How to deal with these spams?

2006-03-15 Thread Jeff Peng
Hello,lists,

Some spammers use SMTP bounce to send lots of spams to our systems.For 
example,they send spams to some mail servers which are lost for antispam 
mechanism (They may send to the uncorrect users on those mail 
servers).Surely,these spams are faked with header,whose return-path are facked 
as our system's real user accounts.Then,those mail servers bounce the spams to 
our users' mailboxes.While we can't reject those mail servers,although they are 
abused by spammers really.
How can I deal with these spams?Thanks.

Jeff


_
Call Anyone, Anytime, Anywhere in the World - FREE!
Free Internet calling from NetZero Voice
Visit http://www.netzerovoice.com today!




Re: SpamAssassin large-scale users willing to comment?

2006-02-23 Thread Jeff Peng

Hi,Jdow,

I think you have misunderstood me.I'm not working for Netzero.In fact,our mail 
systems' end-users numbers are more than a hundred million far away.We use SA 
as part of our antispam mechanism. Certainly,it's not the original one written 
with Perl.We disable the Bayes currently,and only use rules for 
filters.While,we'll continuely to develop to support Bayes and more others 
featers of SA. We use network test too,so we should reject most of the spams in 
front of SA.
btw: When I say,we have more than 30 MTAs,it don't really mean we ONLY have 
that numbers.Some datas including users numbers and messages per day,are a 
little sensitivity to the commercial company,I hope you understand me,thanks.

Jeff



_
Call Anyone, Anytime, Anywhere in the World - FREE!
Free Internet calling from NetZero Voice
Visit http://www.netzerovoice.com today!




Re: SpamAssassin large-scale users willing to comment?

2006-02-23 Thread Jeff Peng
Hi,jdow,

I think you have misunderstood me.I'm not working for Netzero.In fact,our mail 
systems' end-users numbers are more than a hundred million far away.We use SA 
as part of our antispam mechanism.Certainly,it's not the original one written 
with Perl.We disable the Bayes currently,and only use rules for 
filters.While,we'll continuely to develop to support Bayes and more others 
featers of SA.
We use network test too,so could reject most of the spams in front of SA.

-- jdow [EMAIL PROTECTED] wrote:
After I replied offline I did some looking around

Er, I just looked up on a size ranking from November 20th last year. NetZero
was set at 8,600,000 members. AOL is only 22,200,000 members. Do you mean
100,000,000 EMAILS PER DAY or something? That makes more sense and is maybe
somewhat low.

860 members * 1000 emails per member/day / 86400 seconds/day gives
10,000 emails a second. That's still a whomping load for only 30 MTAs,
300 or so emails a second when I get about 3 seconds per email throughput
at my low non-overlapping volume.

Of course you probably greylist first. And with a network average of 80%
spam that puts you down into the 60 emails a second range. It's still
an amazing push.

Is your version of SA still in perl or is it in machine language? I also
presume you optimized it for your needs. Does it still do Bayes and rules
or only one of them?

{^_^}
mumble The whole US internet seems to come out around 100,000,000
at the time of that survey. Of course, Hotmail is listed at 140,000
members. But there are some accounts that receive no mail so that
cuts them way down from what an ISP might run. (I have one I
abandoned as useless to me. When accounts cost money they get used
more.) {^_-}
- Original Message - 
From: Jeff Peng [EMAIL PROTECTED]


Yes,we have more than 30 MTAs running the SA.While it's not the original SA,it 
has been 
improved by us for more better performance.


-- jdow [EMAIL PROTECTED] wrote:
From: Jeff Peng [EMAIL PROTECTED]

 I really mean a hundred million.

Jeff, if you are ever permitted to and are moved to describe that
installation sometime I'd be an avid reader. You must have an amazing
array of equipment involved in the filtering and email handling.

{^_^}



_
Call Anyone, Anytime, Anywhere in the World - FREE!
Free Internet calling from NetZero Voice
Visit http://www.netzerovoice.com today!



_
Call Anyone, Anytime, Anywhere in the World - FREE!
Free Internet calling from NetZero Voice
Visit http://www.netzerovoice.com today!




Re: unable to connect to database: DBI module not available

2006-02-23 Thread Jeff Peng
Hi,
That seems your Perl installation lack the DBI module,which is needed for 
accessing mysql in Perl.You should install the DBI and DBD::Mysql modules by 
hand.Certainly,you could go to http://search.cpan.org and get them.

