I went ahead and tried modifying chkuser. these two little quick and dirty
changes seem to make it work. I'm still open to hearing alternatives (if there
are any).
--- ../chkuser-2.0.8b-release/chkuser.c 2004-12-08 17:10:36.000000000 -0800
+++ chkuser.c 2005-06-30 17:10:44.192650158 -0700
@@ -435,15 +435,10 @@
#endif
#if defined CHKUSER_RCPT_LIMIT_VARIABLE
- maxrcpt_string = env_get (CHKUSER_RCPT_LIMIT_VARIABLE);
- if (maxrcpt_string) {
- maxrcpt_limit = atoi (maxrcpt_string);
+ maxrcpt_limit = CHKUSER_RCPT_LIMIT_VARIABLE;
if (maxrcpt_limit < 1) {
maxrcpt_limit = 0;
}
- } else {
- maxrcpt_string = "";;
- }
#endif
#if defined CHKUSER_WRONGRCPT_LIMIT_VARIABLE
--- ../chkuser-2.0.8b-release/chkuser_settings.h 2004-12-08
17:10:36.000000000 -0800
+++ chkuser_settings.h 2005-06-30 17:01:28.512240856 -0700
@@ -160,7 +160,7 @@
* The first reached, between CHKUSER_RCPT_LIMIT_VARIABLE and
CHKUSER_WRONGRCPT_LIMIT_VARIABLE,
* makes chkuser rejecting everything else
*/
-#define CHKUSER_RCPT_LIMIT_VARIABLE "CHKUSER_RCPTLIMIT"
+#define CHKUSER_RCPT_LIMIT_VARIABLE 25
/*
* Uncomment to define a variable which contains the max unknown recipients
number
--
Igor
On Thu, Jun 30, 2005 at 04:21:24PM -0700, Igor Grinchenko wrote:
-> Hello everyone,
->
-> I have a pop-before-smtp server running vpopmail 5.4.5. That server only
serves clients that supply a valid login/password pair.
-> Some clients, however, started abusing this system by sending tons of spam
which gets us blacklisted in various RBL DBs.
-> what I'm trying to do is use CHKUSER_RCPTLIMIT option to at least limit the
number of emails that gets sent at once.
->
-> it seems like for that I need to have the following line in tcp.smtp.cdb for
every entry: RELAYCLIENT="",RBLSMTPD="",CHKUSER_RCPTLIMIT"<some number>"
-> since vpopmail only creates RELAYCLIENT="",RBLSMTPD="" - it seems like the
only way is to hack vpopmail.c around line 2526.
->
-> another way is to hack chkuser to hardcode some value to
CHKUSER_RCPT_LIMIT_VARIABLE instead of looking at the env variable from the cdb
file.
->
-> which way should I go? I don't think I'll have a problem modifying either
source code, since it seems like a pretty straightforward change.
->
-> am I missing some other, simplier way?
->
-> --
-> Igor
->
->
