[exim-dev] [Bug 2335] New: unexpected "past final cutoff" af…

Top Page

Reply to this message
Author: admin
Date:  
To: exim-dev
Subject: [exim-dev] [Bug 2335] New: unexpected "past final cutoff" after greylisting
https://bugs.exim.org/show_bug.cgi?id=2335

            Bug ID: 2335
           Summary: unexpected "past final cutoff" after greylisting
           Product: Exim
           Version: 4.91
          Hardware: x86
                OS: Linux
            Status: NEW
          Severity: bug
          Priority: medium
         Component: Delivery in general
          Assignee: nigel@???
          Reporter: wbreyha@???
                CC: exim-dev@???


We got a report from one of our users that he was unable to send E-Mail to a
recipient @cims.nyc.edu.

I checked our logs and IMO Exim failed to maintain the retry DB correctly in
this case.

The timeline...

On 30.10.2018 21:25:23 our user FROM_A@??? sent email to
RCPT_A@???.

cims.nyc.edu has two MX RRs. The first one greylisted with
Oct 30 21:25:31 grace exim[21251]: 1gHaZn-0005Wi-H1 H=MX.cims.nyu.edu
[128.122.49.99]: SMTP error from remote mail server after RCPT
TO:<RCPT_A@???>: 451 4.7.1 Greylisting in action, please come back
later

The second accept the email:
Oct 30 21:25:40 grace exim[21249]: 1gHaZn-0005Wi-H1 => RCPT_A@???
R=external_delivery T=remote_smtp M8D=0/0 H=MX2.cims.nyu.edu [216.165.22.57]
X=TLSv1.2:DHE-RSA-AES256-GCM-SHA384:256 CV=yes
DN="/C=US/postalCode=10012/ST=New York/L=New York/street=251 Mercer
Street/O=New York University/OU=Courant Institute/CN=mx2.cims.nyu.edu" C="250
2.0.0 w9UKPbdM022878 Message accepted for delivery" QT=17s DT=8s
Oct 30 21:25:40 grace exim[21249]: 1gHaZn-0005Wi-H1 Completed QT=17s

Yesterday, 06.11.2018 17:36:58, the same sender sent E-Mail to
RCPT_A@??? and a second RCPT_B@???.

This time both recipients got greylisted by both MX hosts:
Nov 6 17:37:06 grace exim[30258]: 1gK4La-0007rv-3d H=MX.cims.nyu.edu
[128.122.49.99]: SMTP error from remote mail server after RCPT
TO:<RCPT_A@???>: 451 4.7.1 Greylisting in action, please come back
later
Nov 6 17:37:06 grace exim[30258]: 1gK4La-0007rv-3d H=MX.cims.nyu.edu
[128.122.49.99]: SMTP error from remote mail server after RCPT
TO:<RCPT_B@???>: 451 4.7.1 Greylisting in action, please come back
later
Nov 6 17:37:13 grace exim[30252]: 1gK4La-0007rv-3d == RCPT_A@???
R=external_delivery T=remote_smtp defer (-44) H=MX2.cims.nyu.edu
[216.165.22.57]: SMTP error from remote mail server after RCPT
TO:<RCPT_A@???>: 451 4.7.1 Greylisting in action, please come back
later
Nov 6 17:37:13 grace exim[30252]: 1gK4La-0007rv-3d == RCPT_B@???
R=external_delivery T=remote_smtp defer (-44) H=MX2.cims.nyu.edu
[216.165.22.57]: SMTP error from remote mail server after RCPT
TO:<RCPT_A@???>: 451 4.7.1 Greylisting in action, please come back
later

To my surprise the next log lines show:
Nov 6 17:37:13 grace exim[30252]: 1gK4La-0007rv-3d ** RCPT_B@???:
retry timeout exceeded
Nov 6 17:37:13 grace exim[30252]: 1gK4La-0007rv-3d ** RCPT_A@???:
retry timeout exceeded
Nov 6 17:37:13 grace exim[30360]: 1gK4Lp-0007tg-CY <= <> R=1gK4La-0007rv-3d
U=mail P=local S=3165 M8S=0
Nov 6 17:37:13 grace exim[30252]: 1gK4La-0007rv-3d Completed QT=15s

In close succession the same sender retried to send the email and failed in the
same way as above. Both hosts greylisted and the email bounced because of
"retry timeout exceeded".

exinext cims.nyu.edu shows:
Route: RCPT_A@???:<FROM_A@???> error -44: H=MX2.cims.nyu.edu
[216.165.22.57]: SMTP error from remote mail server after RCPT
TO:<RCPT_A@???>: 451 4.7.1 Greylisting in action, please
first failed: 30-Oct-2018 21:25:40
last tried: 06-Nov-2018 17:38:46
next try at: 06-Nov-2018 23:38:46
past final cutoff time
Route: RCPT_B@???:<FROM_A@???> error -44: H=MX2.cims.nyu.edu
[216.165.22.57]: SMTP error from remote mail server after RCPT
TO:<RCPT_B@???>: 451 4.7.1 Greylisting in action, please com
first failed: 06-Nov-2018 17:37:13
last tried: 06-Nov-2018 17:38:46
next try at: 06-Nov-2018 17:43:46

Note the "past final cutoff time" for RCPT_A and "first failed: 30-Oct". RCPT_B
has no "past final cutoff"

There are no entries for the hosts themselves in my retry DB.

I "clear" my retry db with "exim_tidydb -t 7d" what is "kind of" recommended in
53.14.

Shouldn't the successful delivery on 30.10.2018 "clear" the retrydb? Shouldn't
the delivery to RCPT_B succeed even if RCPT_A is "past final cutoff"?

--
You are receiving this mail because:
You are on the CC list for the bug.