Hi,
One of my customers sends out an email to her customer database of nearly 4,000 people via the osCommerce store php email function.
Only about 2/3rd's of her customers are getting the email before her send page locks up and seems to stall.
Does anyone know where to look for this problem?
On CentOS 3.7 and plesk 8. This is the settings in /etc/php.ini
max_execution_time = 30 ; Maximum execution time of each script, in seconds
max_input_time = 60 ; Maximum amount of time each script may spend parsing request data
memory_limit = 32M
For an example I sent out a test email through another client's osCommerce store to 598 people and it took about 25 minutes to complete without any problems that I could see. This seems like a long time to me and would be incredibly long for nearly 4,000 emails.
Any help is greatly appreciated.
php emails are timing out
Hi,
I had looked at the maillogs right after a client sent out to her database and there were no errors.
A reply from swsoft said that qmail only sends out 20 emails at a time and that was why it took over an hour or that I had tweaked 7.5.4.
The problem with that logic is that on Plesk 7.5.4 I did not have qmail-scanner, clamav, clamd, or the regular spamassassin on my server. I was using psa-spamassassin and dr. web and never had a complaint about sending out emails from my php stores. Nor did I tweak any settings for qmail to send out more emails as swsoft suggested.
Then, after swsoft updated my server to plesk 8 without my permission I removed the psa-spamassassin and dr. web and installed the spamassassin in your psa-compatible section plus qmail-scanner, clamav, and clamd.
Also, I have watchdog but it is disabled and have not changed anything in the firewall.
My spam problem is way down using your programs but is there anyway that qmail-scanner could be causing this slow down? I also have your razor-agents, dcc, and pyzor running.
CentOS 3.7
Intel 2.4 GHz
1 GB RAM
Each time someone sends out emails to their customer database in osCommerce stores my server loads hit the ceiling and the server slows down. I am planning on upgrading my server so will this help?
I had looked at the maillogs right after a client sent out to her database and there were no errors.
A reply from swsoft said that qmail only sends out 20 emails at a time and that was why it took over an hour or that I had tweaked 7.5.4.
The problem with that logic is that on Plesk 7.5.4 I did not have qmail-scanner, clamav, clamd, or the regular spamassassin on my server. I was using psa-spamassassin and dr. web and never had a complaint about sending out emails from my php stores. Nor did I tweak any settings for qmail to send out more emails as swsoft suggested.
Then, after swsoft updated my server to plesk 8 without my permission I removed the psa-spamassassin and dr. web and installed the spamassassin in your psa-compatible section plus qmail-scanner, clamav, and clamd.
Also, I have watchdog but it is disabled and have not changed anything in the firewall.
My spam problem is way down using your programs but is there anyway that qmail-scanner could be causing this slow down? I also have your razor-agents, dcc, and pyzor running.
CentOS 3.7
Intel 2.4 GHz
1 GB RAM
Each time someone sends out emails to their customer database in osCommerce stores my server loads hit the ceiling and the server slows down. I am planning on upgrading my server so will this help?
-
- Atomicorp Staff - Site Admin
- Posts: 8355
- Joined: Wed Dec 31, 1969 8:00 pm
- Location: earth
- Contact:
Sure, content scanning isn't a free process. The more tests and scanners you use, the more processing overhead and time it takes. The logs will tell you how long SA is taking to process each message. It will also automatically bypass SA if it takes longer than 30 seconds. Generally on my systems its taking about 2-4 seconds per message.
Also look into bumping up your concurrencyremote, just keep in mind that you're pumping even more data into the content scanners. It could also be that you've just reached the capacity of the box to handle all the stuff you're doing on it.
Also look into bumping up your concurrencyremote, just keep in mind that you're pumping even more data into the content scanners. It could also be that you've just reached the capacity of the box to handle all the stuff you're doing on it.