I am looking at using SES for a small 1,500 email list that is not super time critical.

Setup went fine - exiting the sandbox was easy - daily quota of 50k more than I will ever need.

14 emails per second - that sounds like it could be an issues - but hard to tell - How fast is the SMTP call in CIVI.Mailer

Can I use throttle time and add 100,000 microseconds between each email to lower my rate to 10 per second. For that matter I could do 500,000 and do 2 per second

Trying to avoid having to use a postfix installation.



  • 1
    Can't answer your question directly but we bailed on SES precisely because of the rate limits. Not so much that they are unmanageble, but that if you ever DO exceed them you'll never know it, because there's no bounce, error or log. There are many SMTP services out there that allow free access for low volume. We are using Sparkpost, which allows up to 100k per month for free. And there are two CiviCRM extensions to manage it.
    – Rob Brandt
    Commented May 4, 2017 at 21:06

1 Answer 1


I've used Amazon SES for a similar sized mailing list to you for a client. I was confused initially and then surprised when they gave me a 14/s send limit.

However, when you work it out that's just over 50,000 per hour so the only issue is how to keep below the 14/s limit. Well, as you've mentioned, CiviCRM has the very helpful "Throttle Time" which you can set to 75000microseconds to limit the send rate from CiviCRM to just under 14/second.

The client has found SES to be extremely fast and reliable. Just make sure you've got access to your DNS so you can configure necessary DNS entries to allow SES to send email for you.

  • Thanks - I also pinged AWS - and they were kind enough to bump to 90 a second so between that and Throttle time - I should be good. Commented May 5, 2017 at 22:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.