error message, email stopped

[Kino]

I sent to a list of just over 1600 subscribers. About halfway through mailing I received this error message (email address removed):

Sending error. Check your mail settings.

Queue Row: queueid=a8e4acc55199bc84d1683f408f741fef msgtype=1 userid=22781 msgid=583 email=[email removed])

edit: I looked up the domain for the email address, and it was not working at the time I sent the email.

Things have been running quite well with the latest version. Sorry to see this happen.

K

[DW]

Were you able to complete your mailing?  It's hard to say what the exact SMTP response was.  If you enable extra debug information in admin.php you will be shown more information during your next mailout.

change this line:

// $bugs = 1;

to

$bugs = 1;

[Kino]

Everything has been working fine, hence no reply to this thread from me for some time. I encountered a similar problem again, whereby the mail out hangs when it finds an email address it does not like.

My remedy is not a good one: try Sendmail instead of SMTP. Not only does it slow down my server's mail queue, but it takes hours to send to a small list of 500 recipients.

***

Here is the message again, with bugs set to ON.


Sending Mail  of 560..
PIPELINE-FROMmsg=250 2.1.0 ... Sender ok .
RCPT TO: <snip@hawaii.rr.com>
RCPTmsg=DATAmsg=.

Sending error. Check your mail settings.

Queue Row: queueid=a9aee0bbc07f9666081be8881e0a84be msgtype=1 userid=26440 msgid=629 email=snip@hawaii.rr.com)

Here's ANOTHER one, same list (they only send out 3 to 4 times per year, so the list of addresses gets a bit stale). I deleted the offending address above, and resent to this list (I moved the ones that already rec'd emails to the bounced list temporarily as there were only 12 at that point).


PIPELINE-FROMmsg=250 2.1.0 ... Sender ok .
RCPT TO: <snip@covad.net>
RCPTmsg=DATAmsg=.

Sending error. Check your mail settings.

Queue Row: queueid=e915d4527b57691374e5e0de501fffb7 msgtype=1 userid=26467 msgid=630 email=snip@covad.net)

Server said:
Go to Config


Dean, can you help on this problem??

Thanks,

K

[Kino]

I have determined that this bug is caused when a web host is down, even temporarily. I sent out an email from a domain that was down (not expired). I get this error and everything grinds to a halt.

There is no workaround with LM Pro except to reset to Sendmail rather than SMTP. The former is slow and unproductive.

Can we get a fix for this so LM Pro does not stop when domains are down?

Thanks.

K

[DW]

Thank you for your excellent insight into this problem.  This will be fixed in the next update.

[drs]

Hello,
we bought listmail recently and find it great except several sending problems.

When I try to send email via the web interface to a large quantity list  the sending is stopped with the message:

Sending error. Check your mail settings.

Server said:

after sending between 200 and 1000 emails.

I activated the $smtp_debug = '1'; option and in the debug log file I see:

> MAIL FROM: <bounce@ourcompany.net>
> RCPT TO: <xxx@concepta.com>
250 2.1.0 <bounce@ourcompany.net>... Sender ok
LM: Undeliverable. RCPT response: . Skipping.
> RSET
> MAIL FROM: <bounce@ourcompany.net>
> RCPT TO: <yyyy@freesbee.fr>
250 2.0.0 Reset state
250 2.1.0 <bounce@ourcompany.net>... Sender ok
> DATA
250 2.1.5 <yyyyy@freesbee.fr>... Recipient ok


and it is there all is stopped. When I click on the resume button, sending is correctly resumed until the next stop.

It appears that this happens when there is an smtp error and the RSET command does not seem to work.

Can you please help us? Our listmail version is 1.8

Thanks in advance.

[DW]

It appears that either the server or ListMail is not doing things in the right order or is delayed.  Would you mind me doing some manual testing (to a test list, of course)?  I would need you to submit your information here with a note about the problem.

Regards

[drs]

Hello,
I sent the infos you nedeed via the submit form.
did you get access to my server?

[drs]

It seems as if after receiving the RSET command the smtp connections fails.
Is there a quick way to close and reopen the connection to the SMTP server?

[DW]

Yes, I have your information - I will be taking a look ASAP.

Regards

[cheresources]

After having a nearly identical problem to the above, I set up a test list with three different domains (total of three addresses).  SMTP mailing worked flawlessly.  When I try it with my 20,000 member list, the mailing gets stuck at various numbers of mailings and the bounces start flying back to me with 550 errors when I know that the addresses are good because they work fine with internal Sendmail.

I know Dean will come through on this one...now I have to go apologize to my tech support person at my ISP.  I doubt he has any hair left

[DW]

I'm working on it now. I thought I had these responses licked. :x

I'm having some technical difficulties getting connected to drs's server.

cheresources, would you like to submit your information too so that I can investigate?

Regards

[DW]

I apologize, I should have provided a link.  PM's are supposed to be disabled and some buttons have been removed from the forum making it difficult to access messages.  Please submit here, thanks!

Regards

[DW]

I am making progress on this issue.  drs has a large live list and I was able to keep resuming his mailing to narrow down the problem.

It seems that the SMTP server (with this particular mail software) 'locks up' when checking that certain domains are deliverable.  After 20-30 seconds of waiting, ListMail is given a "blank" reponse to the RCPT command.  Unlike a normal failure such as "User Unknown", this one seems to stop the SMTP server from responding to future commands.

I'm almost finished with a workaround that skips the user and reconnects to SMTP to continue sending.  However, I wonder if I can further improve the new change.

Each time a bad address does this it takes, as I mentioned, 20-30 seconds before ListMail can continue.  If you have a lot of these it can add up and your overall speed will decrease.

So, what should I do?

1) Ignore and accept the delays - mailings will go out unterrupted but with a few delays.

2) Automatically and immediately flag the user as Bounced

3) Run the user through the bounce feature, so if they fail X times in X days they are flagged as Bounced

I would say choice #3 is optimal - how about you?  Should  I also use the bounce feature with other non-success RCPT messages?

Regards!

[DW]

I have updated the v1.8 files in the members area with my workaround.  Currently, the failing users will be skipped and ListMail will reconnect.

Regards!