HTH

-- Joshua Tinnin [EMAIL PROTECTED] wrote:
Using FreeBSD 6.0-RELEASE-p4, mysql 4.1.18, p5-Mail-SpamAssassin-3.1.0_6

Using the debug flag, I get:

[25618] dbg: bayes: unable to connect to database: DBI module not 
available: No such file or directory
[25618] dbg: bayes: using username: krinklyfig

Now, I don't know why it would be using that username, as I specifically 
entered in to /usr/local/etc/mail/spamassassin/local.cf (sa_user* is 
just the generic example):

clear_report_template
report_safe 0   
lock_method flock
dns_available yes

bayes_store_module  Mail::SpamAssassin::BayesStore::MySQL
bayes_sql_dsn   DBI:mysql:sa_bayes:localhost
bayes_sql_username  sa_user
bayes_sql_password  sa_user_password

auto_whitelist_factory  Mail::SpamAssassin::SQLBasedAddrList
user_awl_dsnDBI:mysql:sa_bayes:localhost
user_awl_sql_username   sa_user
user_awl_sql_password   sa_user_password


Mysql is running, and the error log doesn't show any errors. I can log 
into mysql just fine manually with the username and password I 
assigned. This worked up until very recently. What could be causing 
this problem?

- jt



_
Call Anyone, Anytime, Anywhere in the World - FREE!
Free Internet calling from NetZero Voice
Visit http://www.netzerovoice.com today!




Re: SpamAssassin large-scale users willing to comment?

2006-02-22 Thread Jeff Peng
I'm interested in this reporter.We use spamassassin's partial features,it's 
original now,while we should improve it.Thre are more than a hundred million 
users are protected under SA here.

-- [EMAIL PROTECTED] (Justin Mason) wrote:
Hey all --

Apache SpamAssassin has won DataMation Product of the Year in the
anti-spam category *again* this year -- for the second year running!
(yay!)

One thing that would be really cool would be some comments from our
customers, for the press surrounding this.

If you, or someone you know, would be willing to talk to a reporter about
how SpamAssassin has helped eliminate spam in your organization, that'd be
great.  (A non-technical organisation would be even better btw.)

Anyone interested?  Please reply here, or if you'd prefer to follow up
confidentially for whatever reason, to [EMAIL PROTECTED].

--j.



_
Call Anyone, Anytime, Anywhere in the World - FREE!
Free Internet calling from NetZero Voice
Visit http://www.netzerovoice.com today!




Re: SpamAssassin large-scale users willing to comment?

2006-02-22 Thread Jeff Peng


A hundred million or a hundred thousand?


I really mean a hundred million.


_
Call Anyone, Anytime, Anywhere in the World - FREE!
Free Internet calling from NetZero Voice
Visit http://www.netzerovoice.com today!




Re: Re: 3.1 seems worse than 2.64?

2006-01-23 Thread Jeff Peng
Are there any optimizing options for SA (I mean the performance)? if we want to 
run SA on our antispam system.
There are more than ten millions of messages coming into our system everyday.





On Monday, January 23, 2006, 8:13:26 AM, Dan Bongert wrote:
 I recently did an email server change/upgrade from Sendmail on FreeBSD (w/
 Spamassassin 2.6.4) to Postfix on RHEL 3 (w/Spamassassin 3.1). On both
 systems, Spamassassin is called from user's .procmailrc files--not every
 user wants to be running SA (I'm not quite sure why).

 I wasn't able to convert people's Bayes databases from one system to the
 other--the Linux system didn't recognize them at all as valid DB files, so
 everyone had to start Bayes over from scratch.

 Here's my problem: the new SA doesn't work nearly as well as the old one.
 Some of my users are reporting 50% false negatives in their inbox in the
 morning, even after their Bayes autolearning has kicked in. We run a nightly
 learning script for them, and have been telling everyone to put any and all
 false negatives in the appropriate mailbox so that sa-learn can snag them.

 For my own experiences, I'm seeing a lot more spam that's being autolearned
 as ham--scores of 0.0 and even negative ones for things that to my eyes are
 very obviously spam.

 It's a pretty vanilla set up so far--are there any recommended optional
 rules sets or tweaks I haven't discovered for 3.1 yet? Unfortunately, I
 don't have any hard numbers comparing the set ups, just lots of complaints
 that the new version isn't as good.

You may want to check for a broken trust path.  (See wiki.)  Also
be sure to enable network tests and apply for rsync access for
RBL and SURBL zone files if you handle a lot of messages (100k
messages/day). 

Cheers,

Jeff C.
-- 
Jeff Chan
mailto:[EMAIL PROTECTED]
http://www.surbl.org/

.





Re: RE: rbldnsd on FreeBSD

2006-01-22 Thread Jeff Peng
when you run ./rbldnsd -h
you should see:
-b address[/port] - bind to (listen on) this address (required)

So you can bind the rbldnsd to another alias IP address,diff from the IP that 
your BIND server is listening to.
I think there is no conflict between the rbldnsd and the BIND.

