June 29, 2014

esProc/Oracle: Single Machine Performance Comparison Test(II)

5 Use Cases Description

For better understanding, all test logic will be described in SQL.

During the test Oracle will execute the SQL statement directly, whereas esProcwill be running the equivalent code we write to complete the same computation.

5.1 Use Case for Large Data Volume Scan
This use case is large task single machine test. It’s designed to test the computation performance for whole table scan with large data volume. Performance for whole table counting, integer sum, float sum, value sum, integer filter, number filter, character filter, date filter are considered respectively.

SQL as shown below ,please Click here to view the full table:



















5.2 Use Case for Large Data Volume Grouping

This use case is large task single machine test. It’s designed to test the grouping computation performance with large data volume, where grouping result record number is small, smaller than the physical memory. Performance for integer grouping, number grouping, character grouping and date grouping are considered respectively.

SQL as shown below ,please Click here to view the full table:
















5.3 Use Case for Large Data Volume Large Grouping

This use case is large task single machine test. It’s designed to test the grouping computation performance with large data volume, where grouping result record number is large, larger than the size of the physical memory, and thus the computation cannot be completed within the memory.















5.4 Use Case for Large Data Volume Join

This use case is large task single machine test. It’s designed to test the external join computation performance with large data volume. Performance for integer, number, character join is considered respectively, including single-tier join and milti-tier join.

SQL is shown blow,please click here to view the full table.


5.5 Use Case for Small Data Volume Concurrent Test

This use case is to test the multi-task concurrent computation performance with small data volume. Each task of the concurrent process is accessing different physical tables/files to avoid OS cache. Tests were done for scanning, grouping and joining, with the same use cases as 5.1, 5.2, 5.4 above. However the data volume is small (See data volume part).

Tests were done for single task, dual tasks concurrent, and four tasks concurrent scenarios respectively.

Related:
esProc/Oracle: Single Machine Performance Comparison Test(I)

esProc/Oracle: Single Machine Performance Comparison Test(III)

esProc/Oracle: Single Machine Performance Comparison Test(IV)

Please Click here to download the full version。

No comments:

Post a Comment