_{Settings in OutBoard Cockpit view}

To see the parameters click on the settings from the Outboard Cockpit. The Outboard Settings Screen should appear (appearance may vary depending on which OutBoard modules or possible service packs are installed).

_{Selecting an InfoProvider and viewing settings}

DEFAULT values are pre-delivered in the OutBoard installation package. 
On the first Tab of OutBoard settings, the user can define global parameters, as well as change the units and currency used for OutBoard Cockpit.
The Authorization checkbox determines if OutBoard™ authorization is activated, which determines the following accesses "display", "edit" or "execute" to the users.
The job start-up parameter determines if the background jobs in OutBoard™ are started with the time conversion. When unchecked the jobs are scheduled and started normally by the system scheduler, when there are available background processes. In the other case, the job is started immediately. Archive size unit, Percent per GB, and Currency are the customizable parameters for the Cockpit statistics. Near-line Objects tab:

_{Nearline Object tab in view settings}

Summary of Object names and Object types with the possibility to define InfoProvider-specific settings.
Object-specific settings
DEFAULT values are pre-delivered in the OutBoard installation package. To change an InfoProvider parameter value, uncheck the "From Defaults" checkbox.

NLS Object – NLS Object name

• Max. package number (DTP) - determines the maximum number of packages that can be processed at once in the loading via a Data Transfer Process (DTP).
  Possible values: integer > 0 e.g. 10, 20, 50
  Default value: 5
• Finished packages for processing - maximum number of finished processes during parallel DTP load.
  Possible values: integer > 0 e.g. 10, 20, 50
  Default value: 2
• Parallel jobs for loading – determines how many parallel jobs will be used for reading from the archive when loading from the archive via DTP. (link to further reading).
  Possible values: integer >= 0 e.g. 0, 4, 15
  Default value: 0 (no parallelization)
• Parallel WP for reporting – this value determines how many work processes are used for reading when data from OutBoard is requested for reporting. For more information see Reporting on archived data.
  Possible values: integer > 0 e.g. 10, 20, 50
  Default value: 0 (no parallelization)
• Filtering type – determines the filtering logic when accessing the NLS data.
  Possible values: SETS, RANGE

Expert settings

There is also a possibility for users to enter the Expert settings. These settings have the form of the database table and have also parameters not visible in the Main settings. 
Each parameter can be either InfoProvider-specific (these can vary between different InfoProvider) or is set only globally. A special InfoProvider is used which is called "DEFAULT". If no parameters are set on the InfoProvider level, the DEFAULT value is used.  Global parameters may only be set in the DEFAULT.

_{Output from Settings}

Parameters

• AUTHORIZATION – this parameter determines if OutBoard Authorization is switched ON (value X) / OFF (value ' ')
  Default value: OFF
• BG_JOB_DIRECT_START – this parameter determines if Background jobs in OutBoard are started "With converting to Time – Based Jobs" (value 'X')/ "Without converting to Time – Based Jobs" (no value). When a job is started "Without converting to time-based jobs", this job is scheduled and started normally when any background job is free. When selecting "With converting to Time – Based Jobs", the job will be started immediately if any background job is free. If there are no free background jobs it will try to execute the job every ten seconds for 30 minutes.
  Default value: OFF.

• CHECK_PACKAGE_SIZE - this parameter checks the maximum size of each data package during archiving to avoid possible problems with the future reload of archived data, a load of archived data to other InfoProviders, or performance problems during reporting on archived data in SP19. This parameter can negatively impact the archiving performance in the copy phase when there are many characteristics used as semantic groups in the Data archiving process. To disable this check, change the value of this parameter to an empty value.

