies NOT to compress them.

     NEVER means that the datablocks of a new subtable will NOT be compressed

     regardless of any specified Query Band option.

     The default value is ONLYIFQBYES.

    Default: ==> ONLYIFQBYES

    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

15.  CompressPermFallbackDBs - This field is used by the File System to determine

     how to handle fallback subtable datablocks in Permanent storage of tables

     using manual compression.  This field does not affect tables using automatic

     compression.  System data, dictionary data, table headers,

     LOBs (except CLOBs), Snapshot Logs, Restore Logs, and all primary row

     secondary indexes are always excluded from compression.  This setting

     and all other compression attributes specified for the base table of

     this table type also apply to sort worktables and index maintenance

     worktables corresponding to this table type.

     ONLYIFQBYES means that the datablocks of a new subtable will NOT be

     compressed unless a Query Band option is provided that specifies to

     compress them.

     UNLESSQBNO means that the datablocks of a new subtable will be compressed

     unless a Query Band option is provided that specifies NOT to compress them.

     NEVER means that the datablocks of a new subtable will NOT be compressed

     regardless of any specified Query Band option.

     The default value is ONLYIFQBYES.

    Default: ==> ONLYIFQBYES

    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

16.  CompressPermPrimaryCLOBDBs - This field is used by the File System to

     determine how to handle primary CLOB subtable datablocks in Permanent storage of

     tables using manual compression.  This field does not affect tables using

     automatic compression.  System data, dictionary data, table headers,

     LOBs (except CLOBs), Snapshot Logs, Restore Logs, and all primary row

     secondary indexes are always excluded from compression.  This setting

     and all other compression attributes specified for the base table of

     this table type also apply to sort worktables and index maintenance

     worktables corresponding to this table type.

     ONLYIFQBYES means that the datablocks of a new subtable will NOT be

     compressed unless a Query Band option is provided that specifies to

     compress them.

     UNLESSQBNO means that the datablocks of a new subtable will be compressed

     unless a Query Band option is provided that specifies NOT to compress them.

     NEVER means that the datablocks of a new subtable will NOT be compressed

     regardless of any specified Query Band option.

     The default value is NEVER.

    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

17.  CompressPermFallbackCLOBDBs - This field is used by the File System to

     determine how to handle fallback CLOB subtable datablocks in Permanent storage

     of tables using manual compression.  This field does not affect tables using

     automatic compression.  System data, dictionary data, table headers,

     LOBs (except CLOBs), Snapshot Logs, Restore Logs, and all primary

     row secondary indexes are always excluded from compression.  This setting

     and all other compression attributes specified for the base table of

     this table type also apply to sort worktables and index maintenance

     worktables corresponding to this table type.

     ONLYIFQBYES means that the datablocks of a new subtable will NOT be

     compressed unless a Query Band option is provided that specifies to

     compress them.

     UNLESSQBNO means that the datablocks of a new subtable will be compressed

     unless a Query Band option is provided that specifies NOT to compress them.

     NEVER means that the datablocks of a new subtable will NOT be compressed

     regardless of any specified Query Band option.

     The default value is NEVER.

    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

18.  CompressGlobalTempPrimaryDBs - This field is used by the File System to

     determine how to handle primary subtable datablocks in Global Temporary storage

     with respect to compression.  System data, dictionary data, table headers,

     LOBs (except CLOBs), Snapshot Logs, Restore Logs, and all primary

     row secondary indexes are always excluded from compression.  This setting

     and all other compression attributes specified for the base table of

     this table type also apply to sort worktables and index maintenance

     worktables corresponding to this table type.

     ONLYIFQBYES means that the datablocks of a new subtable will NOT be

     compressed unless a Query Band option is provided that specifies to

     compress them.

     UNLESSQBNO means that the datablocks of a new subtable will be compressed

     unless a Query Band option is provided that specifies NOT to compress them.

     NEVER means that the datablocks of a new subtable will NOT be compressed

     regardless of any specified Query Band option.

     The default value is NEVER.

    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

19.  CompressGlobalTempFallbackDBs - This field is used by the File System to

     determine how to handle fallback subtable datablocks in Global Temporary storage

     with respect to compression.  System data, dictionary data, table headers,

     LOBs (except CLOBs), Snapshot Logs, Restore Logs, and all primary

     row secondary indexes are always excluded from compression.  This setting

     and all other compression attributes specified for the base table of

     this table type also apply to sort worktables and index maintenance

     worktables corresponding to this table type.

     ONLYIFQBYES means that the datablocks of a new subtable will NOT be

     compressed unless a Query Band option is provided that specifies to

     compress them.

