Hi all

i just installed the new vesion of clamwall to 1.4 and upgrade to clamav 095a

 

all work fine!!

 

thanks a lot for this great support!!!

 

Victor

search mailixf and you find a few cgi tools

[quote user="bwhite"]Hi PaulW,  not sure how that would work in my situation in that we currently have a pop3 account that is our Domain Mailbox.  All mail sent to an address in our domain is collected there.  We then have mercury retreive the messages from that account to distribute to our internal users.[/quote]

That's a fairly standard setup, but has various drawbacks.

[quote]To use MercuryS we would have to have our Email host setup to periodically send all mail collected to our email server, right? [/quote]

No, that wouldn't gain you anything.  You must be the email host by changing the domain's mx record to point to your server.

[quote]With our ISP we have a problem with our "Static" IP addresses not being so Static, so that we be another problem.[/quote]

Ah, that would cause complications.

<title>blank_page</title><link href="/tiny_mce/jscripts/tiny_mce/themes/CommunityServer/css/editor_content.css" rel="stylesheet"><base href="http://community.pmail.com/forums/">

You really shouldn't have an AV solution that intercepts traffic at the tcp

level (most do unfortunately).

In a client this is perfectly normal, but in a mail-server you need a server

product, as the connection is supposed to be initiated from the other end (like

a magnet - opposite polarity) - the sockets are always open for new inbound

connections, and multiple connections has to be allowed.

For this very reason you should use an AV product that examines the content

for viral content after the content has been received.

I'm not familiar with Mercury on Netware or how different it is from Mercury/32 so YMMV.

You should turn on session logging for the SMTP client module to capture a session transcript of a message that causes the error (using a very small attachment is a good idea [:)]) and post it here.

 

> Today I saw again (it is still) that on some reason MercuryC
> (installed on NW 5.1) stuck on message delivery. I load another
> instance (as Mr. Stephenson suggested) and then e-mails get delivered
> again (beside this one which stuck with MercuryC 1-st instance). As I
> have Mercury/32 installed also (on WXP box with ASSP on front) I
> wonder is may be reasonable to load on this Mercury/32 box Novell
> Client and log into NW server and add to M/32 additional queue pointed
> to NW server Mercury smtpqueue directory? Is that OK? M/32 is probably
> newer and could handle messages in more ... efficient way!? Make this
> sense?

Can't hurt but the problem is that Mercury/32 MercuryC must be handling the mail exclusively and when using MercuryC the NLM version is more efficient.  There is no way that a single instance MercuryC of Mercury/32 is going to be more efficient than 5 instances of MercuryC.NLM handling the queue in a round robin fachion.  This is especially true if you are using multiple relay hosts but even if using a single relay host you will be using 5 paths instead of just one.  

RTFM [:D]

Have a look in the help file under "filters" and "policies".

How to employ these depends on the detail of what you are trying to achieve.

eGroupware is not a mail client.

Nope. Egroupware is a mail client! Maybe in some older versions it wasnt one but now it is (but its not so comfortable as an real mail-client is).

A 'smarthost' is an SMTP server that handles ALL outgoing mail for the eGroupware application, "there can be only one..."

Also Nope. I can use as many SMTP-servers (or smarthosts or however you like to call it) as I want to. The problem is the authentification. As i wrote some posts ago: When I configure mercury to relay only mails for my domain than I can send mails from egroupware regardless of which account I use (thats what I want) and all is working well. But I cannot use SMTP AUTH. My intention was to configure mercury to only relay for its own domain and to use SMTP AUTH, but if I enable SMTP AUTH all other opitons are greyed out and mercury relays all mails that are authenticated (equal for which domain).

 

Greetings thomas, that conf change corrected the problem.

 

 

[quote user="Thomas R. Stephenson"]This has been passed to David but the fix will not get into v4.72 since it's in the final release cycle.[/quote]

 

Gismo22, please read Thomas's answer above - the fix is NOT in the latest release. There wasn't enough time. I'm hoping for the NEXT release of Mercury. Thanks.

Is there a way to make the mercury imap server expunge itself say every hour? On another event? 

Thanks.

Thanks for the quick reply.

The VERP address actually decodes to binary, and not a correct email

address. Is there a log somewhere where I can find out what the real

original VERP address was? Does Mercury figure these things out on its

own?

