Re: [Exim] ETRN again...

Top Pagina
Delete this message
Reply to this message
Auteur: Odhiambo Washington
Datum:  
Aan: exim-users
Onderwerp: Re: [Exim] ETRN again...
* Tamas TEVESZ <ice@???> [20010903 14:05]: writing on the subject 'Re: [Exim] ETRN again...'
| On Mon, 3 Sep 2001, Marcus Fleige wrote:
|
| > is it possible to forward the emails from that pop-mailbox to the
| > customers mailserver via etrn?
|
| no. not from the 'pop-box'. when it's in the 'pop-box' (ie had been
| delivered by exim) it's essesntially out of exim's control.
|
| you can, however "hold_domains = foo.com". this makes the messagaes to
| that given domain not to be delivered, but just held on the queue. now
| when the client comes in and says "etrn foo.com" then they're flushed.


I have some situation which I'd also like to bring to light.
I have some domains whose mail I keep in the queue. I've given them static
IP addresses and configured my DNS such that they are their own primary
MX, like the one below:

<snip>
mail            IN      A       62.8.66.25
@               IN      MX      5       mail.domainX.co.ke.
@               IN      MX      10      mail.wananchi.com.
@               IN      NS      ns1.wananchi.com.
@               IN      NS      ns2.wananchi.com.
@               IN      A       62.8.64.4
gw        IN    CNAME    mail
www             IN      CNAME   mail
</snip>



Now when Exim receives mail addressed to user@???, it tries to
deliver it directly but if that remote server is not online, Exim says
this:

2001-09-03 00:29:02 15demQ-0004Cn-00 == user@??? T=remote_smtp
defer (60): Operation timed out

I don't really like this.

When I use hold_domains = lsearch;/some/file, Exim says

2001-09-03 15:30:34 15dss4-0003rp-00 == user@??? defer
(-24): domain is held

Why shouldn't Exim just say "250 OK" or something?
Should I change the priorities for MX???


One last thing: I get so many of these in my mainlog:

2001-09-03 15:48:27 15dro1-0001EF-00 SMTP timeout while connected to
mx1.mail.yahoo.com [64.157.4.87] after sending data block (221103 bytes
written): Operation timed out


Any clues as to what I should suspect here??


PS: Sorry for jipoxing and seizing the opportunity as if it belonged to
me.

| ---
|
| actually there's something in etrn that bothers me. being able to
| connect to an smtp server and having access to the ETRN command, i can
| start a queue run to any domain i wish to. isn't that bad ? i mean, if
| there are lots of mails, and the mxes for that domain are not
| available, this will eat mtas resources; and this also raises a
| question (be this exim-specific; i have a domain "on hold". someone
| issues an etrn to that domain; this invokes exim -Rf (anything else
| to try to achieve this functionality?); and this forces
| the delivery.



This really sounds bad, since anyone can telnet to the smtp port and do
it.


| this is ok, if i'm the one who is supposed to ask for my mails. but if
| i'm just playing abusive, how will this affect the further handling of
| these messages ? will those that are already in the queue be held
| until the next ETRN (the ones that will arrive later will be held,
| that's clear), or does a one-time -Rf cease the effect of
| hold_domains ? (the latter can give a lot of opportunity to be
| abusive)


Hmm, I'll wait to hear what the gurus say about this.



-Wash

--
Odhiambo Washington
Wananchi Online Ltd.,
wash@??? 1st Flr Loita Hse.
Tel: 254 2 313985 Loita Street.,
Fax: 254 2 313922 PO Box 10286,00100-NAIROBI,KE.

When you were born, you cried and the world rejoiced. Live your life in such a
manner that when you die the world cries and you rejoice.
-old Indian saying