Analyzer

The function of SAP BW InfoProvider Analyzer helps the user to analyze which data is most suitable for archiving. Data unsuitable for archiving would be data of the current fiscal year, or data from very small tables, whose size would not be reduced significantly upon archiving. Data that is suitable for archiving would be data older than current fiscal years or in tables that range in size of several GB.

The advantages of using OutBoard are that:

Data can be reloaded into their former InfoCube or a DSO from OutBoard.
Data can be loaded into a different InfoCube or a DSO from OutBoard.
Reporting can be done on data archived by OutBoard.

_{Introductory screen to InfoProvider Analyzer}

The InfoProvider analyzer provides information about the DataStore Objects. Select which DSO(s) and/or InfoCube(s) to be included in the Run Name.
The SAP BW InfoProvider Analyzer allows you to Display (F5), Create (F6), Copy (F7) and Delete (F8) a particular Run Name.
⇒ To display a Run Name, type the name of the Run Name or press F4 to select values from a list, and then press "Display".
When the data in the DSOs / InfoCubes you have selected is changed users can create new Run Names to reflect these changes.
⇒ To create a Run Name, press "Create" and fill in the following attributes:
When creating or displaying a particular Run Name, you will be presented with the following parameters.

_{Creating a Run Name}

Creation of run is possible also in Background, after clicking on Create button there will be an option In dialog or In Background. The run can be displayed when background processing is finished.

To copy a Run Name, press "Copy" and fill in these attributes:

_{Copying a Run Name}

"Description" is optional but highly recommended to fill out. To delete a Run Name, type the name of the Run Name or press F4 to select values from a list, then press, "Delete". You will be prompted whether you wish to delete this Run Name.

_{Deleting a Run Name}

Analyzer Run view

When creating a Run Name, tables of InfoProviders are selected and the statistics of their tables (database statistics) such as table size, count of rows, etc. are used to estimate the size of an archived object.

_{Parameters of objects analyzed in the Analyzer}

InfoProvider – Name of the InfoProvider.
Technical name – Technical name of the InfoProvider
Selection – Selection that is used for the request in archiving.
Type – Can be either a DSO or an InfoCube (designated by their specific icons). Data contained in a DSO is drawn from active tables and data in InfoCubes is from fact tables. Their current statistics do not have to correspond to their actual statistics.
Table size – Size of the entire table.
Possible values: KB, MB, and GB
Default value: GB
Indexes size – Size of the indexes of the analyzed tables. (For column-oriented tables on HANA is this column empty)
Possible values: KB, MB, and GB
Default value: GB
Count rows – How many rows does the table contain?
Total size – this is the total size of the table and includes the size of the table's indexes.
Possible values: KB, MB, and GB
Default value: GB
Compressed size – Expected size of the table after archiving.
Possible values: KB, MB, and GB
Default value: GB
Calculate ratio – In OutBoard Analyzer, the ratio of compression is a fixed value of 90% unless recalculation will be started. In case a DAP is created for an InfoProvider, an archive test of 100.000 rows will be executed and the correct ratio will be calculated and used in the analysis besides the constant.

Ad-hoc analysis

The selection allows the user to create an ad-hoc analysis of the data in the InfoProvider i.e. to determine the count of rows for any given selection. To create an ad-hoc analysis, click on the gray arrow.
Press:

"Automatic analyze " button and select all InfoObjects which should be used in the analysis. The InfoProvider is selected in the lower part of the screen.

_{InfoObjects for Analysis}

In the following screen you may restrict the values of some of the InfoObject to predefined value ranges or values:

_{Define ranges for Analysis}

In the last third screen define, which of the InfoObjects chosen in the first screen should be used as 'distinct' and which not. If you mark an InfoObject as 'Use distinct', in the final result, there will be a result row for each possible value of this InfoObject, if it fits the value ranges pre-defined in the previous screen. This option is beneficial if e.g. you want to find the data distribution for each calendar month in a pre-defined calendar year.
If multiple InfoObjects are chosen here, the analysis result will be a row for each possible combination of values for chosen InfoObjects.

_{Distinct InfoObjects values for Analysis}

The field "Maximum selection" represents the maximum number of analyses, which will be created (default 100).

_{Automatic analysis result}

Column headings:

Number of selection – Unique ID of selection
Selection – Selected selection can be shown when clicking on the icon
Count of records – Number of records in this particular selection.
Size – Size of this particular selection.
Possible values: KB, MB, and GB
Default value: GB

Expected compressed size – Expected size of the selection after archiving.
Possible values: KB, MB, and GB
Default value: GB

Saved money (EUR) – Money saved in EUR based on archiving this particular selection.

Press:
"Delete button" to delete selected analyses. The user can delete each analysis separately or choose the option "Yes All" and delete all selected analysis

Old data Analysis

Old data Analysis

Before archiving the InfoProvider, it is recommended that the user should know as much as possible, such as the distribution of the data in InfoProvider, data growth, reporting characteristics, filters used for InfoProvider, and flow of the data into InfoProvider.
The purpose of this analysis is to help the user by providing information regarding old data coming into the system by analyzing current requests.

Without the NLS writer old data is problematic, as it could cause the activation or insertion to fail. Even when using the NLS Writer, it is not recommended to choose time characteristics, where data is changing on a high frequency. Therefore as a result of this analysis will be also recommended characteristic for DAP time slice definition.
To start the analysis run TA: SE38 and execute the report /DVD/NLS_ANA_OLD_DATA.

