Invision Community 4: SEO, prepare for v5 and dormant account notifications By Matt Monday at 02:04 PM
npnchicago Posted June 5, 2018 Posted June 5, 2018 After switching to SMTP method of sending mail, we frequently get 421 Timeout errors when sending mail. We switched on June 1 from using PHP method and sendmail on the server, to SMTP direct. Both methods are using AWS SES as the primary mail server. Once we switched we get many "421 Timeout" and "Mail could not be sent" errors. We send about 600 mails per day, well below the limits imposed by SES. The full logs in the IPS Email Logging are not much help unfortunately, containing only simple or sometimes no information: We can resend the mails in the Error Log and they seem to send without issue, not bouncing or having the same timeout issues. IPC says there is not much more that can be done on the IPC side b/c it is a server issue. Are there any other ways you can suggest to help us resolve this issue?
bfarber Posted June 6, 2018 Posted June 6, 2018 This type of issue would definitely be a server-side problem that you would need to relay to your host. Let them know that SMTP emails are intermittently failing with a 421 Timeout message when using SES.
npnchicago Posted June 6, 2018 Author Posted June 6, 2018 Thank you for responding. The host is AWS EC2 and mails are being sent directly via IPS SMTP to SES on their servers. So it never really leaves their servers. And oddly, with "Mail cannot be sent errors" there is no log, no information at all. This suggests a connection issue, sure, but why is there no log at all? There is no error, no connection information, etc. It seems to me that perhaps IPS is only recording a certain portion of the conversation, because there is no connection, no handshake, nothing (in the log). Since mails DO go through all the time (errors are only about 1% of our total), there must be some connection, handshake, etc that is happening correctly. Mike
bfarber Posted June 7, 2018 Posted June 7, 2018 Because it's timing out when attempting to connect, so there would be no handshake information (i.e. the software never sent any further commands). https://docs.aws.amazon.com/ses/latest/DeveloperGuide/smtp-issues.html Possibly you're hitting their default sending limits and need to request them to be raised (see the second bullet point). That would definitely explain why it works 99% of the time.
BomAle Posted June 7, 2018 Posted June 7, 2018 If you would a solution to avoid exceed limit you can install my app and configure with restrictions imposed by smtp service. It manage 421 emails and write into mail error logs some detail to help you what happen on delayed emails.
Recommended Posts
Archived
This topic is now archived and is closed to further replies.