Jeff Peng wrote:
 hi,Irina,
 rbldnsd is really a simple dns server.you can use it directly,no any
 need to bind.and,you can use rsync to download the rbl files. 
 
I have both rbldnsd and bind running on my 2 nameservers.  I had to
bind(pardon the pun) rbldnsd
To a separate alias IP, as I couldn't seem to make bind9 do the forward
correctly.

Rbldnsd is in FreeBSD ports (although it seems to be a release or 2 down,
I'll probably submit
An update soon).

LER


-- 
Larry Rosenman http://www.lerctr.org/~ler
Phone: +1 512-248-2683 E-Mail: ler@lerctr.org
US Mail: 430 Valona Loop, Round Rock, TX 78681-3683 US

.





Re: rbldnsd on FreeBSD

2006-01-21 Thread Jeff Peng
hi,Irina,
rbldnsd is really a simple dns server.you can use it directly,no any need to
bind.and,you can use rsync to download the rbl files.



 --- Ursprüngliche Nachricht ---
 Von: Irina [EMAIL PROTECTED]
 An: users@spamassassin.apache.org
 Betreff: rbldnsd on FreeBSD
 Datum: Sat, 21 Jan 2006 14:57:02 -0500
 
 Hello all,
 
 Thank you for your answers on SURBL (few days back).  I decided to install
 rbldnsd with rsync and have few things to ask.
 
 It will run on FreeBSD 5.4 with no named running.  Server uses
 resolve.conf
 with 2 our DNS servers.
 
 Do I need to use BIND with rbldnsd and rsync?  Or only rbldnsd and rsync?
 
 If I don't really need it with BIND, but would it be beneficial?
 
 Thank you,
 Irina
 
 
 

-- 
DSL-Aktion wegen großer Nachfrage bis 28.2.2006 verlängert:
GMX DSL-Flatrate 1 Jahr kostenlos* http://www.gmx.net/de/go/dsl


Re: How SpamAssassin recognize chinese character?

2006-01-09 Thread Jeff Peng
If you are using the perl of 5.8.0 or higher,it process the unicode 
characters well.So you should not worry about that Perl how to interpret 
the Chinese character.Just use the rules as normally as english language.




From: Vincent Li [EMAIL PROTECTED]
To: users@spamassassin.apache.org
Subject: Re: How SpamAssassin recognize chinese character?
Date: Mon, 9 Jan 2006 14:26:09 -0800


On 9 Jan 2006, at 10:08 PM, Jon Armitage wrote:


Vincent Li wrote:
I have been using SpamAssassin for quite a while, and used SARE  
rules and other custom rules. I am interested in writing my own  
chinese spam rules to block chinese spam email.
I cheat and use an Exim acl statement to reject messages composed  
in unwanted character sets. However, I don't know which other MTAs  
would be able to do this, or even if this blanket approach would  
suit you.


Jon


Hi Joh:

I am in academic enviroment, we do receive some legitemate chinese  
email and the chinese rules I downloaded works well. I am just  
curious how SpamAssassin or Perl interpret the rules written in 
Chinese?


Vincent


_
与世界各地的朋友进行交流,免费下载 MSN Messenger:  
http://messenger.msn.com/cn 



RE: URIBLFP? [Was: SA or Commercial AntiSpam products]

2006-01-07 Thread Jeff Peng
I agree with Track.In fact all popular email service providers including 
hotmail,yahoo or our 163.com and 126.com in China,are heavily abused by 
spammers.So we'll try our best and spend much time and money to get rid of 
spams.And,because of some numbers of spams,someone think that Netease's IP 
and domain should be listed into black lists,it's really unfair!



--
Jeff Peng[EMAIL PROTECTED]
http://mail.163.com  http://mail.126.com  http://wazzy.nease.net 
values of β will give rise to dom!







From: List Mail User [EMAIL PROTECTED]
To: [EMAIL PROTECTED], [EMAIL PROTECTED],

users@spamassassin.apache.org

CC: [EMAIL PROTECTED]
Subject: RE: URIBLFP?  [Was: SA or Commercial AntiSpam products]
Date: Sat, 7 Jan 2006 00:29:38 -0800 (PST)

 -Original Message-
 From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
 Sent: Friday, January 06, 2006 1:31 PM
 To: mouss
 Cc: Jeff Peng; users@spamassassin.apache.org
 Subject: Re: URIBLFP? [Was: SA or Commercial AntiSpam products]

 mouss wrote:
  (top posting because not a reply:)
 
  this message triggers:
  URIBL_BLACK (nease DOT net. found in the message footer)
  MIME_CHARSET_FARAWAY RCVD_IN_NJABL_PROXY (because of
 X-Originating-IP:
  [218.19.159.186])
 
  and also
  DNS_FROM_RFC_ABUSE
  DNS_FROM_RFC_POST
  MSGID_FROM_MTA_HEADER
 
  is the uribl listing justified or is it too aggressive?
 