• CREATE_PARTITIONING - this parameter applies only to transparent storages. We currently support Oracle and Sybase IQ. When enabled ('X'), the system creates a table of DAP, which is partitioned based on a request.
  Please note this setting doesn't influence already existing DAPs and it doesn't initialize partitioning on already existing DAPs.
• CURRENCY – the currency in which money saved is calculated.
  Possible values: Any currency available in SAP.
  Default value: EUR
• DATA_CLASS – determines the Data Class for a storage table and its indexes. It can be changed only when no data is archived. For the Oracle database, the tablespace for a particular storage table can be selected. It is determined by DATA_CLASS. The link between a tablespace and Data Class can be found in table TAORA. Possible values: Any Data Class available in the SAP system.
  Default value: APPL1
• DISPLAY_COMP_RATIO_TYPE – this parameter determines the way of computing the saved storage by archiving. Two possible settings are possible:LOCAL_ONLY: when the saved storage is calculated, the ratio for calculating the saved storage in external storages is set to 100%. For data archived in the online database, the compression ratio was determined during archiving. When the saved storage is calculated, the compression ratio used for calculation is taken and is determined during archiving. In case the parameter is changed, the statistics need to be recalculated with the report /dvd/nls_upd_clu_ratio_size.
• DUMMY_VARIANT_DATE - Bottom range date limit used for generated data archiving process variants.
  Possible values: date >= 19010101 (YYYYMMDD)
  Note: If a fiscal variant is year-dependent, the value of this setting has to be inside the defined range of years for that variant, otherwise the date will not be translated.
• EXLCUDE_TRANP_REQ_IN_BIN_READ - enable/disable the performance optimization that excludes all requests stored in transparent storage from processing in the binary relevant functionality and vice versa. 
  Default Value: X
• FILTERING_TYPE – determines the way, how is the archived data filtered when accessed. Can be based on RANGE or SETS.
  Possible values: RANGE, SETS
  Default Value: RANGE
• HNA_UNLOAD_PRIO – determines the Unload Priority of Outboard tables for SAP BW running on SAP HANA. It can be 0 ~ 9, where 0 means not-unloadable and 9 means the earliest unload.
  Default value: 9
• LOAD_MAX_GRAN – determines the maximum number of packages that can be processed at once in the loading via a Data Transfer Process (DTP).
  Possible values:  integer > 0 e.g. 10, 20, 50
  Default value:  5
• LOAD_USE_ADD_SORTING – when enabled, data loaded from archived WODS is sorted in the same order in which it was loaded. This applies when the setting has the value ‘X’, L mode is used, source WODS is archived. Setting may be switched ON (value X) / OFF (value ‘ ’)
  Default value: ‘ ‘
• LOAD_PAR_RET – maximum number of finished processes during parallel DTP load.
  Possible values:  integer > 0 e.g. 10, 20, 50
  Default value:  2
• LOAD_PAR_WP – determines how many parallel jobs will be used for reading from the archive when loading from the archive via DTP. 
  Possible values:  integer >= 0 e.g. 0, 4, 15
  Default value: 2 (parallelization using 2 jobs)
• NAV_ATTR_BUFFER – size of the buffering table when the Navigation attribute support is switched on.
  Possible values:  integer > 0 e.g. 10, 20, 50
  Default value: 3000000
• NAV_ATTR_FILTER_POS – defines the filtering order/ priority when the Navigation attribute support is switched on. Value 0 means that filtering on the Navigation attribute is done at first then the filtering on other attributes is done. Value 1 represents the opposite scenario.
  Default value: 0
• NAV_ATTR_SELECTION – this parameter is for turning the navigation attribute support functionality: on (1) or off (" "). This setting can be edited on the InfoProvider level. It is recommended to turn off this parameter for SybaseIQ archiving storage because needed filtering is done on the database level and there is no need to use the standard Outboard navigational functionality in this case.
  Default value: 1
• NLS_STATISTICS – this parameter determines if OutBoard Statistics is switched ON (value X) / OFF (value ' ')
  Default value: ON
• REBUILD_AGGREGATE – this parameter defines if OutBoard aggregates are rebuilt automatically (value X) during the archiving process, or deactivated instead (value ' '). If the automatic rebuild is switched off, the OutBoard aggregates can be activated and filled again using the report /DVD/NLS_FILL_AGGREGATE_IN_BG.
  Default value: ON (X)
• REP_PAR_WP – this value determines how many work processes are used for reading when data from OutBoard is requested for reporting. For more information see [#Reporting Reporting on archived data].
  Possible values:  integer > 0 e.g. 10, 20, 50
  Default value:  2 (parallelization using 2 jobs)
  Note: For an IQ transparent storage, the parameter for REP_PAR_WP cannot be set to higher than 1.
• SIZE_DIFF_THRESHOLD – if using a size-dependent OutBoard license, after reaching this value of free space, the user will get a warning message by each archiving process or creation of DAP.
  Default value: 60 GB
• SIZE_TO_MONEY_COEFF – determines how much money 1 GB costs. Money saved is computed by multiplying SIZE_TO_MONEY_COEFF by storage saved.
  Possible values: integer > 0 e.g. 10, 20, 50
  Default value: 10
• SIZE_UNIT – the value used as a base unit of measurement for all volumes shown in GUI. If changed, data is recalculated accordingly.
  Possible values:  KB, MB, GB    
  Default value:  GB
• STORAGE_UPDATE_PACK_LOG_SECS - storage updates can take quite a long time even for one request, therefore we also provide information about processed packages. The value of this parameter will affect the time-frequency of how often (in seconds) a message about processed packages should be written into the application logs. However, the messages are written only after the complete package is processed, therefore it can happen that the message is not written right after the specified value. 
  Possible values:  integer > 0
  Default value: 600
• UPDATE_SIZE_STATISTICS – this parameter determines if the object size statistics are updated while archiving or reloading data. In the case of substantial archiving requests, recalculation has a noticeable performance impact. Setting may be switched ON (value X) / OFF (value ‘ ’)
  Default value: ON
• USE_DATA_CLASS_BADI - determines if BAdI will be used to define DATA_CLASS. If the value is 'X', the parameter DATA_CLASS cannot be edited in Settings for a given InfoProvider. (The default value can still be changed.) If the BAdI is used, the implementation for BAdI /DVD/NLS_DETERMINE_DATACLASS has to be created. Parameters of BAdI: iv_dapname – Technical Name of InfoProvider ev_data_class – Data class in technical settings. The output format of the BAdI implementing class method GET_DATA_CLASS should be of type TABART.

