Topic: SMTP unavailable problem

[stdean]

I've been having a problem sending out mails with my SME box. I will occasionally (every 2 or 3 mails) receive an error message telling me that my SMTP server is either unavailable or refusing connections. Here's the message I receive:

Send Message Error

Sending of message failed.
The message could not be sent because connecting to SMTP server mail.mydomain.com failed. The server may be uavailable or is refusing SMTP connections. Please verify that your SMTP server setting is correct and try again, or else contact your network administrator.

I've tested it using different mail clients on different platforms (majority of users are using Thunderbird 1.5 on Centos 4.4, a handful are using Thunderbird 1.5 on WinXP SP2) and I've pretty much come to the conclusion that it has something to do with the smtpd & ssmtpd Instances and InstancesPerIP settings.

I followed the thread here http://forums.contribs.org/index.php?topic=34645.0
and increased my Instances and InstancesPerIP settings for both smtpd and ssmtpd.

Here is what they are set to now:

smtpd=service
    Authentication=disabled
    Instances=40
    InstancesPerIP=20
    MaximumDateOffset=0
    PatternsScan=disabled
    Proxy=enabled
    TCPPort=25
    TCPProxyPort=25
    VirusScan=enabled
    access=public
    status=enabled
    tnef2mime=enabled

ssmtpd=service
    Authentication=enabled
    Instances=40
    InstancesPerIP=20
    TCPPort=465
    access=public
    status=enabled

I would imagine that these should be more than enough, I've got a approx 60 users spread across 3 locations with a split of 30/20/10.

The question I have is, if I increase the InstancesPerIP and Instances to something very high like 100 would this resolve my issue? And if so could it also affect performance (increased overhead etc..)

This issue is happening across all the locations, so it does not appear to be tied to and location connections (2 have fiber connections, the other has DSL).

Any help is appreciated.

Cheers,

Conor

[mmccarn]

Just to check the obvious - have you run signal-event email-update to activate your new settings?

My first guess is that you're getting this message from users at sites that  have a NAT device between them and your SME server - so your 10 (or 20, or 30) remote users all look like the same IP address to your SME server.

CharlieBrady (in the link you referenced) seems averse to increasing InstancesPerIP; I think you might want to run netstat -an | grep :25 and look at /var/log/qpsmtpd/current to try to figure out where your smtp connections are coming from before making huge changes to your config...

In past versions of SME you could easily change the behavior of lots of internal settings by adding the IP of a remote site under "Local Networks", but this a) may not apply at all and b) has important security implications, so be careful...