If one is running multiple SMTP servers, is there any harm running
"cleanup" from each one
at the same time via a cron job, or is it better to stagger the times a bit?
-
This SF.net email is sponsored by: Splunk Inc.
Still gr
Geert Hendrickx wrote:
> On Tue, Aug 28, 2007 at 12:04:50PM -0700, Tobias Kreidl wrote:
>
>> If one is running multiple SMTP servers, is there any harm running
>> "cleanup" from each one
>> at the same time via a cron job, or is it better to stagger the times
Are there problems with a lot of invalid recipients, or just large numbers per
mailer envelope?
In any case, in postfix, try setting these in main.cf:
smtpd_recipient_limit=512 (max. number per envelope)
smtpd_junk_command_limit=1000 (lets real addresses in envelopes get delivered
to,
while
Jordi Espasa Clofent wrote:
> Hi all,
>
> I want to establish a maximum number of messages that the accounts
> behind SMTP can recieve per hour.
> I understand that this berhaviour is controled by:
>
> RECIPIENTTHROTTLE=1
> RECIPIENTMSGLIMIT=1000
> RECIPIENTTIMELIMIT=1h
> RECIPIENT_QUOTA_REJECTION
R Wahyudi wrote:
> Hi Leon,
>
> I don't think the problem is caused by policyd, and you are not the
> only postfix users complaint about this.
>
> Try setting smtpd_timeout = 60s.
> I had this setup for ages and never giving me any problem.
> Read the whole thread :
> http://www.security-express.co
Cami Sardinha wrote:
> Tobias J. Kreidl wrote:
>
>> So, in the case of 5k connections per minute, the server is probably
>> saturated... does policyd log that it's reached its limit?
>>
>
> Connections per minute do not mean much. Concurrent connections
> is what matters. It should log when
Benny Pedersen wrote:
> On Fri, September 7, 2007 10:46, Xavier Beaudouin wrote:
>
How many concurrent smtpd processes do you have on your MX?
>>> Up to 1.5k per MX in peak during bot attacks.
>>>
>> Hum. Anvil on postfix is a very good idea to avoid too mutch ressource
Xavier Beaudouin wrote:
> Hi,
>
>
>>> How many concurrent smtpd processes do you have on your MX?
>>>
>>>
>>>
>> Up to 1.5k per MX in peak during bot attacks.
>>
>
> Hum. Anvil on postfix is a very good idea to avoid too mutch ressources
> utilization.
>
> /Xavier
>
>
But anvi
In that case, there should at least be a comment in the start-up scripts
provided. :-) As root, it's usually a limit that you can easily
override and that would be done best in the startup script. Also, the
documentation
doesn't discuss the limit/file descriptor constraints really -- you sort
So, if you remove the "LIMIT" restriction, it then just runs longer,
with no other adverse effects? That aside, I assume you want to run the
cleanup application at night because the load is generally a lot lower.
Just to get an idea what you're up against, how many messages are we
talking abo
d would suggest that this could perhaps at least be
overridden through a command line variable.
--Tobias
Nate wrote:
> At 09:59 AM 10/5/2007, Tobias Kreidl wrote:
>
>> So, if you remove the "LIMIT" restriction, it then just runs longer,
>> with no other adverse effect
Cami Sardinha wrote:
> Benny Pedersen wrote:
>
>> On Mon, October 8, 2007 08:29, Cami Sardinha wrote:
>> if you have to do it
>>
>> do
>> {
>> } while count(bar() > 10)
>>
>>
>> did you not learn comal in school ?
>>
>
> No.
>
> Cami
>
> -
Rianto Wahyudi wrote on 2007-10-07 01:25:
> | %yahoo.com%|
Shouldn't that read "%yahoo.com" (no trailing "%")?
You'd think some spammers would get wise to using hostnames like
"ns1.xyz.com" for sending spam -- you really don't see much, if any of
this sort of thing, Rianto?
Sloan wrote:
> Leonardo Rodrigues Magalhães wrote:
>
>> Tobias Kreidl escreveu:
>>
>>> Rianto Wahyudi wrote on 2007-10-07 01:25:
>>>
>>>
>>>> | %yahoo.com%|
>>>>
>>>>
>
I am getting some odd results when I try to send mail that contains one
or more invalid addresses -- I don't want the mail client to reject
sending the message altogether, just report back the errors. It seems as
if I get greylisted almost immediately, which shouldn't happen (I
thought) unless
Jon,
The MySQL DB has various quota varibled defined as INTEGER (not sure why
not UNSIGNED INTEGER), and this limits the maximum value to
2,147,438,647. I do not think MySQL suports any larger integer values.
Oracle, on the other hand, supports both Int64 and UInt64 integers.
In your case, y
Changing the DB definitions is something Cami would need to consider,
not me! :-)
The various quota parameters are defined as "unsigned int," so they'd
have to be redefined
in the source code, as well as for the table spaces. Again, not my call!
--Tobias
P.S. When I refresh the policyd mailin
One more thing, Jon... You're right that in the policyd DB that the
various quota fields are of the type "unsigned integer," and the C code
also seems to specify unsigned int (I'm looking at the 1.9X beta code --
however, I'm running 1.82 in production), so unless there's something
else, I'm no
Maybe it was a TLS/SSL connenction that couldn't be completed, if you
don't have that enabled in postfix?
You can add logging for TLS connections by setting smtpd_tls_loglevel to
be greater than 0 (the default) up to a max. of 4.
