 | Start a set with this search |
 | Include this search in one of my sets |
 | Exclude this search from one of my sets |
 | Permalink to these results Paste this link in email or IM: |
 | Atom feed for tracking future search results Paste this URL into your reader: |
7 messages in net.sourceforge.lists.courier-maildrop[maildropl] Re: maildrop-1.5.3 and EX...| From | Sent On | Attachments |
|---|---|---|
| Deim Agoston | May 2, 2003 3:51 pm | |
| Sam Varshavchik | May 2, 2003 4:17 pm | |
| Deim Agoston | May 2, 2003 5:03 pm | |
| Deim Agoston | May 3, 2003 5:41 am | |
| Aleksey Perov | May 5, 2003 3:00 am | |
| Ken...@muspellsheim.net | May 5, 2003 1:51 pm | |
| Marcio Merlone | May 6, 2003 7:55 am |
| Permalink for this message Paste this link in email or IM: |
| Permalink for this thread Paste this link in email or IM: |
| Atom feed for this thread Paste this URL into your reader: |
| Subject: | [maildropl] Re: maildrop-1.5.3 and EX_TEMPFAIL error | Actions... |
|---|---|---|
| From: | Sam Varshavchik (mrs...@courier-mta.com) | |
| Date: | May 2, 2003 4:17:33 pm | |
| List: | net.sourceforge.lists.courier-maildrop | |
Deim Agoston writes:
Hello!
I've just upgraded to version 1.5.3 from 1.3.7 (it worked stable so there was no reason to upgrade). Right now, I just get the EX_TEMPFAIL message instead of EX_NOUSER. It would be OK if I didn't touch the maildrop/main.C file and replace the static value of nouser from EX_TEMPFAIL to EX_NOUSER - for the sake of newcommers: that's the proper value for a non-existent user. So, I touched the files and replaced the EX_TEMPFAIL with EX_NOUSER. Built the binaries, installed them.... and gone mad. Why the f*@&{ maildrop can implement the correct behaviour
It is the correct behavior. If the recipient mailbox does not exist, your mail server should reject the recipient address as nonexistent and nondeliverable, and maildrop will never even see it.
As such, if the LDA is unable to determine the location of the recipient's mailbox, it is interpreted as a local misconfiguration, and is deferred accordingly.
when it would be the best LDA available. Right now, it violates sysexists.h and I'm not sure but maybe an RFC which describes how to behave in this situation. Sorry, Sam but I can only speak right from my heart. You gave the community this great piece of software. But why, tell me, why you can't beleive if you just listen to others your respect won't going down and you can make mistakes? Nobody's perfect, you know.
I believe that I am correct.
Nor you, nor me. So many people told you that this ISN'T the correct behavior, more, I say, with this behaviour maildrop is a security risk.
Oh really? I'll tell you what a security risk really is: for poorly designed or configured mail server to accept mail to nonexistent recipients, only to bounce them later. As such, your server can now be used as a mailbomb proxy.
You know why? Because somebody can flood my filesystem. And you know why I don't make my user's list available to the open world? Because so many "great" harvesting tools are out there in the wild.
My user list is also available. Right now, I bounce more mail addressed to nonexistent recipients -- no doubt as a result of buggy harvestware -- then I accept real mail. I shudder to imagine what's happening to all Qmail relays out there which stupidly accept every turd that's flung at them.
Beside that, it makes bigger load on the machine (because of double checking of users).
And you think that's less of the load than to accept the entire load of garbage, only to be forced to bounce it later? Is that what you're arguing, that it's a "bigger load" to reject nonexistent recipients right up front, rather than happily swallow every mail to nonexistent recipients, say "thank you", and then figure out where to spit back the crap.
Here's a security risk for you: mailbomb any Qmail server with a return address set to a carefully chosen spam domain that's been shutdown, but still has valid MX record, pointing to a dead IP address. Just keep sending mail for a few hours, which should be enough to fill up the disk with bounces that can't go anywhere.