![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
We have a Webuntu server that sends out email campaigns via an Interspire Email Marketer program. It sends the email campaigns to a HAProxy load balancer Ubuntu server which then distributes emails in round robin between SMTP relay servers. The Interspire program sends out emails on an individual basis, 1 by 1 to whatever SMTP server is specified.
When an email campaign starts, netstat shows a SYN SENT for whatever email is currently being attempted to be sent to the HAProxy server, but the HAProxy server doesn't establish a connection until about around 100s to 1000s of emails have failed to send because HAProxy doesn't respond to the SYN SENT request.
I've sent out small sized test email campaigns to other SMTP servers (Exchange, internal SMTP relay server, and third party SMTP server), and the emails get sent out without a problem.
So I'm fairly certain that the problem source is the HAProxy server, but I don't understand why it takes the HAProxy server so long to finally respond properly to a SYN SENT message. Once it finally does, the remaining email campaign QTY (~220K) gets sent out successfully.
Both servers are on the same network. They are both VMs located on separate ESX 4 hosts.
Thanks for any help.
When an email campaign starts, netstat shows a SYN SENT for whatever email is currently being attempted to be sent to the HAProxy server, but the HAProxy server doesn't establish a connection until about around 100s to 1000s of emails have failed to send because HAProxy doesn't respond to the SYN SENT request.
I've sent out small sized test email campaigns to other SMTP servers (Exchange, internal SMTP relay server, and third party SMTP server), and the emails get sent out without a problem.
So I'm fairly certain that the problem source is the HAProxy server, but I don't understand why it takes the HAProxy server so long to finally respond properly to a SYN SENT message. Once it finally does, the remaining email campaign QTY (~220K) gets sent out successfully.
Both servers are on the same network. They are both VMs located on separate ESX 4 hosts.
Thanks for any help.