logs archiveIRC Archive / Freenode / #exim / 2010 / January / 6 / 1
In a string expansion, how can I test whether a variable is part of a domain list?
Do I need to use forall or forany?
Nevermind. :_)
Now I have another problem: why is *@* ${if!match_domain{$2}{+local_domains}{$1@$2}fail} rfF
yielding "@" as an address when local_domains isn't matched?
I have a working exim4 MTA on Debian Etch and I have configured it to work with mailman according to this guide: http://www.debian-administration.org/article/Mailman_and_Exim4 but I still don't get messages to the list; I really don't know how to start troubleshooting it
my template: http://hiperactivo.com/~kandinski/exim4.conf.template if someone thinks they can help
and thanks in advance
Is it possible, using exim, to find out if my server is being blocked by another smtp server?
You could look in the logfiles.
Can someone help me decode the logfiles of possible non-delivered mails? It is about a bulk mailing going out, receiving 452 Too many recipients from mx*.hotmail.com, but also many what I interpret as second delivery attempts seperately afterwards.
I have grepped and awk'ed lists of both the 452'ed addresses and the 250'ed addresses, but both lists contain addresses not on the other list.
So it seems some undelivered mail wasn't delivered succesfully again, but also some mail that didn't throw any errors was somehow sent in seperate attempts rather than collected bulks.
I just want to be able to produce a list of people who didn't receive the mail.
are the undelivered mails not still in the queue?
petemc: I don't think so, this was on 30 dec
you cant check?
I don't receive the mailman bounces, and I'm not that familiar with the logfile format.
I'm mostly trying to explain the difference in both directions between the lists I have created.
There is some mail in the queue 8 total
Oldest 42 hours

not quite, what you're reading is for exim 3
so obsolete
Stupid google.
First hit on "exim log format"
aye, the exim site is/was badly set up
its being worked on
I thought the 3.2 meant HTML 3.2
I'll look up how I created my 452 and my 250 lists.
Actually I have since making those lists refined my regexps, I'll regenerate them with my newer more precise ones.
Hmm.. two different formats of logline depending on which 4XX error I got.
grep -B1 Completed /var/log/exim4/mainlog
I am only looking at a specific timespan.
The logfiles are per week or so.
But that yields 3954, of which 3556 in the timespan.
easy to filter on the timestamps
I know.
fantastic then
That's now what confused me.
What confused me is that some 452-errors didn't get a subsequent 250-success, and some 250-successes happened without a previous 452-error.
I am not sure whether exim has delivered the mails in blocks or seperate, and how large th blocks are, etc.
Which i why i am greping and awking and doing all sorts of stuff
exim -bP | grep max
Now I need to grab email-addresses from the same output but in different locaties in the line.
Let's just hope noone has a quoted string address.
I need a second pair of eyes to help me spot the error in:
# cat mainlog.1 | grep -E '^2009-12-30 23:(1[89]|2[0-4]).*: 4[0-9][0-9].*' | sed 's/.*\([^ <]+@[^ >]+\).*/\1/'
Mostly the sed part
Aprogas: Use exigrep instead of grep.
The grep works fine, it's the sed that doesn't.
This will eat at my brain until I figure it out.
It was -E
I was using extended regex format and sed uses basic by default.
Oh.. gnu sed doesn't have -E
I fixed it with sed -r
What would cause a simple mail (via mailman) to sit in the queue for say 24 minutes.
I am getting really confused today.
The conclusion of my research if you are interested is that only 15 of the 362 temporary errors wasn't deliverable. And in total 97.3% of all mail was delivered.
« prev next »