_{Old Data Analysis input screen}

Screen options:

Run ID – Name of analysis run

Option Buttons:

Start Analysis – For starting a new analysis under a specified Run ID
Display Results of Analysis – For displaying results of previous analysis under specified Run ID
Delete Analysis – For deleting results from previous analysis for specified Run ID

Fill parameters manually:

InfoProvider Tech. Name – Technical name of the InfoProvider, which is to be analyzed
InfoObject Tech. Name – Technical name of an InfoObject, which is going to be analyzed in requests
Fiscal year variant – Specification of Fiscal year variant
Archived records older than – Definition of the date that will be taken as a boundary between old and new data
Requests in last XX days – Specify what request is relevant for analysis (based on request date)
Detailed Analysis – When checked, the results of the analysis will be displayed based on InfoProvider, InfoObject, and requests, otherwise it is based only on InfoProvider and InfoObject.

Fill parameters from the table:

Using this option, the user is available to specify all parameters using the database table /DVD/NLS_ANA_OIP.

RUNID – Name of analysis run
INFPRV – Technical name of the InfoProvider
IOBJ – Technical name of an InfoObject
FISCV – Specification of Fiscal year variant
DATA_RET_TIME – Definition of the date that will be taken as a boundary between old and new data
LAST_DAYS – Specify what request is relevant for the analysis

_{Table view in Data browser}

Analysis example:
Let's say 3 InfoCubes are to be analyzed based on all InfoObjects from the Time dimension. Enter a new Run ID that will represent this group and choose the "Start Analysis" option button. During the first step these InfoCubes in "InfoProvider Tech. Name" should be listed. The next step is to add all InfoObjects from Time dimensions in "InfoObject Tech. Name". It is mandatory to specify "Fiscal year variant", in case there are "fiscal objects" in the Time dimension. Continue with choosing a boundary date for defining the age of data.
In the field "Requests in last XX days" enter the number of days for which will be the requests checked retrospectively, then execute the analysis (F8).
Output example

_{Locked data analysis}

History:
To display results from previous analysis runs, choose the "Display Results of Analysis" option button, then select Run ID from search help (F4) and execute (F8). The "Detailed analysis" checkbox option allows you to display the results in request-level detail.
For deleting the results of previous runs, select the option "Delete Analysis", specify a Run ID, and execute.
Restrictions:
The analysis logic is designed to scan for data in the Changelog of the DSO and F-table of the InfoCube. Therefore compressed data in E-tables are not counted as affected.

Deletion of requests

The Archiving request can be deleted by calling the report /DVD/NLS_DELETE_REQUEST.

_{Request deletion screen}

The following parameters can be specified:

OutBoard Object name – Name of Object(s) for which requests are to be deleted.
Delete by Request – Press F4 to display request(s) that are to be deleted.
Delete by Date (absolute) – Type in the desired date of request creation or press F4 to select a date of request creation from the calendar.
Delete by Date (relative) – Specify a number and select the Time period (Day, Week, Month, or Year) of request creation. Calculated from the current date.

To include InfoCube or DSO request "InfoCube, DSO" must be selected.
To include PSA or Change Log request "PSA, Change Log" must be selected.

_{Delete OutBoard request screen}

When clicking on the "Delete" button only checked Objects will be deleted in the background job.
Support of standard SAP deletion of archiving requests is described in the SAP Note 1571882. If your BW release is supporting the standard deletion, on the Archiving tab of InfoProvider, after double-clicking on archiving request, you should be able to see the option "Delete Near-Line Request".

_{Process request}

Automatic deletion of requests

For automatic deletion of requests the parameter "Delete by date (relative)" has to be specified. Variant with this setting will need to be set up. Automatic deletion of requests can be done by scheduling the report /DVD/NLS_DELETE_REQUEST periodically using the particular variant.

It is recommended not to use this functionality. When an Outboard Archiving request is deleted the Data (archived Data) within the request is also deleted.

For the automatic deletion of requests, there are a few parameters to be specified. Fill in the object name, select the type of InfoProvider (InfoCube, DSO or PSA, Change Log), and decide which option is the most suitable one:

Delete by Request number
Delete by Date – relative
Delete by Date – absolute

Each of these options has to be specified.

Drop NLS Data

Data stored in NLS can be dropped by calling report /DVD/NLS_DROP_DAP.

_Drop DAP

When executing this report, all data stored in NLS will be dropped and the requests in the archive tab of InfoProvider will be marked as invalid. This report will also delete all the locks on the archived area. Running a report in "Test mode" will not delete the data. By checking the checkbox "Delete DAP", the Data archiving process will be deleted with NLS data. If the option “Recreate DAP is checked, DAP is recreated after DAP deletion.

“Ignore Connection” allows deleting DAP when the storage is inaccessible.

“CTO Check” allows deleting DAP, recreating DAP, and reloading arch. request on a closed system.

“Force mode” will execute DAP deletion with force mode ON.

Logs from the deletion can be found in the transaction SLG1 under the object /DVD/NLS_OTB and the subobject /DVDV/NLS_RELOAD_LOG. It is also possible to drop the data via the Process chain. In this scenario, class /DVD/NLS_DROP_DATA will need to be used in this process type. In the case of problematic behavior read Authorizations and Access part of this User Guide, to check if your user has enough access privileges.

_{Drop the Data}

Browser not supported