Our implementation of bitemporal database is based on KDB (produced by Kx systems www.kx.com) . KDB is a database system implemented on top of the K language environment, a fast array language. Our implementation takes advantage of the feature of ordered data in KDB. The version we are using is K2.8/KDB2.2. On a Dell Precision 410 PC (PIII 550MHz) with 128M of RAM, we ran the spytime benchmark at different scales and produced the result of table 1.
Table 1. SpyTime benchmark results (in seconds)
|
Scale |
1 |
10 |
100 |
|
Number of records |
10,000 |
100,000 |
1,000,000 |
|
Query 1 |
0.006 |
0.012 |
0.08 |
|
Query 2 |
0.011 |
0.031 |
0.25 |
|
Query 3 |
0.008 |
0.024 |
0.21 |
|
Query 4 |
0.009 |
0.031 |
0.29 |
|
Query 5 |
0.009 |
0.043 |
0.46 |
|
Query 6 |
0.011 |
0.063 |
0.58 |
|
Query 7 |
0.012 |
0.066 |
0.59 |
|
Query 8 |
0.022 |
0.107 |
0.80 |
|
Query 9 |
2.6 |
2.7 |
2.7 |
|
Query 10 |
0.052 |
0.021 |
0.14 |
|
R |
0.021 |
0.056 |
0.38 |
Last updated May,2001