The original message has these data in the  Errors-to: line and if you are archiving all of your mail this line will be the original RCPT TO: address.  In addition, the MercuryE/C log will also show this address as originally sent.  Should be able to do a case insensitive string search to match the lower case string to a U/L case string.

Does Mercury figure these things out on its

own?

I do not know but if you are blocking the bad addresses you'll know soon enough if it does.  ;-)

 

 

Is there any way to safely scan outbound mail via Symantec AV? Is this

the type of problem that SAV would cause? Or is this potentially

another problem?

If you have to scan outbound mail that means there is some sort of internal virus is bouncing around your LAN and you have a lot more problems than just email.  ;-)

That said, you might want to checkout Virprot from Martin Ireland http://community.pmail.com/files/folders/mercadd/entry803.aspx  It uses your own anti-virus software to scan the mail under the control of Mercury/32.  There is also ClamWall comes with Mercury/32 that uses ClamAV to do the mail scanning as a daemon under the control of Mercury/32.  Both of these are safe since they do not interfere with Mercury's access to the mail files and folders.

[quote user="blallen58"]

here is a copy of the log file:

T 20090317 114334 840 Begin processing job MO000004 from
E 20090317 114334 840 530 5.7.0 Must issue a STARTTLS command first. 9sm2166941yxs.6
[/quote]This suggests that you do not have the security settings correct[quote]

T 20090317 142642 6ac Begin processing job MO001E26 from postmaster@localhost
E 20090317 142647 6ac 550 5.7.1 Unable to relay[/quote]This will be a problem. Email addresses with a domain part of "localhost" will get rejected. You need to have a proper domain part.

[quote]T 20090317 142647 6ac Begin processing job MO000002 from postmaster@ecbhlme.org
E 20090317 142657 6ac 550 5.7.1 Unable to relay[/quote]As well as authenticating correctly, you will need to set up your Gmail account to relay mail for the given addresses.[quote]
T 20090317 160734 6ac Established ESMTP connection to 192.168.10.15
T 20090317 160734 6ac Begin processing job MO001E2D from postmaster@localhost
T 20090317 160734 6ac MAIL FROM:<postmaster@localhost> SIZE=370
T 20090317 160734 6ac 250 2.1.0 Sender OK
T 20090317 160734 6ac RCPT TO:<ballen@ecbhlme.org>
T 20090317 160734 6ac 250 2.1.5 Recipient OK
T 20090317 160734 6ac Job MO001E2D from postmaster@localhost processed OK.
T 20090317 160807 6ac Established ESMTP connection to 192.168.10.15
T 20090317 160807 6ac Begin processing job MO001E2F from postmaster@localhost
T 20090317 160807 6ac MAIL FROM:<postmaster@localhost> SIZE=363
T 20090317 160807 6ac 250 2.1.0 Sender OK
T 20090317 160807 6ac RCPT TO:<ballen@ec.rr.com>
E 20090317 160812 6ac 550 5.7.1 Unable to relay

[/quote]

If you go into the MercuryC config menu & enable session logging you will get a transcript of the actual exchange of messages between the client and server, which is usually a lot more helpful than the General Log. (Also remember to turn it off after troubleshooting as it eats disk space fairly rapidly.)

Good to hear! We have very well behaved users that won't attempt to paste a zillion copies of the same address into the To: field, but I've been running the modified daemon in production here anyway for the past days. Looks like it's quite stable.

It was mentioned in the Beta group discussion that Outlook only delivers a single copy of a message in a similar situation, so perhaps this should be considered for Pegasus as well. 

 /Rolf 

Trace the message path. Meaning, if you use MercuryD to pop from the isp, do enable session log to see that you actually get the message down to Mercury. Pause all other modules etc, and examine the queues, so you see what actually happens in each step.

Don't forget to turn off session logging once you're done, as session logs eats performance and disk rapidly...

Thank you, Peter.  Yes, a single router would be the most simple solution, which I intend to use when I can find the time.  I now realize that I don't need connection control in Mercury.

Gordon.

 

How can I accomplish the reverse of this? How can I import messages

into the Mercury Mail server format? I have many messages that need to

be added to the new mercury mail server mailbox.

Depends on the files.  If these are simply *.eml files then just drop them in the new mail directory as *.CNM files and you'll be able to read them with any IMAP4 e-mail client or PMail directly.  If they are not, there are a number of translators, tell us what they are.

I suspect the problem is that he's entered the E_Mail From address with the personal name in the E-mail address field that is supposed to be a simple SMTP address.  See