https://issues.apache.org/SpamAssassin/show_bug.cgi?id=5590

--- Comment #48 from Quanah Gibson-Mount <[email protected]> ---
Thanks to the amavis-logwatch script, I have some nice timing data for rules
processing at my site as well:

========================================================================================================================
SA Timing Percentiles         % Time    Total (ms)        0%        5%      
25%       50%       75%       95%      100%
------------------------------------------------------------------------------------------------------------------------
tests_pri_0 (7650)            84.64%    5165461.00     88.00    233.00   
338.00    402.00    534.00   1759.40  24756.00
check_pyzor (7650)            28.59%    1744883.00     31.00    139.00   
145.00    151.00    158.00    218.00  19208.00
check_razor2 (7650)           19.22%    1173107.00     24.00     27.00    
94.00    132.00    162.00    209.00  14023.00
learn (3985)                   5.91%     360940.90      3.40      7.00    
28.00     38.00     60.00    248.80   3674.00
b_learn (3985)                 5.42%     330508.35      1.49      2.20    
25.00     34.00     52.00    204.00   3481.00
tests_pri_-400 (7650)          4.66%     284643.00      7.00     12.00    
17.00     22.00     34.00     87.00  14435.00
check_bayes (7650)             4.49%     273827.00      7.00     11.00    
16.00     21.00     32.00     84.00  14434.00
extract_message_metadata (7650)   1.74%     106200.12      0.93      1.99     
3.90      8.00     14.00     45.00    528.00
b_tokenize (7650)              1.61%      98333.80      1.85      3.20     
4.80      7.00     12.00     33.00    531.00
b_count_change (3605)          1.40%      85147.90      3.20      8.00    
11.00     16.00     24.00     47.00   2919.00
get_uri_detail_list (7650)     1.00%      60895.61      0.08      0.64     
1.76      2.90      7.00     23.00    509.00
b_tok_get_all (7650)           1.00%      60866.52      1.51      3.00     
4.30      6.00      8.00     19.00    699.00
b_comp_prob (7650)             0.79%      48176.27      1.16      2.20     
3.60      4.70      7.00     14.00    123.00
b_finish (7650)                0.52%      31588.31      0.24      0.46     
0.52      0.56      0.63      0.96  14408.00
parse (7650)                   0.42%      25687.28      0.73      1.20     
1.79      2.40      3.50      9.00     40.00
tests_pri_-1000 (7650)         0.41%      25285.90      1.95      2.20     
2.60      2.80      3.10      6.00     42.00
tests_pri_500 (7650)           0.41%      24944.14      1.66      2.20     
2.50      2.90      3.20      6.00     31.00
expire_bayes (2)               0.40%      24705.00  10441.00  10632.15 
11396.75  12352.50  13308.25  14072.85  14264.00
tests_pri_-900 (7650)          0.18%      11028.16      0.93      1.02     
1.16      1.21      1.29      1.84     40.00
tests_pri_-950 (7650)          0.16%      10020.39      0.77      0.86     
0.98      1.05      1.36      1.73     22.00
get_report (7650)              0.09%       5729.38      0.25      0.41     
0.48      0.56      0.69      1.48     25.00
check_spf (7650)               0.05%       3037.06      0.18      0.21     
0.24      0.27      0.46      0.65     27.00
b_tok_touch_all (7650)         0.02%       1254.65      0.04      0.06     
0.09      0.13      0.19      0.25     14.00
------------------------------------------------------------------------------------------------------------------------
Total                                   6103202.00    118.00    293.00   
395.00    475.00    667.00   2086.70  25856.00
========================================================================================================================

We can see that test_pri_0 eats up an inordinate amount of time comparatively.

-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to