"no_action" vs "no action"

[Alexander Moisseev]

I found lines with action "no action" in rspamd logs.
What the difference between "no_action" and "no action"?

2014-04-28 02:35:02 #91436(normal) accept_socket: accepted connection from 127.0.0.1 port 33891
2014-04-28 02:35:02 #91436(normal) composites_foreach_callback: <mailman.174.13986...@dovecot.org>, insert symbol FORGED_RECIPIENTS_MAILLIST instead of symbols: -MAILLIST, FORGED_RECIPIENTS
2014-04-28 02:35:02 #91436(normal) write_check_reply: id: <mailman.174.13986...@dovecot.org>, qid: <2CC0E3999B>, (default: F (no_action): [-3.69/15.00] [FORGED_SENDER,MAILLIST,FORGED_RECIPIENTS_MAILLIST,BAYES_HAM]), len: 18759, time: 309ms real, 25ms virtual, dns req: 57, part: C7139ECF

2014-04-28 04:08:39 #91436(normal) accept_socket: accepted connection from 127.0.0.1 port 32206
2014-04-28 04:08:40 #91436(normal) print_metric_data_rspamc: action: no_action
2014-04-28 04:08:40 #91436(normal) write_check_reply: id: <2901139...@mxfront7h.mail.yandex.net>, qid: <6834B3999B>, (default: F (no action): [0.00/15.00] []), len: 1067, time: 1469ms real, 3ms virtual, dns req: 14, part: C5B40601

Also rsapmc stat shows "no action" in total messages count, but not in any particular actions count:

# rspamc stat
Results for host: localhost:11334: 200, OK
Messages scanned: 822
Messages with action reject: 34, 4.13%
Messages with action soft_reject: 0, 0.00%
Messages with action rewrite_subject: 0, 0.00%
Messages with action add_header: 176, 21.41%
Messages with action greylist: 67, 8.15%
Messages with action no_action: 533, 64.84%
Messages treated as spam: 210, 9223372036854775808.00%
Messages treated as ham: 600, 9223372036854775808.00%
Messages learned: 144
Connections count: 822

34+176+67+533=810
I parsed rspamd logs since last start-up and found 12 "no action" lines.
810+12=822

Alexander

[Vsevolod Stakhov]

It seems to be a bug actually. I'll check the code and provide an update.

--
Vsevolod Stakhov

[Vsevolod Stakhov]

In the upcoming rspamd 0.7 this issue seems to be fixed. I'll check 0.6
branch shortly.

--
Vsevolod Stakhov