_{The implementation of BAdI for Data class (Transaction(TA): SE80)}

• USE_PART_KEY_BADI - if the value is 'X', the user has the possibility to define his own partitioning keys for the OutBoard table for the DB6 database. If the BAdI is used, the implementation for BAdI /DVD/NLS_STORAGE_PARTITION_KEY is to be implemented. The output format of the BADI implementing the class method /DVD/NLS_DB6_STORAGE_PART should be the table type /DVD/NLS_TT_FIELD. The input parameter iv_tabname is the name of the table to be partitioned.
• USE_DV1_NAMESPACE – if turned on ('X'), the needed local cluster table in DAP creation is created in the namespace /DV1/. If there are any issues with this namespace, the namespace /BIC/ is used instead – by setting the parameter to ' '.
  Possible values: ON (value X) / OFF (value ' ')
  Default value: ON (X)

OutBoard Settings can be transported between the systems.

• VP_SUPP_NAV_ATTR_IN_LOC_FILT – in case the local filters are defined on navigational attributes, some tuning can be done by setting the parameter to a specific value:
  S(recommended): support filtering on navigational attributes in the local filters
  N: local filters will be removed/cleared from the filtering completely
  R: only local filters parts connected to navigational attributes will be removed
  

             _{OutBoard settings}

• WRITER_ADSO_INDEX_AUTOBUILD - this setting initializes secondary indexes on all key fields of an ADSO during a DAP creation. This setting is supported only for Writer, for ADSO delta calculation type ("DSO like").
  
• WRITER_AUTO_RELOAD_EMPTY_REQ - enables the automatic reload functionality. At the beginning of a DTP load to cube / DSO activation, empty archiving requests are reloaded. Data, therefore, stays in the active table, in the postprocessing, new archiving requests are created with the same archiving condition.
  Possible values: ON (value X) / OFF (value ' ')
  Default value: ON (X)
• WRITER_AUTO_RELOAD_THRESHOLD - sets the maximum number of rows for creating a new archiving request in the postprocessing. In the case of an archiving request having more rows than the set value, a new archiving request has to be created manually.
  Default value: '0'.

Outboard for Analytics with storage management specific settings

Aging profile assignments
Go to Storage management settings > Aging profile assignments
In this tab, you can assign or change an aging profile of a specific InfoProvider.

Archive storages
Go to Storage management settings > Archive storages
Only with OutBoard with the NLS solution, you can decide about the preferred storage location of the NLS database.

_{Storage Definition}

OutBoard Analytics is ready to handle (doesn't mandatory require) additional, an external database or a non-SAP®-based software for data management thus keeping the system landscape simple and easy to maintain.

By clicking on the field Storage Type, you may choose the preferred storage location from the following predefined possibilities:

Cluster storage

The storage saves data within the database tables of the BW system. This solution is not supported if the primary system is running on the SAP HANA database.

RFC storage

The storage saves data in separate database tables of a remotely available system. RFC connections defined earlier can be set through the transaction SM59.

File storage

The storage saves data in separate files on a file system. The file name is determined by the request and package.
Supported operating systems:  Windows© Server, Unix, Linux, AIX©, HP-UX©, WebDAV. The destination must be available for the system and the user.

External database

The storage saves data in separate database tables in a remotely available database. You must also define the secondary connection to this database (Please see the SAP Note 808505). The name of the connection mustn't contain any special characters (e.g. space).

External Database MaxDB

Starting with OutBoard SP20, it is possible to use the MaxDB database as external binary storage.

DB2 transparent storage

Starting with OutBoard SP17, it is possible to create transparent storage for a DB2 database using a column-oriented approach. Data is not compressed, the compression ratio is not comparable to the non-transparent archiving (when using cluster tables).

Transparent Storage Sybase IQ

It is possible to create transparent storage for a SybaseIQ database.

 _{Transparent Sybase IQ - storage definition}

• Connection mode: O (stands for an open connection, OpenSQL is used)
• DB Connection Name: Name of a database connection
• Load stripe size: the size of Stripe, multiple of SYBASE_IQ_BUFFER_SIZE (optional)
• Load stripe width: number of parallel processes (optional)
• SYBASE_IQ_BUFFER_SIZE could be changed at the database connection level (transaction DBCO).

