logs archiveIRC Archive / Freenode / #exim / 2010 / February / 3 / 1
verblendet
hi, is there a way to use condition = $acl_m_something in a router?
for now i do it with writing something in the header, then check header in the router and maybe remove the header again
condition = $acl_m_isham e.g. gives only errors. i found that if condition a bit complicated
also its not a good idea to always parse the header, that need a lot of performance for things that are already done for the mail, marked as spam e.g., then also have to setup a statical pattern for signing spam as spam.... e.g. subject. i use exiscan, and sa settings and header rewrite has no effect here
i do all spam reporting in exim
[1]plex
Hi. Is there a way to establish a begin retry rule based on origin IP address?
Everything I have read details how to apply it to domains which are receiving email, not those who are sending. We host many clients who use our mail server for outgoing mail through our database application. I would like specific clients informed immediately if there was a problem contacting the recipient mail server.
DukeDrake
hi
how can I have pop3-accounts with exim?
I already have a server up and running, but it only delivers to local users currently... how can I set up a pop3 box?
[1]plex
DukeDrake, you need an MTA and a pop3 server
something like exim+dovecot will work
unlofl
I'm running multiple exim virutal domains for several users domains. Is it possible to use a seperate ip for each? I think I can use acls so that exim will only accept mail for each host on the correct ip for it, but I'm trying to figure out how to make exim initiate outbound smtp connections from the appropriate ip depending on the domain of the from header.
This is the only way I can think of to guarantee that one users email list being reported as spam won't cause other users/my main domains email form being blacklisted.
petemc
you can have a router for each domain with a corresponding transport, the transport can use the interface = option
Groenleer
Hi All, I was wondering how i can see within the mainlog if a message is rejected because of spam...
petemc
did you look?
Groenleer
I am currently looking at my mainlog :)
         

petemc
grep for reject
Groenleer
i am tailing it now, so new message fly by at the moment
petemc
right, thats why its better to use grep
Groenleer
most of them just have reject RCPT
it is not giving me a reason about spam/virus/blacklisted etc.
petemc
then you need to configure it do so
Groenleer
ok.
unlofl
generally, remote servers won't say that "I thought it was spam" will they? Doesn't smtp time spam scanning use temp/perm rejects when it is "strongly suspected"?
Groenleer
unlofl, i am looking for messages which are blocked by my exim.
I want to graph the number of mails in different classes
like, blacklisted anyway, virus, whitelist, spamscore to high, spam-tagged, bounces, and normal mail
but currently anything rejected is tagged with 'rejected rcpt' and i have not yet a clue about how to change that.
unlofl
sorry, misunderstood
DukeDrake
[1]plex: thanx
unlofl
hm, I'm seeing those messages in my production log, its running with the default debian distro log settings
Groenleer
maybe i iam searching for the wrong text in the log?
looking for 554 5.7.1 (spam blocked) gives me only 8 mesages in a whole day, on a production server...
Either the spammers don't like me, or we just dont get enough spam ;)
unlofl
I'm using sa-exim, I see "Action: scanned but message isn't spam: score=0.0 required=5.0 (scanned in 1/1 secs" in my exim mainlog
a positive match looks like "Action: flagged as Spam but accepted: score=11.1 required=5.0 (scanned in 0/0 secs "
Groenleer
strange, i don't have that, even while Spam Assassin is running also as a transporter in exim (afaik)
unlofl
but spamassassin debug from spamd is written to syslog, that might be more along the lines of what you're looking for
if you don't really want more data on exim, just counts of how many messages are accept/rejected
Groenleer
incomming mail is however tagged correctly if i look at the headers
[1]plex
Hi. Is there a way to establish a begin retry rule based on origin IP address? Everything I have read details how to apply it to domains which are receiving email, not those who are sending. We host many clients who use our mail server for outgoing mail through our database application. I would like specific clients informed immediately if there was a problem contacting the recipient mail server.
petemc
what is it that you've read?
the best thing to read would be the exim spec, the retry configuration section
[1]plex
that is what I have read
it only focuses on retry intervals
« prev next »