When Should You Use the Scalable Performance Data Engine (SPDE)?

[#1]

SPD Engine or SPD Server?

The SPD Engine and the SPD Server product share a common heritage and, therefore, share a great number of features and performance benefits. However, there are some important differences, primarily in the execution environment.

SPDE should be considered an entry-level scalable product. It runs as a libname engine in the SAS environment. SPD Server is a standalone client/server product. Applications initially developed on SPDE can be migrated to SPD Server with ease, as the need to move to a full client/server environment arises. Compared to SPDE, SPD Server:

• requires its computer be a mostly dedicated server; the more dedicated, the better.
• requires more skills to set up and administer than SPDE.
• supports multi-user client/server access.
• supports an Access Control List–based security model.
• supports the SQL functions:
  • parallel BY-group (PBG) processing.
  • implicit pass-through.
• is not available on Linux (LNX), OS/390 (MVS), HP/UX for the Itanium Processor Family (H6I), or OpenVMS Alpha (ALP).

For more information, visit the SPD Server Learn & Support page. 

SPDE Engine or Base Engine?

SPDE is optimized for the storage and sequential access of large and very large data sets (millions of rows, many GB of data). For medium to small data sets, the base engine is often a better performer. Compared to the base engine, SPDE:

• supports more than 32K columns in SAS 9 and later. The base engine supports more than 32K columns in SAS 9.1 and later.
• is the only SAS engine that supports more than 231 - 1 (approximately 2 billion) rows on 32-bit hosts.
• supports the implicit sort for BY processing.
• supports optimization of the WHERE expression with multiple indexes.
• supports optimization of the WHERE expression containing OR.
• supports partitioned data sets.
• locks at the member level; the base engine locks at the record level.
• requires an index-reorganization utility to rebalance the index tree.
• does not support some of the base engine features:
  • utility (byte) files.
  • catalogs.
  • views.
  • MDDBs.
  • integrity constraints.
  • data set generations.
  • CEDA.
  • audit trail.
  • options for national language support.

Content originally published in 2003.
If you're looking for a solution for advanced analytics and real-time data processing, see SAS with SingleStore.