Further information can be found in the SAP Note 2128579.

Load striping across multiple disks is an essential technique for good performance.

Oracle Transparent Storage

It is transparent storage for an Oracle database.

_{Archive storage types - Oracle}

Hadoop storage

Apache Hadoop is implemented as binary storage (compressed). OutBoard for Analytics is compatible with Cloudera's implementation of Hadoop, Hive, and Impala interface components starting with the version CDH 5.x.

_{Archive storage types - Hadoop}

MSSQL BLOB storage

Enables storing data in a compressed BLOB structure into a remote MSSQL database. This reduces the amount of data physically stored, but may also lead to a longer processing time in comparison to a transparent solution. Due to the limitations on the MSSQL side, the maximum size for one BLOB line is 8000 bytes. 

MSSQL transparent storage

Archiving into the transparent MSSQL is supported in OutBoard storage management. Using this option, data in the archive can be accessed more quickly through transparent tables. However, the compression ratio is lower compared with archiving into cluster tables.

Storage management settings – Aging profiles

Aging profiles provide customization for:

• Retention policies: time period until the specified storage type is relevant.
• Storage type link: location where the data is automatically migrated to when the data fulfills the criteria of the retention policy.
• Deletion indicator: if the data can be deleted during the last phase of the chain.

_{Aging profile in Outboard Settings}

Each "Aging profile" has a default storage type used, in the case of no storage destination determined. The definition of an aging profile therefore must be linked to a target storage destination. All InfoProviders linked to the specific profile must have a DAP created or DAP must be able to be determined,  (s-tcode: RSA1) determination is done by using standard time characteristics.
Available profile categories:

• with only time characteristic à maintenance of storage destination is available

with more characteristics à requires BAI implemented to determine the storage destination for each request

Example: Each storage destination has defined boundaries based on the given selection characteristics of the profile. All boundaries are closed from the right and open from the left side so they combine the complete timeline.

1. (0, 3> years – cluster
2. (3, 6> years – remote system
3. (6,10> years – file
4. (10, infinity) – delete

OutBoard on SAP HANA

The default setting for Unload priority for OutBoard tables is defined as 9. Unload priority sets the priority of the table to be unloaded from the memory. It can be 0 ~ 9, where 0 means not-unloadable and 9 means the earliest unload. The standard unload priority of the tables stored in the memory is 5, which means in case of need to free some memory space, the archived data would be unloaded to disks preferably.
For allowing OutBoard users to define the Unload priority other than the default value, a setting is available. In the Near-line objects tab of OutBoard with Storage Management when running on SAP HANA settings, the user should have the "Unload Priority of Archive Table" option. Here the user may define the new Unload Priority value.

_{Near-line objects tab with SAP HANA settings}

By clicking on the "Set Unload Priority" button, the user may specify for which InfoProviders this setting will be updated. When using OutBoard without storage management, this value is one of the parameters that can be found in the settings.

_{HNA_UNLOAD_PRIO setting}

To activate this value, the user should run the report /DVD/NLS_HNA_SET_UNLOAD_PRIO. By running this report, the user may define specific InfoProviders for which the new Unload Priority value should be changed.

_{Set Unload Priority for Table on HANA}

All archived data are divided into SAP HANA partitions based on NLS Request ID using range partitioning. Range partitioning creates dedicated partitions for certain values or certain value ranges. Range partitioning is used to actively manage the partitioning of a table; partitions may be created or dropped individually.

Language support

English and German are currently supported.

Size-Dependent License

If using OutBoard with a size-dependent license: As of spring release in May 2013, the license can be also size-dependent. A size-dependent license is based on the agreement, prior to the Outboard installation.

The Outboard consultant will provide information about the Outboard license size and validity of the system. In most cases, the license is based on the size that is occupied in the system.
There are various places within Outboard to find information and monitor license size:
1. By checking the Outboard connection (e.g. in TA rsa1 in DAP Display mode or Archiving Tab of an InfoProvider

_{License size - Outboard Connection}

2. By the creation of a new DAP or in Logs of an archiving process.

_{License Size / Status - Log}

Setting a Size Threshold warning for OutBoard

In OutBoard Settings there is a new parameter corresponding to the new licensing of OutBoard. This parameter is called SIZE_DIFF_THRESHOLD. Here the threshold size for warnings can be defined. When the free size within the system passes the threshold, a warning will be triggered during each archiving process or creation of DAP.

Reaching Outboard storage size limit.
In the case of reaching the storage limit of Outboard, it will no longer be allowed to archive any additional data. However, all data that was archived previously within the limit size will still be accessible. If the size limit is reached, it is recommended to contact your OutBoard consultant in order to discuss how to proceed.