     UNLESSQBNO means that the datablocks of a new subtable will be compressed

     unless a Query Band option is provided that specifies NOT to compress them.

     NEVER means that the datablocks of a new subtable will NOT be compressed

     regardless of any specified Query Band option.

     The default value is NEVER.

    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

20.  CompressGlobalTempPrimaryCLOBDBs - This field is used by the File System to

     determine how to handle primary CLOB subtable datablocks in Global Temporary storage

     with respect to compression.  System data, dictionary data, table headers,

     LOBs (except CLOBs), Snapshot Logs, Restore Logs, and all primary

     row secondary indexes are always excluded from compression.  This setting

     and all other compression attributes specified for the base table of

     this table type also apply to sort worktables and index maintenance

     worktables corresponding to this table type.

     ONLYIFQBYES means that the datablocks of a new subtable will NOT be

     compressed unless a Query Band option is provided that specifies to

     compress them.

     UNLESSQBNO means that the datablocks of a new subtable will be compressed

     unless a Query Band option is provided that specifies NOT to compress them.

     NEVER means that the datablocks of a new subtable will NOT be compressed

     regardless of any specified Query Band option.

     The default value is NEVER.

    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

21.  CompressGlobalTempFallbackCLOBDBs - This field is used by the File System to

determine how to handle fallback CLOB subtable datablocks in Global Temporary storage with respect to compression.  System data, dictionary data, table headers,

     LOBs (except CLOBs), Snapshot Logs, Restore Logs, and all primary

     row secondary indexes are always excluded from compression.  This setting

     and all other compression attributes specified for the base table of

     this table type also apply to sort worktables and index maintenance

     worktables corresponding to this table type.

     ONLYIFQBYES means that the datablocks of a new subtable will NOT be

     compressed unless a Query Band option is provided that specifies to

     compress them.

     UNLESSQBNO means that the datablocks of a new subtable will be compressed

     unless a Query Band option is provided that specifies NOT to compress them.

     NEVER means that the datablocks of a new subtable will NOT be compressed

     regardless of any specified Query Band option.

     The default value is NEVER.

    Default: ==> NEVER
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

22 - 31.  (Reserved for future use)

32.  EnableTempBLC - Controls all parts of temperature based BLC

     This includes the AutoTempComp background task responsible for

     compressing and uncompressing data based on temperature, and any

     decisions to automatically compress COLD data as it is loaded.

     If this value is enabled, it is highly recommended that the

     tunable DisableAutoCylPack be set to FALSE and AutoCylPackColddata

     be set to TRUE.  If set to FALSE, the rules for compression revert

     to the inheritance model used by tradional BLC(if enabled).

     EnableTempBLC might be set to FALSE during a performance critical

     window.

     Valid Values are TRUE and FALSE.  The default is FALSE.  Changes are

     effective immediately.

    Default: ==> FALSE
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

33. DefaultTableMode - Indicates whether the default for tables with

    no specific table level attribute is to manage their compression

    manually, automatically or never compress the table.  If MANUAL

    is chosen, it is the user's responsibility to compress or

    decompress the table.  If AUTOTEMP is chosen, then data in the

    table will be compressed or decompressed automatically by the

    AutoTempComp background task based on the temperature of the data.

    In this mode, parts of the table(such as partitions) may have a

    different compression state based on their frequency of access.

    If NEVER is chosen, then data in the table will never be compressed.

    Valid values are MANUAL, AUTOTEMP and NEVER.  The default is

    MANUAL.  Changes are effective immediately.

    Default: ==> MANUAL
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

34. TempBLCThresh - Indicates the temperature threshold at which data in

    tables using AUTOTEMP compression management are compressed.  Any

    data at or colder than the specified temperature will be compressed.

    Temperatures can be specified by keyword or a floating point value

    representing a percentage of the coldest allocated

    user data (PERM TABLES) that should be compressed.  The percentage is

    calculated based on the number of cylinders of PERM data used,

    without regard to how fully occupied the cylinders are.

    Valid Values are COLD, WARM, HOT, or a floating point value greater

    than 0 and less than or equal to 100.  The default is COLD.

    Changes are effective immediately.

    Default: ==> COLD
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

35. TempBLCSpread - A percentage that is added to or subtracted from

    TempBLCThresh to determine if a cylinder should change compression state.

    For example, if the TempBLCThresh is defined as COLD and the TempBLCSpread

    is 5% then the data must be 5% colder than COLD to be compressed and

    5% warmer than COLD to be uncompressed.  This value can be used to

    prevent data whose natural temperature is close to the threshold

    from being repeatedly compressed and uncompressed.

