(SP21) FAQ

Owned by SNP
Apr 13, 2018

What is NLS?
Nearline storage ("near-online") is used to describe an intermediate type of data storage that represents a compromise between online storage (supporting frequent, very rapid access to data) and offline storage/archiving (used for backups or long-term storage, with infrequent access to data). Both archiving and Nearline allow a reduction of database size that results in improved speed of performance for the online system. However, accessing archived data is more complex and/or slower than is the case with Nearline storage, and can also negatively affect the performance of the main database, particularly when the archive data must be reloaded into that database.


What is so special about Datavard OutBoard?
Datavard's "OutBoard for SAP® BW" provides near-line NLS storage functionality with reporting capabilities without the need of any additional database or software, thus keeping the system landscape simpler and easier to maintain. OutBoard uses compression technology which is not only very fast, but which can easily compete with external near line systems in terms of compression ratio - OutBoard can provide a ratio of up to 20:1! Finally, the OutBoard Cockpit offers analytic functions to give transparency as to the data volume (both gross and net) and the compression ratio for the data, which is stored in the OutBoard NLS tables.


Is it time-consuming to install OutBoard?
Not at all – the installation is a simple SAP Add-on that needs to be imported. 


Is it time-consuming to use OutBoard?
It is not time-consuming: OutBoard features an administrator cockpit that is fully integrated into the SAP BW Backend. This cockpit facilitates all tasks related to using OutBoard. For end-users, OutBoard is 100% transparent and they don't even notice whether data in reporting comes from the online database or from OutBoard.


What is the typical compression ratio?
Compression ratio differs depending on InfoProvider design, type of data and the database system. Typical rates are between 1:8 and 1:20.


How is performance in reporting on data in OutBoard?
Typically performance is comparable to normal reporting – but same as in the online database, speed heavily depends on data volumes and type of queries. However, we gladly assist in optimizing query performance based on indexes, aggregates and other standard BW performance measures.


Why the archiving performance in copy phase may be influenced in Outboard DataTiering SP19?
When the parameter CHECK_PACKAGE_SIZE is by default set to "2147483648", it checks during archiving the maximum size of each package according to semantic group(s). In the case many characteristics are selected as semantic groups in the Data archiving process, it may have a negative effect on performance. This check can be turned off by changing the parameter to an empty value.


How are support and updates handled for OutBoard?
Support is available via our hotline and our Customer Support Center on the Internet. From the Support Center, customers can download updates and additional materials.


What Objects are suitable for Semantic Groups in DAP?
It is recommended to choose the Objects used for filtering in reporting. At most 5 InfoObjects should be added to Semantic Group. As first, an InfoObject should be added with the lowest cardinality, at last the one with the highest cardinality. In case a Semantic Object is too big (more than 200.000 rows), the chosen InfoObjects should be reconsidered again.


It is reasonable to limit the Size of Data Package in DAP?
Yes, Data Package is the smallest entity processed in Queries and Loads. Since the Data in InfoCubes is accessed more often, we recommend restricting this size with 50MB. For DSO we recommend 100MB.


Can I load data from OutBoard back into the online database?
Yes, you can. See chapter "Reloading data" to learn more.


Can I load data from OutBoard back into any info provider in the online database, just like in datamarts?
Yes, you can. See chapter "Loading data" to learn more.


What happens if I change the structure of the InfoProvider?
If new fields are added to an InfoProvider, the DAP has to be reactivated. In order to add new data to the archive, the archived data has to be reloaded into this new structure, updated and archived again.


Can I add the NLS data in reporting automatically?
Yes, but the InfoProvider or Queries have to be specified. See the chapter "Reporting on archived data" to learn more


In SAP BW Release less that 7.3 Virtual Providers have to be built to include archived data in reporting on MultiProviders. Is there a smart way, how to build these VirtualProviders automatically?
Yes, see the chapter "Reporting on archived data" to learn more. Using report /DVD/NLS_VP_GUI the Virtual InfoProviders can be created automatically with all needed adjustments.


Do I have to schedule update of Aggregates in Process chains, when new data is archived?
No, aggregates are updated automatically. If you want to change the Field Selection for Indexes, see chapter "Field selection for InfoProvider" to learn more.


How should I setup DAP correctly?
You can find the detailed information in the OutBoard best practices document. Very important is to choose adequate Time Slice Characteristic, appropriate Objects for Semantic groups and limit Size of Data Package.


Can I parallelize the archiving by running more archiving request for one InfoProviders at once?
Yes, but not to the deletion phase. Parallelization of more archiving requests is only done for requests created to verification phase. Deletion phase cannot be parallelized.


How do I find detailed logs for DAP creation in Mass Archiving?
In logs for Mass Archiving you can find the name of Background Job where the DAP creation was started. You copy the name of the BG Job and in TA SM37 you can see the detailed logs for the Job. The SAP note 46272 could help you to create a new tablespace and connect it with data class. Sometimes, it is needed to reset corresponding table buffers (e.g. TAORA for Oracle) to make system aware of the changes.


Can we talk to another Datavard customer before we start with a project?
Yes, of course you can. However, a reference customer call needs to be planned in advance. A lead-time of two to three weeks is required to allow for careful planning without disrupting any of our customer's business or projects.


Example: How to implement a BAdI to define variants?
You can create and activate a BAdI implementation for the BAdI definition /DVD/NLS_INTIAL_RUN_VARIANTS for particular InfoProvider. An example is the following:

After the creation of initial run you can see the variants to be created in Mass archiving cockpit.  


After execution of the initial run you can see which archiving requests are done in Mass archiving cockpit or via RSA1.



Datavard Ltd. was founded in London by a group of IT professionals with extensive knowledge of the SAP market and practical experience with implementing SAP products on a national and international scale. They noticed a growing gap as organizations where struggling to overcome issues created from using SAP Systems. As a result, Datavard was created and is growing dynamically, broadening its business to three different locations. We are constantly developing new tools and training more consultants to help more customers.

In addition to NLS for BW, Datavard also engages in System Landscape Optimization (SLO) and Data Quality Optimization for SAP systems.