So, in the postfix main.cf file, you'd need to insert:
smtpd_tls_r
Alejandro,
The whole idea of the throttle table is to limit the amount of email of
a sender for a certain period of time. As the time expires, so do the
records. You need only establish limits that are applied to all users
(thoug you can set exceptions in the whitelist), but the throttle and
In our casem wqe don;t care about the recipients, just the senders. In
your case, if you wnat to limit what recipients can receive, then the
table size will be determined eventually by how many indovidual
recipients you have on your host(s).
If you want to really limit email based on a monthly
No, the RECIPIENT_INACTIVE_EXPIRE variable applies to recipients (on
your email hosts). Senders are regulated with
SENDER_INACTIVE_EXPIRE (which can in principle be anybody who sends mail
to your site). We use, for example, SENDERTIMELIMIT=1h and
SENDER_INACTIVE_EXPIRE=4h which is generously
regulate which messages get
sent to policyd and which bypass it.
--Tobias
John Beaver wrote:
> Tobias Kreidl wrote:
>
>> Alejandro,
>>
>> The whole idea of the throttle table is to limit the amount of email of
>> a sender for a certain period of time. As the time
You can, of course, have a cron job that re-populates the table with the
exceptions for your few users, which might be the easiest way to do this
(and it sounds like this is more or less what you are doing now).
Otherwise, I don't see an easy way to preserve the information unless
you expire t
From my understanding of RFC 1870, there is nothing wrong with a
message with a content of zero size.
One option to trap such messages would be to use the body_checks postfix
filter, but that's "expensive".
Another option that might be more efficient would be to use a procmail
filter on the ma
t line:
:0:
* ! B ?? .
zero-length-body-folder
--Tobias
Tobias Kreidl wrote:
> From my understanding of RFC 1870, there is nothing wrong with a
> message with a content of zero size.
>
> One option to trap such messages would be to use the body_checks postfix
> filter, but
Carlos,
If you want to impose email size limits, you need to use the
"throttle_recipient" table, not the "throttle table". Look at the very
bottom of the policyd.conf file. Plus, there are only restrictions on
numbers of messages, not sizes.
The "throttle" table refers to the origination of the
I've set the "cleanup" routine to run hourly via cron, but with the
expiration set such that records from
the DB for the "sendertimelimit" expire if over 4h but are retained in
the DB until over 8h:
# sender time limit:default: 24 hours
#
# after how long does a
For some reason, this got posted to the wrong topic back on Jan 31, so
here it is again:
I've set the "cleanup" routine to run hourly via cron, but with the
expiration set such that records from
the DB for the "sendertimelimit" expire if over 4h but are retained in
the DB until over 8h:
# sender
Dietmar,
SENDERQUOTALIMIT=21 is what I use, and it is defined to be in
bytes. It can't exceed
what can be stored in a signed, long integer (or around 2.1 x 10^9). I
would say your setting of around 200 MB is certainly reasonable. Since
I don't really want to be bothered with this limi
Dietmar Braun wrote:
>
> Ok, so if it's in bytes, the description in policyd.conf is wrong,
> unfortunately. Thank you for clearifying.
>
>> What's your SMTP server's setting 00 if any -- on the maximum size of
>> any email message? In postfix, it's defined in the main.cf file in the
>> variable "
IPtables (Netfilter) is a nice throttling mechanism, however, it's not
supported on all platforms.
There are other options, including modules for Web daemons. One such
module is mod_evasive, which can act as a throttle on Apache 1.3.X and
2.X servers.
--Tobias
[EMAIL PROTECTED] wrote:
> Does
As an aside, postfix has all sorts of rate-limiting and throttling
capabilities built in, as well.
So, one could certainly regulate the SMTP flow independent of what
policyd does regarding rate limiting. There is, for example, the
smtpd_client_connection_rate _limit parameter. There are numero
The mySQL Administrator (free download from mysql.org) is also a
convenient user interface. There are implementations for Windows, Mac
OS X, and several flavors of Linux. The download is currently at:
http://dev.mysql.com/downloads/gui-tools/5.0.html
We run two instances of policyd, one DB ser
Hello, Roland:
You didn't mention what your retention times were for the triplets, etc.
You might try cutting the expiration limits down to be no more than
about twice the limit value (so, say, 4 hours and 2 hours). We get over
1 M messages a day and even though my DBs grow to as much as 50 MB
One other thing: Mail gateways do a lot of nslookup operations, so make
sure you run "nscd" (name service cache daemon) and adjust it so it
caches more than the default number of hosts. So for example, the one
section in /etc/nscd.conf dealing with host names might contain
something like this:
I've noticed a lot of people try to put policyd further "down" in the
list of
smtpd_recipient_restrictions entires in an attempt to load the system
less. Well,
in most cases, the extra load needed to -- for example -- see if the
connection is whitelisted is so low that we decided some time ago
Sounds like you need to set up your email servers to reject relayed mail
(see the postfix documentation for details).
Sam Przyswa wrote:
> Hi,
>
> After installing Postfix-Policyd it work fine and seems very efficient
> but I got a lot of mailer-daemon returned mail from spamers who used
> some va
The FQDN can be easily implemented on the mail gateway, such as
postfix. The HELO requirement is optional.
Jim Wright wrote:
> On Jun 10, 2008, at 7:23 PM, Hugo Monteiro wrote:
>
>> Recently i've been receiving quite a lot of spam, from different
>> places,
>> with a common denominator. The he
39 matches
Mail list logo
