 |
|
|
SyncSort can reduce the number of passes through the data by speeding up or combining these steps:
- Data Extraction - SyncSort's INCLUDE/OMIT quickly identifies the exact records you need, reducing the total data to be processed.
- Data Conversion - FilePort, a related product, converts mixed-type mainframe data to UNIX format (or vice versa) without the ballooning of datasizes or destruction of record layout that comes from unpacking mainframe data.
- Data Transformation - SyncSort's REFORMAT quickly changes your legacy records into load record images.
- Summarization - SyncSort's SUMMARIZE aggregates records with duplicate sort keys into a single record, counting records and totaling numeric fields.
- Join - SyncSort's JOIN takes records from two different sources and creates target records containing data from both sources.
- Data Partitioning - SyncSort's multiple OUTFILE splits your data into separate partitioned table ranges.
- Load Pre-Sorting - SyncSort's world-record sort performance accelerates the database load itself.
If your relational database is very large, a lot of time can be spent loading data, extracting data for handoffs or reports, and running reorgs. SyncSort can cut significant time from these operations.
- Bulk loading for high performance relational databases like Oracle and Sybase can be accelerated. Presorting data with SyncSort prior to bulk loading can cut 50% or more from the average load time.
- Where a bulk load is not appropriate, loads can still be accelerated by presorting the data, using the clustered index as the sort key.
- Extract processing can also be accelerated. SyncSort's SUMMARIZE and KEYS functions are much faster alternatives to complex SQL SELECT, GROUP BY, SUM and ORDER BY statements.
- The DB2 accelerator speeds database load functions on DB2 databases.
- Output record numbering adds unique record identifiers to output data. This simplifies data loading for database tables that have uniqueness constraints.
Database reorgs will run significantly faster when database utilities are only used to unload and reload data, and SyncSort is used for the sort.
COBOL applications frequently use flat or indexed files, which require a great deal of sorting. SyncSort makes these sorts run faster whether they are hidden inside programs using the COBOL SORT verb or are done in shell scripts using the UNIX sort command.
- SyncSort can improve the performance of COBOL applications by more than 65%. SyncSort's COBOL Sort Accelerator automatically replaces COBOL SORT and MERGE verbs with SyncSort, resulting in major performance gains.
- Significant time can be saved by using SyncSort instead of system utilities. A compatible interface lets you sort with SyncSort, without editing your shell scripts. In addition, SyncSort's DATADICTIONARY feature allows you to reference fields in your data files by their COBOL names, instead of by position, length, and format.
- SyncSort's indexed file access capability lets you sort files easily at the command prompt, whether they are Micro Focus indexed, ACUCOBOL, C-ISAM or UniKix VSAM format files.
To request a free Proof-of-Concept, more information, or to order SyncSort, please contact us.
|
|
|
