Mautic says 'emails sent', but nothing has actually gone out

[modifiedcontent]

Your software
My Mautic version is: v4.4.7
My PHP version is: 8
My Database type and version is: MariaDB 10.3.38-MariaDB

Your problem
My problem is:

Channels>Emails says about 1100 emails have been sent, but nothing actually has gone out - 0 read, 0 bounced emails, etc.

I have been sending emails with this Mautic installation for years, but have never seen this. What could have happened? What should I check to troubleshoot?

I tried a small test mailing to a test segment with 3 of my own addresses. Same result; Mautic says ‘sent’, but nothing actually gone out. With ‘Send Example’, the email does arrive fine - email configuration still works.

These errors are showing in the log:

Which log…? Looking for it now…

‘Today’s Mautic error log is empty. Check server error log for error messages Mautic didn’t have a chance to catch.’

error_log in the installation folder is empty.

Steps I have tried to fix the problem:

Don 't know where to start yet.

Edit:

Based on this, when Mautic says emails are sent, could it still mean the cronjob hasn’t run? Where is the limbo where these emails are stuck?

Could I send emails stuck in limbo with an SSH command? I’ve tried:

php bin/console mautic:emails:send

Something has now gone out. I’ve received one bounce message; it still does not look like a normal full mailing. Why did/do my emails get stuck in spool limbo?

[joeyk]

Correct. They were all in the spool.

[modifiedcontent]

Thanks. But why were they stuck there? Why didn’t they go out? What should I check/fix?

Only one or two emails have gone out now. Most of the mailing was lost, probably deleted from the server after a number of seconds you can set in Mautic configuration?

Mailings worked fine before. What broke/changed? I think my upgrade to PHP 8 was after my last successful mailing. Or did anything change in recent Mautic upgrades that requires email config or cronjob changes?

[joeyk]

Hi,
You need to test this:

1. check how many emails are in the pool.
   This command counts the files:
   ls | wc

2. If there are any try to run your mautic:emails:send command.
   Did it send out the emails?

• If yes, then this command is configured wrong in your crontab. (probably not the case). Check for typos.
• If no, then check error log.
  – Any related entries?
  – Do you get an error? For example some of the files cannot be read?
  – Do you see any files, with NOT .message extension?

3. If there are no emails in your spool folder, then make a test send and start from step 1.

[modifiedcontent]

Thank you @joeyk !

This is an error that shows up in logs when I try to force send via the command line:

“rename(): supplied resource is not a valid Stream-Context resource”

Someone here mentions the same error. I’ll look into this further later…

The initial sending attempt produces errors in logs similar to these - Error 503: Valid RCPT command must precede DATA - but I’m confused if that applies to the recipient or my server. I have never had problems with SMTP authentication before and haven’t changed any settings.

Initial send attempt error in logs:

[2023-04-24 22:25:23] console.CRITICAL: Error thrown while running command “mautic:emails:send”. Message: "Expected response code 354 but got code “503”, with message "503-All RCPT commands were rejected with this error: 503-The mail server could not deliver mail to jane.doe@recipient.co.uk. 503-The account or domain may not exist, they may be blacklisted, or missing 503-the proper dns entries. 503 Valid RCPT command must precede DATA " Log data: >> . << 250 OK id=1pr4cd-00043a-1E >> MAIL FROM:mod@mydomainname.com >> RCPT TO:another.recipient@website.org.uk >> DATA << 250 OK << 250 Accepted << 354 Enter message, ending with “.” on a line by itself >> . << 250 OK id=1pr4cd-00043a-23 >> … etc.

[joeyk]

@modifiedcontent

rename(): supplied resource is not a valid Stream-Context resource

Sorry, no idea.

[modifiedcontent]

Working on this again this weekend, trying to restart my business from another stand-still. Sent out another mailing to a small test segment of other email addresses I own - gmail etc. Nothing arrives. No errors in logs.

