Mailwasher doesn't work, locally sent mails aren't delivered

[kirberich]

I just tried getting mailwasher to work, on Scalix 11 and Suse 10.1.

After following the howto (and following the suse specific hint regarding the linux.mc sendmail source file) mailwasher doesn't recieve any mails.

Additionally, I don't seem to be able to deactivate the changes I made.

With the SMTPFILTER=TRUE directive, no mail at all gets delivered, witout it, mails fetched and forwarded to a scalix user with fetchmail do get delivered, but locally send mails don't. (A cron script is supposed to locally send a mail every hour through sendmail, it worked before mailwasher and doesn't anymore.)

Sending mails from the scalix server doesn't work at all.

So my two questions are:
Where could I have gone wrong in the mailwasher configuration? (My guess is it's suse/sendmail related)
and how do I get my mailserver to work again?

[kirberich]

Oh, I almost forgot: omshowlog does not show any errors, at all.

[kirberich]

So apparently sendmail is the culprit, this is an exerpt from the mail.err logfile

Jun 27 11:38:41 mailserver sendmail-client[23240]: l5R9Xcv5023902: to=<test@blubbel.de>, delay=00:05:03, xdelay=00:00:00, mailer=relay, pri=570959, relay=[127.0.0.1], dsn=4.0.0, stat=Deferred: Connection refused by [127.0.0.1]

[Shredder]

I'll just make a informative post here, that I have used mailwasher and found it to be OK at doing what it is supposed to do. But if it fails for any reason, mail will not be delivered to the scalix box.

I find that it is way easier and better to have a seperate box for handling spam and virus scanning. This gives you a few benifits: 1) No extra load on the mail server. Big plus for large installs. 2) If the scalix server goes down (upgrade, something else), your mail will still be cached on the mail gateway. When the Scalix server comes up, the mail gets delivered. 3) No changes to the scalix box. If you do an upgrade to your scalix server with mailwasher installed on it, you will have to redo your smtpd.cfg file and fix the sendmail.mc files again.

Just my ramblings,
Shredder

[kirberich]

I very much like that idea. It's not practicable for me right now, but the concept sure makes a lot of sense to me, mostly because of the added redundancy. I'll keep it as an option for future scaling.

Thanks a lot.

The problem is solved btw, the reason was that sendmail commited suicide somewhere along the installation, slight reconfiguration and a clean restart of scalix fixed it.

[Shredder]

No problem. Glad everything got worked out for you.

If you have a VMWare server, there is a VM you can get for free from www.global-domination.org/ESVA.php that works fairly well. I have no affiliation with them, just liked the product (add free always helps!!!!)

Shredder

[kirberich]

On a side note, it would be a pretty nice addition to scalix if there was a way to get the server to fall back to normal e-mail routing when a plugged in service like an antispam or antivirus solution fails.

But oh well, it works now and I'm quite happy with it

[Shredder]

That won't happen as Scalix is not aware of the MailWasher stuff. All it does is send the message to a port and hope that something is there to receive it.

Shredder