    Valid values are from 1 to 25%.  The default is 5%.  Changes are

    effective immediately.

    Default: ==> 5%
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

36. TempBLCInterval - The amount of time for AutoTempComp to wait after

    compressing or uncompressing data before checking if other data

    needs to be processed.  If the whole MI is scanned and no data is

    found to process then AutoTempComp will sleep for a duration

    indicated by TempBLCRescanPeriod.

    Valid values are from 1 to 120 minutes.  The default is 10 minutes.

    Changes are effective the next time AutoTempComp wakes up.

    Default: ==> 5%
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

37. TempBLCIOThresh - The maximum number of outstanding I/O's on a node

    above which AutoTempComp will sleep for a short interval before running

    again.  This setting limits the performance impact of AutoTempComp

    on foreground work loads.  If the system remains busy, AutoTempComp

    will sleep according to the TempBLCInterval setting.

    Valid values are from 1 to 1000.  The default is 1.  Changes are

    effective immediatelly.

    Default: ==> 1
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

38. TempBLCPriority - The priority at which AutoTempComp runs. Priorities

    can be specified using their old or new names.

    Valid values are: LOW, MEDIUM (or DEFAULT), HIGH,

    or RUSH(or TOP).  The default is MEDIUM (or DEFAULT).  Changes are

    effective the next time AutoTempComp wakes up.

    Default: ==> MEDIUM /DEAUL
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

39. TempBLCRescanPeriod - The length of time in days to allow before

    revalidating the state of cylinders using automatic compression to

    see if they are properly compressed or uncompressed.  This rescan is

    necessary as the temperature of a cylinder can change, even if the

    cylinder is not directly modified.  At least one scan is performed

    after each startup.

    Valid values are from 1 to 90 days.  The default is 7 days.

    Changes are effective the next time AutoTempComp wakes up.

    Default: ==> 7
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

           

DBS Control Record - Performance Fields:

1.  DictionaryCacheSize - This field defines the size of the dictionary
    cache for each PE Vproc on the system. The valid range of values
    is 64 .. 16384. The default value is 3072 (kilobytes).
    Default: ==> 3072 KB
    Reference Article: ==> KAP1ACE1E
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> YES

2.  DBSCacheCtrl - This field is used to enable or disable the performance
    enhancements associated with Cache Control Page-Release Interface. The
    default value is  TRUE.
    Default: ==> TRUE
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

3.  DBSCacheThr - This field is used to specify the percentage value to
    use to calculate the cache threshold when the DBSCacheCtrl field is
    enabled. The valid range of values is 0 .. 100. The default value
    is 10.
    Default: ==> 10%
    Reference Article: ==> KAP1ACE2A
    Recommended setting: ==> 1% (see Reference Article)
    Tpareset Required? ==> NO

4.  MaxParseTreeSegs - This field defines the maximum number of 64KB tree
    segments that the parser will allocate while parsing a request. The
    valid range of values is 12 .. 12000. The default value is 2000.
    32-bit platform, the default is 1000.
    Default: ==> 32bit = 1000 Range = 12 to 6000

·          

o     

§   

§  64bit = 2000 Range = 12 to 12000

    Reference Article: ==> N/A
    Recommended setting: ==> Default (Change only under the direction of Teradata Support Center)
    Tpareset Required? ==> NO

5.  ReadAhead - This field is used to enable or disable the performance
    enhancements associated Read-Ahead Sequential File Access Workload.
    The default value is  TRUE.
    Default: ==> TRUE
    Reference Article: ==> C10000D2126
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

6.  StepsSegmentSize - This field defines the maximum size of the plastic
    steps segment (aka "OptSeg") in kilobytes. The valid range of values
    is 64 .. 1024. The default value is 1024 (kilobytes).
    Default: ==> 1024
    Reference Article: ==> C10000D2132
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

7.  RedistBufSize - This field determines the size (in bytes/kilobytes)
    of hashed row redistribution buffers used by MultiLoad, FastLoad,
    Create Unique Index, and SMP-only SQL step responses. Each such
    redistribution in an AMP will use one such buffer for each AMP in
    the configuration. The ranges of valid values are:
                 -1: the optimal buffer size to avoid extra memory overhead
                     for sending row redistribution messages on the
                     current system (3832 bytes).
                  0: the default buffer size for current system (3832 bytes).
            1 .. 63: the buffer size in units of kilobytes. The buffer size
                     in this value range may be adjusted internally
                     for better memory utilization.
       512 .. 65024: the buffer size in units of bytes, which is equivalent
                     to 0.5 .. 63.5 kilobytes. The buffer size in this
                     case is used exactly as is without further adjustment.
    The default value is 3832 bytes.
    Default: ==> 3832 bytes
    Reference Article: ==> C10000D213E
    Recommended setting: ==> Default (Customer tunable, see Reference Article)
    Tpareset Required? ==> NO