Is the listing justified?  Hrmm..

An SA Dev (which will remain nameless) requested this listing at
2005-07-19 21:10:16.  It was accepted at 2005-07-20 01:03:49.Its
been listed since then...  No requests for change or delist.

We'll be happy to move it to grey if need be... But nobody has spoke up
since its listing.

Thanks,
Dallas

I'm sure the original listing was requested in good faith.  And
many of the Netease domains are seriously abused (mostly by Americans).
In particular 163.com-M and 126.com-M are often favorites for spam 

domains'

contacts' email accounts and for drop-boxes.  But in general listing any
Netease domain is like listing HotPop domains or even (stretching a bit
here) mail.com domains.  They are so very large, and have millions of
non-spamming customers and on just that basis should not be in black.
Whether or not nease.net-M should be grey is an open issue, but I can
easily see how the extreme abuse would make 126.com-M and 163.com-M have
grey listings, but neither of those domains are listed, so likely all
of the Netease domains should have the same status.

I'm not even sure if normal users can get accounts at nease.net-M;
It is used for name service and email contacts for other Netease domains
(e.g. lookup Whois records and DNS for 126.com-M and 163.com-M).

Just my opinion, but at worse I think any Netease domain should
only be grey, and not a black listing;  Just like mail.com or HotPop, they
have a share of abusers, but it isn't large compared to their size (though
they are almost as bad as yahoo.com-M or hotmail.com-M for spam domains'
contacts' email accounts frequency;  Before switching to using Hotmail,
163.com-M seemed to have been Ralsky's favorite).


Paul Shupak
[EMAIL PROTECTED]


_
与联机的朋友进行交流,请使用 MSN Messenger: http://messenger.msn.com/cn 



Re: URIBLFP? [Was: SA or Commercial AntiSpam products]

2006-01-07 Thread Jeff Peng

I believe it is good practice to periodically lookup one's IPs and
domains in dnsbl/rhsbl/uribl/surbl/... This will help reporting false
positives, and may also help detecting real problems (you could get
infos on who is abusing your system/network/site for instance).



Yes,in fact we continue to keep touch with some RBL organizations to 
prevent the situation of our domains or IPs listed into their 
blacklists.And,we also keep contact with the primary email service 
providers such as yahoo,gmail to stop the tragedy of joining each other's 
IP into blacklists.








From: mouss [EMAIL PROTECTED]
To: Jeff Peng [EMAIL PROTECTED]
CC: [EMAIL PROTECTED],  [EMAIL PROTECTED],  [EMAIL PROTECTED], 

users@spamassassin.apache.org

Subject: Re: URIBLFP? [Was: SA or Commercial AntiSpam products]
Date: Sat, 07 Jan 2006 16:24:58 +0100

Jeff Peng a écrit :
 I agree with Track.

you mean Paul:)

In fact all popular email service providers including
 hotmail,yahoo or our 163.com and 126.com in China,are heavily abused by
 spammers.So we'll try our best and spend much time and money to get rid
 of spams.And,because of some numbers of spams,someone think that
 Netease's IP and domain should be listed into black lists,it's really
 unfair!

I don't think anyone wanted to be unfair. Since the site is in asian
lang, it's hard to guess whether it is spammy or not and the reporter
didn't check what type of network is netease. Such things do happen, and
will certainly happen again. And this is why I asked (note that it was
just out of luck that I've seen this: I was playing with Matt and Paul
whitelisting suggestions, and I got them wrong, so your message got to
the Junk folder).

I believe it is good practice to periodically lookup one's IPs and
domains in dnsbl/rhsbl/uribl/surbl/... This will help reporting false
positives, and may also help detecting real problems (you could get
infos on who is abusing your system/network/site for instance).

Note that netease is no more listed on uribl.



_
享用世界上最大的电子邮件系统― MSN Hotmail。 http://www.hotmail.com 



SA or Commercial AntiSpam products

2006-01-06 Thread Jeff Peng

Hi,Lists,

I'm new to SpamAssassin.How about SA's TOP capability on antispam? Can I 
use it instead of some commercian antispam products?Thanks.



--
Jeff Peng[EMAIL PROTECTED]
http://mail.163.com  http://mail.126.com  http://wazzy.nease.net 
values of β will give rise to dom!


_
与联机的朋友进行交流,请使用 MSN Messenger: http://messenger.msn.com/cn