Trying to force sending from command line with php bin/console mautic:emails:send gives this error:

In ProcessEmailQueueCommand.php line 141:

  rename(): supplied resource is not a valid Stream-Context resource

mautic:emails:send [--message-limit [MESSAGE-LIMIT]] [--time-limit [TIME-LIMIT]] [--do-not-clear] [--recover-timeout [RECOVER-TIMEOUT]] [--clear-timeout [CLEAR-TIMEOUT]] [--lock-name [LOCK-NAME]] [--bypass-locking] [-t|--timeout TIMEOUT] [-x|--lock_mode LOCK_MODE] [-f|--force] [-h|--help] [-q|--quiet] [-v|vv|vvv|--verbose] [-V|--version] [--ansi] [--no-ansi] [-n|--no-interaction] [-e|--env ENV] [--no-debug] [--] <command>

What is this rename() function? Where is it used in the process? How can I find out?

When I use ls | wc as suggested by @joeyk I get this as response:

26      26     309

What does that mean? If 26 or 309 is the number of emails in the pool/queue (?) that is way more than I would expect.

With php bin/console mautic:messages:send I get this response:

Processing message queue
Messages sent: 0

The rename() error is also mentioned here and here - will research there further…

[modifiedcontent]

Line 141 in ProcessEmailQueueCommand.php is the following:

        //set spool message limit
        if ($msgLimit = $input->getOption('message-limit')) {
            $commandArgs['--message-limit'] = $msgLimit;
        } elseif ($msgLimit = $container->getParameter('mautic.mailer_spool_msg_limit')) {
            $commandArgs['--message-limit'] = $msgLimit;
        }

I don’t see rename() there, so where does that come in? What does ‘supplied resource is not a valid Stream-Context resource’ mean? How does this work together?

I have not made any changes to message limits configuration btw; sending worked fine before with the same settings.

Here’s another one with the same error message. Unfortunately @mindeswx’s solution there does not work for me; I have always had that ‘Use owner as mailer’ setting set to ‘No’.

I guess to use ls | wc you have to change directory to the spool folder first? That is why I got unexpected results earlier I guess.

Where is the spool folder? Here according to Google:

You should find it in either /var/spool/mail/ (the traditional location) or /var/mail (new recommended location).

Could it be that the spool folder on my VPS was changed in an upgrade and that Mautic now can’t find the folder?

Result of ls | wc on /var/spool/mail/ is 0 0 0

Result of ls | wc on /var/mail/ is also 0 0 0

In both cases I would expect two emails that Mautic claims have been sent, but never arrived (in my test email boxes).

[mzagmajster]

Mautic has its own spool dir., usually in /var/spool

If I would have to guess something interrupted the process and now some files are in invalid state.

Another option is that the spool folder/content has wrong owner (make sure is the same user as runs cronjob)

[modifiedcontent]

Thanks @mzagmajster. There were a lot of .message files stuck (?) in that spool folder. I am deleting them now. How can files end up in an ‘invalid state’? What does that mean?

If there were an ownership mismatch between cronjob and spool, I should still have been able to send out the emails with php bin/console mautic:emails:send, right? So I think I can eliminate that issue?

I found the cron jobs I set up years ago in /var/spool/cron. Two of them look like this:

*/3 * * * * /usr/local/bin/ea-php74 /home/myusername/public_html/mautic/bin/console mautic:emails:send >/dev/null 2>&1

I guess /ea-php74 could cause problems, since I have remove php7 and am on php8 now. Other lines have /usr/local/bin/php, so I’ll change it to that.

I also saw this “error” in the logs. I am trying to send from an email address that is an addon/alias of the main domain on my server - always worked fine in the past. Maybe there is a handover or certificate problem there somewhere. I have tried to clean that up on my server, whitelist my IP address as allowed sender, etc.

[mzagmajster]

Like I said, something probably went wrong or the process got interrupted and as a result the script cannot pick up where it left of. From what you have provided its not possible to know how exactly it went wrong.

@modifiedcontent

If there were an ownership mismatch between cronjob and spool, I should still have been able to send out the emails with php bin/console mautic:emails:send, right? So I think I can eliminate that issue?

Not necessarily. It depends on which user on OS you use to send and how mautic permissions / ownership on the files is configured.

Mautic filesystem should be properly locked down and most of the files should only be readable from the webserver´s point of view.

Cronjob should run by the same user that runs the webserver.

I would stop all the sending from mautic create a test segment with just you as a contact and try to first send one email (to yourself). When this works I would again try to send to real contacts.

[modifiedcontent]

Thanks @mzagmajster. Having ea-php74 in the cron job caused the problems - I had upgraded to php8 and removed 7 since the last mailing. After replacing it with /usr/local/bin/php the test mailing went through.