After another look of the maillog, I see there are also at least another
attack like this, instead of AUTH, there are EHLO, RCPT, UNKNOWN,
STARTTLS and CONNECT, among these, EHLO is for sure anther attack. The
others does not have many attempts in a short period.
May be I should make it a wildcard in the filter. But I don't want block
the normal user by mistake.
Now, I set findtime=120 and maxretry=5. Also the filter now is like this:
failregex = ^%(__prefix_line)slost connection after AUTH from (.*)\[<HOST>\]
^%(__prefix_line)slost connection after EHLO from (.*)\[<HOST>\]
^%(__prefix_line)slost connection after UNKNOWN from
(.*)\[<HOST>\]
Gao
On 15-09-30 04:42 AM, Nick Howitt wrote:
> I think it can be pretty bad behaviour if you get too many attempts too
> quickly. I've seen over 1000 in 24h from a single IP. I generally see
> nothing but every now and then someone has a go and it is a bit
> irritating.
>
> I also had one IP doing about 6 attempts an hour for days until I
> spotted it and blocked it. I was blocking 10 occurrences in an hour but
> dropped it to 5 an hour to combat the IP.
>
> On 2015-09-30 12:08, Darac Marjal wrote:
>> On Tue, Sep 29, 2015 at 02:30:47PM -0700, Gao wrote:
>>> Hello, all
>>>
>>> I have the postfix-sasl jail enabled and it works well against
attack,
>>> such as
>>> "Failed login".
>>>
>>> I just notified that my email server's maillog flood with this:
>>> ...
>>> Sep 29 14:19:21 szeta postfix/smtpd[19940]: connect from
>>> ns3366447.ip-37-187-77.eu[37.187.77.147]
>>> Sep 29 14:19:22 szeta postfix/smtpd[19940]: lost connection after
AUTH
>>> from
>>> ns3366447.ip-37-187-77.eu[37.187.77.147]
>>> Sep 29 14:19:22 szeta postfix/smtpd[19940]: disconnect from
>>> ns3366447.ip-37-187-77.eu[37.187.77.147]
>> "Lost connection after AUTH" means that postfix sent "AUTH" to the
>> client, and the client disconnected. In other words, the client
>> probably
>> attempted some action which you've configured that only authorized
>> users
>> can perform (usually, this is something like sending mail to a
>> different
>> serveer (relaying)). Postfix said "authorize yourself in order to
>> perform this action", and the client just dropped the connection
>> (rather
>> than cleanly quitting and waiting for postfix to close the
conecction).
>>
>> In other words, no authorization was attempted.
>>
>> I suspect that fail2ban doesn't block this normally because it's not
>> really bad behaviour. It's akin to someone connecting to your SSH port
>> and disconnecting upon finding that it's asking for a password :)
>>
>>> ...
>>>
>>> And the fail2ban does nothing about this! No new entry about this in
>>> fail2ban.log. The attack is still going and I am going to manual kill
>>> it in
>>> iptables.
>>>
>>> What should I do about this in fail2ban? Please help.
>>>
>>> Thanks.
>>>
>>> Gao
>>>
>>>
------------------------------------------------------------------------------
>>> _______________________________________________
>>> Fail2ban-users mailing list
>>> [email protected]
<mailto:[email protected]>
>>> https://lists.sourceforge.net/lists/listinfo/fail2ban-users
>>
>>
------------------------------------------------------------------------------
>>
>> _______________________________________________
>> Fail2ban-users mailing list
>> [email protected]
<mailto:[email protected]>
>> https://lists.sourceforge.net/lists/listinfo/fail2ban-users
>
------------------------------------------------------------------------------
> _______________________________________________
> Fail2ban-users mailing list
> [email protected]
<mailto:[email protected]>
> https://lists.sourceforge.net/lists/listinfo/fail2ban-users
------------------------------------------------------------------------------
_______________________________________________
Fail2ban-users mailing list
[email protected]
<mailto:[email protected]>
https://lists.sourceforge.net/lists/listinfo/fail2ban-users