Closed emaliv closed 7 years ago
What's the performance impact of this change.. ?
The performance impact is negligible. You can find below a comparison between the default sql query, without dash and null, and the sql query proposed.
$sqlite3 /NVDS/nvd.db sqlite> .timer on sqlite> .stats on
sqlite> SELECT ID FROM PRODUCTS WHERE PRODUCT = 'linux_kernel' AND (VERSION = '3.8.0' OR VERSION IS NULL OR VERSION = '-') COLLATE NOCASE;
sqlite> SELECT ID FROM PRODUCTS WHERE PRODUCT = 'linux_kernel' AND VERSION = '3.8.0' COLLATE NOCASE;
Memory used (with NULL, dash) | Runtime (with NULL, dash) | Memory used | Runtime |
---|---|---|---|
2620280 bytes | real 0.047 | 2619784 bytes | real 0.055 |
2620280 bytes | real 0.062 | 2619784 bytes | real 0.033 |
2620280 bytes | real 0.096 | 2619784 bytes | real 0.073 |
2620280 bytes | real 0.076 | 2619784 bytes | real 0.058 |
2620280 bytes | real 0.090 | 2619784 bytes | real 0.041 |
2620280 bytes | real 0.074 | 2619784 bytes | real 0.045 |
2620280 bytes | real 0.055 | 2619784 bytes | real 0.034 |
Added CVEs with NULL or dash at product version into the report. These CVEs should be taken into account because in some cases they are valid for the scanned product.