1 |
On Dec 4, 2011 10:10 AM, "Michael Orlitzky" <michael@××××××××.com> wrote: |
2 |
> |
3 |
> On 12/03/2011 09:48 PM, Pandu Poluan wrote: |
4 |
>> |
5 |
>> |
6 |
>> |
7 |
>> Thanks! Very helpful resources. |
8 |
>> |
9 |
>> You mentioned amavisd-new. What's their relationship? I mean, if I |
10 |
>> deploy postscreen, how will it affect amavisd-new? |
11 |
>> |
12 |
> |
13 |
> Postscreen sits in front of smtpd, and handles all incoming connections. |
14 |
It hands the "good" connections off to the real smtpd daemon. Amavisd-new |
15 |
(in both before/after-queue configurations) interacts with the real smtpd, |
16 |
so postscreen doesn't directly affect it at all. |
17 |
> |
18 |
> What was I talking about? |
19 |
> |
20 |
> With amavisd-new, a before-queue filter is generally nicer, because you |
21 |
can reject spam, notifying the sender, rather than discarding it or |
22 |
backscattering. But, amavisd-new is a hog, and with a before-queue filter, |
23 |
an amavis process gets used every time ANY connection is made. Since 95% of |
24 |
your connections will be crap (that is a technical term), you waste tons of |
25 |
resources creating/killing amavisd-new processes for botnets and other scum |
26 |
that will be rejected quickly. |
27 |
> |
28 |
> On a busy server, it will kill you. |
29 |
> |
30 |
> Postscreen only passes the "good" connections to a real smtpd, so with |
31 |
postscreen running, new amavis processes only get used for those good |
32 |
connections. If postscreen can get reject 90% of the incoming connections, |
33 |
you'll use an order of magnitude less resources doing before-queue |
34 |
filtering than you would without postscreen. |
35 |
> |
36 |
> So, in essence, postscreen is what allows you to run the before-queue |
37 |
filter with comparable resources to the after-queue filter. |
38 |
> |
39 |
|
40 |
Thanks for all the information. You really should write a wiki.g.o article |
41 |
about the new setup :-) |
42 |
|
43 |
Rgds, |