To use Rule or Preprocessor profiling, you must build snort using the

# –enable-perfprofiling option to configure.

Snort FAQ

README.PerfProfiling

To use Rule or Preprocessor profiling, you must build snort using the

# –enable-perfprofiling option to configure.

Rule Profiling Configuration

# # syntax: # config profile_rules: print [all | num], sort sort_option, filename file_option [append] # - where num is the number of rules to print # - where sort_option is one of: # checks, matches, nomatches, avg_ticks, # avg_ticks_per_match, avg_ticks_per_nomatch, total_ticks # - where file_option is the output filename # - where append dictates that the output will go to the same file each time (optional) # # examples: # # 1) Print all rules, sort by avg_ticks (default configuration # if option is turned on) # config profile_rules # # 2) Print the top 10 rules, based on highest average time # config profile_rules: print 10, sort avg_ticks # # 3) Print all rules, sorted by number of checks # config profile_rules: print all, sort checks # # 4) Print top 100 rules, based on total time # config profile_rules: print 100, sort total_ticks # # 5) Print with default options, save results to performance.txt each time # config profile_rules: filename performance.txt append # # 6) Print top 20 rules, save results to perf.txt with timestamp in filename # config profile_rules: print 20, filename perf.txt

When you use this (see configuration info with example output), snort will print the following table at exit. The column info is: - Number (rank) - Sig ID - Generator ID - Checks (number of times rule was evaludated after fast pattern match within portgroup or ANY->ANY rules) - Matches (number of times ALL rule options matched, will be high for rules that have no options) - Alerts (number of alerts generated from this rule) - CPU Ticks - Avg Ticks per Check - Avg Ticks per Match - Avg Ticks per Nonmatch

Example, using config profile_rules: print 4, sort total_ticks Rule Profile Statistics (worst 4 rules) ========================================================== Num SID GID Rev Checks Matches Alerts Ticks Avg/Check Avg/Match Avg/Nonmatch === === === === ====== ======= ====== ===== ========= ========= ============ 1 2389 1 12 1 1 1 385698 385698.0 385698.0 0.0 2 2178 1 17 2 0 0 107822 53911.0 0.0 53911.0 3 2179 1 8 2 0 0 92458 46229.0 0.0 46229.0 4 1734 1 37 2 0 0 90054 45027.0 0.0 45027.0

Interpreting this info is the key… The Ticks column is important because that is the total time spent evaluating a given rule. But, if that rule is causing alerts, it makes sense to leave it alone.

A high Avg/Check is a poor performing rule, most likely contains PCRE. High Checks and low Avg/Check is usually an ANY->ANY rule with few rule options and no content. Quick to check, the few options may or may not match. We are looking at moving some of these into code… Especially those with low SIDs.

By default, this information will be printed to the console when Snort exits. You can use the “filename” option in snort.conf to specify a file where this will be written. If “append” is not specified, a new file will be created each time Snort is run. The filenames will have timestamps appended to them. These files will be found in the logging directory.

Preprocessor Profiling Configuration

# # syntax: # config profile_preprocs: print [all | num], sort sort_option, filename file_option [append] # - where num is the number of rules to print # - where sort_option is one of: # checks, avg_ticks, total_ticks # - where file_option is the output filename # - where append dictates that the output will go to the same file each time (optional) # # examples: # # 1) Print all preprocessors, sort by avg_ticks (default configuration # if option is turned on) # config profile_preprocs # # 2) Print the top 10 preprocessors, based on highest average time # config profile_preprocs: print 10, sort avg_ticks # # 3) Print all preprocessors, sorted by number of checks # config profile_preprocs: print all, sort checks # # When printing a specific number of preprocessors all subtasks info # is printed for each layer 0 preprocessor stat.

When you use this (see configuration info with example output), snort will print the following table at exit. The column info is: - Number (rank) The number is indented for each layer. Layer 1 preprocessors are listed under their respective caller (and sorted similarly). - Preprocessor Name - Layer - Checks (number of times preprocessor decided to look at a packet, ports matched, app layer header was correct, etc) - Exits (number of corresponding exits – just to verify code is instrumented correctly, should ALWAYS match Checks, unless an exception was trapped) - CPU Ticks - Avg Ticks per Check - Percent of caller For non layer 0 preprocessors – i.e., subroutines within preprocessors, this identifies the percent of the caller’s ticks that is spent for this subtask.

Example, using config profile_preprocs: print 10, sort total_ticks Preprocessor Profile Statistics (worst 10) ========================================================== Num Preprocessor Layer Checks Exits Microsecs Avg/Check Pct of Caller Pct of Total === ============ ===== ====== ===== ========= ========= ============= ============ 1 detect 0 338181 338181 9054573 26.77 64.62 64.62 1 rule eval 1 256978 256978 2570596 10.00 28.39 18.35 1 rule tree eval 2 399860 399860 2520629 6.30 98.06 17.99 1 pcre 3 51328 51328 505636 9.85 20.06 3.61 2 byte_jump 3 6 6 7 1.30 0.00 0.00 3 content 3 1077588 1077588 1123373 1.04 44.57 8.02 4 uricontent 3 106498 106498 79685 0.75 3.16 0.57 5 byte_test 3 9951 9951 5709 0.57 0.23 0.04 6 isdataat 3 8486 8486 3192 0.38 0.13 0.02 7 flowbits 3 135739 135739 35365 0.26 1.40 0.25 8 flags 3 2 2 0 0.20 0.00 0.00 9 preproc_rule_options 3 15499 15499 1939 0.13 0.08 0.01 10 flow 3 394817 394817 36420 0.09 1.44 0.26 11 file_data 3 15957 15957 1264 0.08 0.05 0.01 12 ack 3 4 4 0 0.07 0.00 0.00 2 rtn eval 2 36928 36928 17500 0.47 0.68 0.12 2 mpse 1 646528 646528 5840244 9.03 64.50 41.68 2 s5 0 310080 310080 3270702 10.55 23.34 23.34 1 s5tcp 1 310080 310080 2993020 9.65 91.51 21.36 1 s5TcpState 2 304484 304484 2559085 8.40 85.50 18.26 1 s5TcpFlush 3 22148 22148 70681 3.19 2.76 0.50 1 s5TcpProcessRebuilt 4 22132 22132 2018748 91.21 2856.11 14.41 2 s5TcpBuildPacket 4 22132 22132 34965 1.58 49.47 0.25 2 s5TcpData 3 184186 184186 120794 0.66 4.72 0.86 1 s5TcpPktInsert 4 46249 46249 89299 1.93 73.93 0.64 2 s5TcpNewSess 2 5777 5777 37958 6.57 1.27 0.27 3 httpinspect 0 204751 204751 1814731 8.86 12.95 12.95 4 ssl 0 10780 10780 16283 1.51 0.12 0.12 5 decode 0 312638 312638 437860 1.40 3.12 3.12 6 DceRpcMain 0 155358 155358 186061 1.20 1.33 1.33 1 DceRpcSession 1 155358 155358 156193 1.01 83.95 1.11 7 backorifice 0 77 77 42 0.55 0.00 0.00 8 smtp 0 45197 45197 17126 0.38 0.12 0.12 9 ssh 0 26453 26453 7195 0.27 0.05 0.05 10 dns 0 28 28 5 0.18 0.00 0.00 total total 0 311202 311202 14011946 45.03 0.00 0.00

Because of task swapping, non-instrumented code, and other factors, the Percent of caller field will not add up to 100% of the caller’s time. It does give a reasonable indication of how much relative time is spent within each subtask.