How do I resolve an execution timeout error when executing a long running php script?

You can add a column (or an ancillary table) in the database, and mark all addresses as initially "to be worked".

Then your script could:

- SELECT the number of worked addresses and total addresses
- display a progress bar
- SELECT the lowest 100, or 50, or N addresses yet to be worked
- send the emails and mark them worked
- issue a javascript to refresh itself.

Then the client will see a page where a progress bar continuously advances, and never times out (as long as "N" is small enough).

You could also do this in AJAX, but that might be overkill.

Moreover, you can employ some tricks (if you don't already do) to speed up operations. One possibility is to "coalesce" several addresses as Bcc:, if the email body is identical; that way you deliver several messages in one call. Take care to not exceed whatever Bcc: limit your ISP might enforce. This strategy uses more resources ISP-side, so check with them.

Another possibility (can be used together) is to send emails sorted by domain. Several mail servers attempt, if possible, to send emails to the same server in one connection to save resources and time. If you send a group of emails all to the same domain, you make things easier for the server.

In this last case you would SELECT emails like this:

SELECT * FROM addresses
    WHERE (worked=0 AND active = 1)
    ORDER BY SUBSTRING_INDEX(email, '@', -1)
    LIMIT 20;

Then (example with mysql functions - PDO are better),

while($tuple = mysql_fetch_assoc($exec))
{
    if (send_mail_to(...))
         $sent[] = $tuple['id'];
    else $failed[] = $tuple['id'];
}
// Now mark both SENT and FAILED as worked
$sentm = implode(',', $sent);
$failm = implode(',', $failed);

// UPDATE addresses SET worked = 1 WHERE id IN ($sentm,$failm);
// UPDATE addresses SET active = 0 WHERE id IN ($failm);

If you saved the start time in a PHP session, you could even display a nice bar-with-dashboard such as

+--------------------+
|########            | 
+--------------------+
40% processed
38% delivered
 2% failed
Expected time remaining 5 min 17"