* Ralf Hildebrandt <[email protected]>: > Hi there! > > TIMING-SA is being logged like this: > > amavis[31273]: (31273-08) TIMING-SA [total 1354 ms, cpu 542 ms] - parse: 3.3 > (0.2%), extract_message_metadata: 76 (5.6%), > get_uri_detail_list: 6 (0.4%), tests_pri_-1000: 68 (5.0%), tests_pri_-950: > 0.92 (0.1%), tests_pri_-900: 0.96 (0.1%), > tests_pri_-400: 37 (2.7%), check_bayes: 35 (2.6%), b_tokenize: 16 (1.2%), > b_tok_get_all: 10 (0.7%), b_comp_prob: 6 (0.4%), > b_tok_touch_all: 0.40 (0.0%), b_finish: 0.39 (0.0%), tests_pri_0: 1150 > (85.0%), > check_spf: 8 (0.6%), poll_dns_idle: 0.08 (0.0%), check_razor2: 826 (61.0%), > tests_pri_500: 5 (0.4%), get_report: 0.58 (0.0%) > > As you can see, "tests_pri_0: 1150 (85.0%)" is taking up most of the > time in SA (with check_razor2 coming in at second place). > > But how can I drill down further, in order to see which tests in the > SA ruleset take longest?
Is http://lists.amavis.org/pipermail/amavis-users/2011-May/000330.html really my best bet? -- Ralf Hildebrandt Charite Universitätsmedizin Berlin [email protected] Campus Benjamin Franklin http://www.charite.de Hindenburgdamm 30, 12203 Berlin Geschäftsbereich IT, Abt. Netzwerk fon: +49-30-450.570.155