8.  DisableSyncScan - This field enables or disables the performance
    enhancements associated with synchronized full table scans.  A value
    of F enables the feature.  A value of T disables the feature.
    The default is FALSE.
    Default: ==> FALSE
    Reference Article: ==> C10000D212E
    Recommended setting: ==> Default (See Reference Article)
    Tpareset Required? ==> YES

9.  SyncScanCacheThr - This field specifies the percentage of file system
    (FSG) cache that the Teradata file system can assume is available for  
    synchronized scans of tables.  It does not reserve cache for this  
    purpose.  Instead, this value specifies the amount of permanent data
    the Teradata file system should try to retain in memory at any one
    time for all tables involved in synchronized scans.  Whether that
    much memory is truly available depends on actual workload.  The valid
    range of values is 0 .. 100, where 0 indicates that the default value
    should be used.  The default value is 10.
    Default: ==> 10%
    Reference Article: ==> S11000DA6A6
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

10. HTMemAlloc - This field specifies the percentage of memory to be
    allocated to a hash table for a hash join.  A value of 0 turns off
    hash joins. Valid range is 0 to 10.
    The default value for 64-bit platform is 10.
    The default value for 32-bit platform is 2.
    Default: ==> 32bit = 2; 64bit = 10
    Reference Article: ==> KAP1ACE2E
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

11. SkewAllowance - This field specifies a percentage factor used by the
    optimizer in deciding on the size of each hash join partition. It makes
    allowance for data skew in the build relation.  Valid range is 20 to 80.
    The default value is 75.
    Default: ==> 75%
    Reference Article: ==> S11000DA6B2
    Recommended setting: ==> Default (See Reference Article)
    Tpareset Required? ==> NO

12. ReadAhead Count - If the ReadAhead field is set TRUE,
    ReadAhead Count is used to specify the number of data blocks
    that will be preloaded in advance of the current file
    position while performing sequential scans.  The valid range
    of values is 1 .. 100. The default value is 1.
    Default: ==> 1 DB
    Reference Article: ==> C10000D211E
    Recommended setting: ==> Default (See Reference Article)
    Tpareset Required? ==> NO

13. PPICacheThrP - This field specifies the maximum amount of FSG memory
    to be used, in tenths of a percent of available FSG memory, for disk read
    operations dealing with multiple partitions.
    If the amount of available FSG memory per AMP exceeds 100 MB,
    the threshold will be calculated based on 100 MB per AMP.
    The valid range is 0 to 500 (that is, 0% to 50.0%).
    The default value is 10 (that is, 1.0%).
    Default: ==> 10
    Reference Article: ==> KAP1ACE32
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

14. ReadLockOnly - This field is used to disable or enable the special  
    read-or-access lock protocol on 1) DBC.AccessRights table during access
    rights validation and on 2) dictionary tables accessed by read-only    
    queries during request parsing.  The default value is FALSE.
    Default: ==> FALSE
    Reference Article: ==> SD1000D07B6
    Recommended setting: ==> Default (See Reference Article)
    Tpareset Required? ==> NO

15. IAMaxWorkloadCache - This tunable field defines the maximum size in
    megabytes of the workload cache used by Index Wizard Analysis.
    Valid range is 64 to 187. The default value is 64 (megabytes).
    Default: ==> 64MB
    Reference Article: ==> SD1000CA9F6
    Recommended setting: ==> Default (See Reference Article)
    Tpareset Required? ==> NO

16. MaxRequestsSaved - This field governs the number of  
    request cache entries allowed on a PE. The default  
    number of entries that can be saved in the cache    
    is 600. The user can modify this field to a fixed    
    value that ranges from 300 to 2000 and, is a multiple  
    of 10.
    Default: ==> 600
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> YES

17. UtilityReadAheadCount - Specifies the number of data blocks the
    Teradata utilities will preload at a time while it performs its
    sequential scan.  The utilities uses this field instead of the ReadAhead
    and ReadAhead Count fields.  The valid range of values is 1 ..  100.
    The default value is 10 (blocks).
    Default: ==> 10 blocks
    Reference Article: ==> KAP1ACE36
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

