Scalix and Sendmail 8.13

[FrogSkin]

Hi,

I installed scalix on Suse Enterprise Server 9 that came with Sendmail 8.12. Being security concious I installed sendmail 8.13

Now I have a problem where the sendmail SMTP listener is fighting with the Scalix SMTP listener for port 25.

I have searched these forums and from that search have completed the following to try and fix the issue.

1. I have issues the omsendin command that rewrites the sendmail.cf file in the /etc directory
2. I have added DAEMON=no to that file
3. I have edited the line that reads O DaemonPortOptions=Port=smtp, Addr=127.0.0.1, Name=MTA
4. I have copied the above file to /etc/mail as well just to ensure something weired isnt happening.

HELP I am going nuts.

To sendmail I need to stop the Scalix smtp listener so that sendmail is happy and to recieve mail I have stop sendmail and start the scalix smtp listener. Any mail that sendmail catched while the Scalix smtp listener is stopped gets qued and keeps getting rejected by Scalix as the listener is partial aborted is I try and run them both.

HELP me obe1kanobe... your my only hope....

[ScalixSupport]

This sounds like /etc/hosts is messed up.

Can you check that your 127.0.0.1 line doesn't include a hostname.

Also, can you post the output of

Code: Select all

grep -v "^#" /etc/sysconfig/sendmail

Cheers

Dave

[ScalixSupport]

First, what you mean you "added DAEMON=no to that file"? Which file? You should have:

Code: Select all

O DaemonPortOptions=Port=smtp,Addr=127.0.0.1, Name=MTA

in the /etc/sendmail.cf file, but *not*:

Code: Select all

O DaemonPortOptions=Port=smtp,Addr=xx.xx.xx.xx, Name=MTA

where xx.xx.xx.xx is your server's IP address. You also need to restart sendmail after making that change by typing:

Code: Select all

/etc/init.d/sendmail restart

Now, stop and restart the Scalix SMTP Relay by typing:

Code: Select all

omoff -d0 smtpd
omon smtpd


Finally, test it all by typing:

Code: Select all

telnet xx.xx.xx.xx 25

and you should see a response that includes "Scalix SMTP Relay" and then type:

Code: Select all

telnet 127.0.0.1 25

and you should see a response that includes "Sendmail 8.13".

Thanks,
Rachel

[FrogSkin]

Hi,

My mailserver host name is: dogfood
The domain name is: speering.net

The /etc/hosts file looks like the following:

dogfood:~ # cat /etc/hosts
#
# hosts This file describes a number of hostname-to-address
# mappings for the TCP/IP subsystem. It is mostly
# used at boot time, when no name servers are running.
# On small systems, this file can be used instead of a
# "named" name server.
# Syntax:
#
# IP-Address Full-Qualified-Hostname Short-Hostname
#

127.0.0.1 localhost

# special IPv6 addresses
::1 localhost ipv6-localhost ipv6-loopback

fe00::0 ipv6-localnet

ff00::0 ipv6-mcastprefix
ff02::1 ipv6-allnodes
ff02::2 ipv6-allrouters
ff02::3 ipv6-allhosts
10.10.10.104 dogfood.speering.net dogfood
dogfood:~ #


I added the DAEMON=no to the /etc/sendmail.cf file

In the same file I added this line O DaemonPortOptions=Port=smtp,Addr=127.0.0.1, Name=MTA

and I commented out all other DaemonPortOptions entries.

Once I had made the changes I issue the rcsendmail restart command and still have the issue with sendmail trying to listen on port 25

Here is the output from grep -v "^#" /etc/sysconfig/sendmail

SENDMAIL_SMARTHOST=""

SENDMAIL_LOCALHOST="localhost"

SENDMAIL_ALLMASQUERADE="no"

SENDMAIL_RELAY=""

SENDMAIL_LUSER=""

SENDMAIL_ARGS="-L sendmail -Am -bd -q30m -om"

SENDMAIL_CLIENT_ARGS="-L sendmail-client -Ac -qp30m"

SENDMAIL_EXPENSIVE="no"

SENDMAIL_NOCANONIFY="no"

NULLCLIENT=""

NODNS="no"

DIALUP="no"

SENDMAIL_GENERICS_DOMAIN=""

MASQUERADE_DOMAINS=""

SMTP_AUTH_MECHANISMS=""

SMTP_AUTH_SERVER=""

STARTTLS=""

SENDMAIL_DB_FOLD="yes"

SENDMAIL_DNSRBL=""
DAEMON=no
dogfood:~ #


I appreciate the help as this is driving me nuts :)


Cheers

[ScalixSupport]

A couple of things. First, you shouldn't have daemon=no in the /etc/sendmail.cf, but I think you were actually referring to the /etc/sysconfig/sendmail file. Go ahead and remove it /etc/sysconfig/sendmail file and restart sendmail. Second, you said you edited sendmail.cf, but you didn't say which one. SLES 9 uses /etc/sendmail.cf, not /etc/mail/sendmail.cf. So, get rid of the file in /etc/mail as it will just cause confusion. Now, restart the Scalix SMTP Relay and post the output of the two telnet commands I posted previously.

Thanks,
Rachel

[FrogSkin]

Hi

I have found the issue. I manually installed sendmail 8.13 and thertefore it uses the /etc/mail/sendmail.cf file.

All the changes I made had been made to the /etc/sendmail.cf file. As soon as I cp this over to /etc/mail/ the sendmail daemon is now only listening to the 127.0.0.1 loopback adress and all is fine.

I now have another small issue in that I cant send mail externally as I get the following error:

Message could not be delivered to the following recipient:

frogskin.net@gmail.com

because: A mail transport failure occurred


Supplementary Information:


WARNING: RunAsUser for MSP ignored, check group ids (egid=1001, want=0)

The informail ion the mail log is:

Mar 28 12:48:41 dogfood sendmail[6719]: NOQUEUE: SYSERR(root): can not write to queue directory /var/spool/clientmqueue/ (RunAsGid=0, required=1001): Permission denied

Thanks for your help so far, its is greatly appreciated

Cheers

[ScalixSupport]

Hi. Okay, well that makes sense. The error you're receiving seems fairly self-explanatory in that the ID running sendmail doesn't have permissions to write to the /var/spool/clientmqueue subdirectory. Who are you running sendmail as? What are the permissions and who is the owner of /var/spool/clientmqueue?

Thanks,
Rachel

[FrogSkin]

Thanks for your help, the inital problem was solved and I greatly appreciate it.

The permission error Im getting is bizzare and has to do with sendmail and how it uses different permissions for different tasks.

I wont bother you with it as its not a Scalix problem.

Cheers

[kwestionit]

I'm having the same problem on the clientmqueue permissions if you solved the problem, please post it.

Regards
Kwestion