@SalvadorS,
In response to your post, let´s break down the text in your post:
The question is why plesk does not limited the send of the mails (is set to 20 per domain per hour) and allow to send hundreds...
The above is not strange and can be caused by a situation in which the domain (as hosted with Plesk) is not actually sending the mail, whereas the server (on which Plesk has been installed) does send mails, with either the server functioning as an (improper) mail relay or a script sending mails from the server.
For instance, a script (i.e. often a hack, resulting in a spam script) can use the php mail() function to send mails (directly) from the server.
Also I check the statistics on the server and in all the domains are 0 mails sent per hour which is not true ...
As stated before, it can be true that the domain does not actually send mails. Hence, it can be the case that statistics for the domain are correct and zero, even though the whole server sends many mails.
Note that the above also implies that the outgoing mail control actually works, but does not apply to mails sent via a spam script.
In essence, many countermeasures have to be initiated and/or deployed to reduce spam to the absolute minimum, even though spam cannot be completely removed.
Some of these countermeasures have already been given in my earlier post.
In addition, one should implement "good practice" in the form of:
a) block specific IP addresses (permanently) in the firewall, given the fact that some spammers are making use of a very consistent IP Block range (note that, with the fail2ban module, you can keep track of bad IPs and it is very likely that you can deduce certain IP blocks, that can consequently be blocked permanently by a firewall rule),
b) disable the php mail() function, which is possible on a domain-per-domain basis and/or for the whole server (note that some basic additional comments can be found here:
http://talk.plesk.com/threads/mail-...its-possible-to-send-spam.332604/#post-777973 ),
c) disable specific script functions on all domains (note that Plesk allows a lot of scripting languages by default and, in most cases, not all of these languages are used. Allowing all scripting languages increases the risk of having spam scripts sending huge amounts of mails),
d) make sure that Wordpress is always or, even better, automatically updated to the newest version (note that Wordpress and primarily Wordpress plugins are vulnerable to script injection, a possibility that has been used by many spammers. Latest versions of Wordpress reduce the risk of script injection and disallowing specific, unreliable plugins reduces the risk on spam even more),
e) investigate php files and certainly those of Wordpress, search for strange lines of code in the php files (these lines of strange code are often spam scripts or code),
f) always use spamfilters, to check for both incoming AND outgoing spam,
and so on.
In short, it may be a lot of work to reduce spam, but in the long run it becomes easier and less time-consuming, IF you have fail2ban and use the method from point a).
Hope the above gives you a general idea of how to "attack spam".
Kind regards....