Amazon SES Send Rate is Slow

I’ve been using PHPlist for a couple years now on a couple of lists. One is roughly 28,000, the other a few thousand. I upgraded the PHPList version in November and had to tinker with some things to get Amazon to send at all, but eventually got it working and send rate was as expected. Went to send a newsletter today, and the send rate is about 14 emails/minute. It should be nearly 1000 a minute based on my account and settings. I’ve updated to the lates PHPList version, updated and checked database, even tried a brand new install on a different server. Same results. I’m using the Amazon SES plugin and the API for sending. Tried without and it wouldn’t send at all. I used to love PHPList because it just worked. Over the last few months…NOT so much. I’ve wasted hours of time troubleshooting this over the last few sends. Anyone have any suggestions? As noted, Amazon SES worked perfectly for 2 years then problem cropped up in November and not again today.

Chris Blair

@videomicloud Which version of the Amazon SES plugin are you using, you can find that on the Manage Plugins page.
You could try enabling the verbose curl output on the Settings page. The log output might give some idea on where the time slowdown is happening.

Latest versions on everything. Just upgraded PHPList version, all plugin versions etc. PHPList is vers. 3.3.9 ; Amazon Plugin is version 1.2.1+20171206 (from Github). Are you talking about the verbose curl on the Amazon settings? Will enable that.

Chris

Yes, although be aware that will generate a log of debug output.
Are you using the concurrent sending with curl multi?

Yes on concurrent/curl multi, which worked fine for months prior.

Chris

@videomicloud Also, what are you batch / throttle settings, and how frequently is processqueue run?

I’ve never used any batch/throttle settings in past. I simply relied on Amazon SES limits (which were insanely high, like 10,000/minute or something like that). Queue is processed every minute I believe.

Here’s a screenshot of log events. I’m looking for more detailed log info but not exactly sure where to get that.

Chris

Another error I’m finding in php-fpm.log

[11-Feb-2019 15:52:26] WARNING: [pool phplist] server reached pm.max_children setting (5), consider raising it

The curl log file is in the phplist $tmpdir directory.

The http code 100 doesn’t look to be handled correctly. Has that always been happening?

My PHPList is on a Bitnami server and NOTHING is where it typically is, so I’m not having any luck finding that log. This is all probably over my head anyway. I’m going to try a fresh install of PHPList on a different server and import in my subscriber lists and see if that fixes. The Bitnami server is getting old anyway and it’s impossible to upgrade PHP on it without jumping through 1000 hoops. The current lists is going out, just really slowly. So will let if finish and start fresh. Hopefully that works as I love PHPList when it works!

Chris

@videomicloud You appear to be processing the queue through the browser. Instead you should use a cron job, which I thought you were using if previously phplist was sending at 1000/minute.

That was the issue. I deleted the existing cron job, created a new one, ran it and poof….entire list went out in a matter of a few minutes. Not sure what was going on with the configuration I had but that fixed it. Thanks so much for your help!

Chris