18. StandAloneReadAheadCount - Specifies the number of data blocks to
    preload when File System startup or a utility runs as a standalone
    task.  The valid range of values is 1 ..  100.
    The default value is 20 (blocks).
    Default: ==> 20 blocks
    Reference Article: ==> KAP1ACE3A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

19. DisablePeekUsing - This field enables or disables the
    performance enhancements associated with exposed using values.
    A value of F enables the feature. A value of T disables the
    feature. The default value is F (FALSE).
    Default: ==> FALSE
    Reference Article: ==> KAP1ACE3E
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO

20. IVMaxWorkloadCache - This tunable field defines the maximum size in
    megabytes of the workload cache used by Index Wizard Validation.
    Valid range is 1 to 32. The default value is 1 (megabytes).
    Default: ==> 1
    Reference Article: ==> KAP1ACE42
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

21. RevertJoinPlanning - Setting this to T (TRUE) causes join planning
    to use a fixed 64 table limit for joins and TD12 join planning heuristics.
    Valid range - T, F
    Default - F
    Default: ==> FALSE
    Reference Article: ==> KAP1ACE46
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

22. MaxJoinTables - The maximum number of tables allowed in a join.  0 is mapped
    to the default value, and any value in 1..16 is mapped to 16.
    Valid range - 0..128
    Default - 128
    Default: ==> 0 (128)
    Reference Article: ==> KAP1ACE4A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

23. DBQLXMLPlanMemLimit - Sets the maximum amount of memory in kilobytes
    the XML document can use per request. If the size of a document exceeds
    this value, then the document will only contain the error message.
    The valid range of values is 0 .. 128000.
    A value of 0 disables memory limit checking of the XML document.
    The minimum recommended value is 8192.
    The default value is 64000.
    Default: ==> 64000 KB
    Reference Article: ==> KAP1B466A
    Recommended setting: ==> Default
    Tpareset Required? ==> NO
 
24. LimitInlistCVal - This field limits the number of combined values in Inlist.
    The valid range of values is 1 ..  1048576.
    The default value is 1048576.
    Default: ==> 1048576
    Reference Article: ==> NTA 2066 (This flag will be moved to Internal settings per DR140383)
    Recommended setting: ==> Default
    Tpareset Required? ==> NO

25. NumStatisticsCacheSegs - This field defines the number of segments for the
    statistics cache maintained at each PE Vproc on the system. The valid
    range of values is 2 .. 200. The default value is 4. The default size
    of the each segment is 1024KB.
    Default: ==> 4
    Reference Article: ==> N/A
    Recommended setting: ==> Default
    Tpareset Required? ==> YES




DBS Control Record - Disk I/O Integrity Fields:

CHECKSUM LEVELS

1.  System Tables            - This field is used to set the checksum level of
                               system tables (data dictionaries, session
                               information, etc.).
2.  System Journal Tables    - This field is used to set the checksum level of
                               system journal tables (transient journal, change
                               tables, recovery journals).
3.  System Logging Tables    - This field is used to set the checksum level of
                               system logging tables (RSS tables, accounting
                               log tables, etc.).
4.  User Tables              - This field is used to set the checksum level of
                               user tables (includes stored procedures, user-
                               defined functions, join indexes, hash indexes).
5.  Permanent Journal Tables - This field is used to set the checksum level of
                               permanent journal tables.
6.  Temporary Tables         - This field is used to set the checksum level of
                               temporary and spool tables.

Valid values for these fields are NONE, LOW, MEDIUM, HIGH, and ALL.  The
default is NONE.

NOTE:  Modifying field 0 results in setting fields 1 - 6 to the specified
value.

CHECKSUM LEVEL DEFINITIONS

    NONE   -  Do not calculate checksums (sample 0% of the disk block to
              generate a checksum).  This percentage cannot be modified.
7.  LOW    -  Calculate checksums by sampling a low percentage of the disk
              block.  Default is to sample 2% of the disk block.  Valid range
              is 1 - 100%.
8.  MEDIUM -  Calculate checksums by sampling a medium percentage of the disk
              block.  Default is to sample 33% of the disk block.  Valid range
              is 1 - 100%
9.  HIGH   -  Calculate checksums by sampling a high percentage of the disk
              block.  Default is to sample 67% of the disk block.  Valid range
              is 1 - 100%
    ALL    -  Calculate checksums using the entire disk block (sample 100% of
              the disk block to generate a checksum).  This percentage cannot
              be modified.
    Default: ==> NONE
    Reference Article: ==> KAP1ACE56
    Recommended setting: ==> Default (see Reference Article)
    Tpareset Required? ==> NO