I need some help (postfix virtual on Mandriva 2008)

[taylor-made]

I've followed both the perfect server guide and the "Virtual Users And Domains With Postfix, Courier And MySQL (Mandriva 2008.0)" step-by-step, which I chose to use because they're the most well-written and practical guides I could find (I didn't really need to use the perfect server guide, but it helped me sort out msec), but I just can't seem to make it work. I'm running into two problems that I'm hoping you'll be willing to help me work out. I've read through about a dozen pages of this forum, and they seem to be pretty common problems, but the posted solutions just didn't help me.

The first, and most crippling, problem is getting the server to authenticate me. It just won't, and I can't figure out why. Here's the log snippit:

Feb 27 14:29:04 mail postfix/smtpd[7530]: warning: SASL authentication failure: could not verify password
Feb 27 14:29:04 mail postfix/smtpd[7530]: warning: SASL authentication failure: Password verification failed
Feb 27 14:29:04 mail postfix/smtpd[7530]: warning: 144-12-178-69.gci.net[69.178.12.144]: SASL PLAIN authentication failed: generic failure
Feb 27 14:29:04 mail postfix/smtpd[7530]: warning: SASL authentication failure: could not verify password
Feb 27 14:29:04 mail postfix/smtpd[7530]: warning: 144-12-178-69.gci.net[69.178.12.144]: SASL LOGIN authentication failed: generic failure

SMTP is definitely listening, and I get asked for a password, but it won't auth. Same goes for POP3:

Feb 27 14:32:06 mail pop3d: authentication error: Input/output error
Feb 27 14:32:45 mail pop3d: authentication error: Input/output error
Feb 27 14:43:30 mail pop3d: authentication error: Input/output error

How do I find out what's causing this? I'm beginning to suspect that this might be a mysql access thing, but I don't know how to find out for sure. Can anyone tell me how to proceed from here?

My other problem was the tls certs, but that seems to have resolved itself (finally). I was trying to access the server as mail.taylor-madeak.org, while the host name was server.taylor-madeak.org. I ended up just changing the host name to mail.taylor-madeak.org, deleting the old certs, and creating new ones as per the guide, and restarting the server (literal reboot). But...it seemed to take a while to actually start offering the new cert. It's fixed now, but I do have a couple questions:

In the future, how do I get that to update sooner?

Is there any way I can have the server's hostname set to something different than the DNS record I access it with and not run into that error from my e-mail client (Thunderbird doesn't like mismatched server names on its certs)? I use the server for more than just e-mail, so it doesn't make any sense to lock the hostname down on mail.localdomain. It doesn't seem to hurt anything, either, but if there's a way to work around it I'd like to know it.

Thanks in advance!

[taylor-made]

Well, I'm pretty sure it's not a MySQL connection problem, or at least not MySQL itself. This is what MySQL says:

mysql> SHOW PROCESSLIST;
+-----+------------+-----------------+------+---------+------+-------+------------------+
| Id | User | Host | db | Command | Time | State | Info |
+-----+------------+-----------------+------+---------+------+-------+------------------+
| 711 | mail_admin | localhost:59626 | mail | Sleep | 3496 | | NULL |
| 844 | mail_admin | localhost:39866 | mail | Sleep | 1312 | | NULL |
| 966 | root | localhost | NULL | Query | 0 | NULL | SHOW PROCESSLIST |
+-----+------------+-----------------+------+---------+------+-------+------------------+
3 rows in set (0.00 sec)
So, at least part of Postfix is connected to it, it's just not doing anything because...well, I don't know why. I really need some help here, I'm feeling like this is probably something simple that I screwed up somewhere, but I've been back over every configuration file edited in the HOWTO and just can't find anything wrong. Even a link to a troubleshooting FAQ would be helpful.

[falko]

Are you using the correct usernames in your email client? Must be an email address.

[taylor-made]

Yep. Here, I'll give you a telnet snippet:

[bobby@mail ~]$ telnet localhost pop3
Trying 127.0.0.1...
Connected to mail.taylor-madeak.org (127.0.0.1).
Escape character is '^]'.
+OK Hello there.
user repairman@taylor-madeak.org
+OK Password required.
pass ********
-ERR Temporary problem, please try again later
Connection closed by foreign host.

I'm migrating over from a qmail/vpopmail system, so I know about using the entire e-mail address as the username. I'm just not familiar enough with postfix yet to be able to troubleshoot this on my own, so I'm asking for help. Anything you can offer is greatly appreciated.

[taylor-made]

I fixed it! Here's where I screwed up:

MYSQL_SERVER localhost
MYSQL_USERNAME mail_admin
MYSQL_PASSWORD mail_admin_password
MYSQL_PORT 0
MYSQL_DATABASE mail
MYSQL_USER_TABLE users
MYSQL_CRYPT_PWFIELD password
#MYSQL_CLEAR_PWFIELD password
MYSQL_UID_FIELD 5000
MYSQL_GID_FIELD 5000
MYSQL_LOGIN_FIELD email
MYSQL_HOME_FIELD "/home/vmail"
MYSQL_MAILDIR_FIELD CONCAT(SUBSTRING_INDEX
(email,'@',-1),'/',SUBSTRING_INDEX(email,'@',1),'/')
#MYSQL_NAME_FIELD
MYSQL_QUOTA_FIELD quota

See what happened there? When I pasted the code in, pico wordwrapped the bold line on me. Once I had a look at my folders and logs and realized that the mail server itself was working properly (had new e-mail in my folders), I sat back and thought about how this postfix installation works. It took me about 5 minutes to see how all the pieces fit together and what order the process happens in - then I started looking at the SASL and courier files and found that line. I knew it was a simple screwup that would be easy to correct once I tracked it down! Once I corrected it, telnet showed me:

[root@mail ~]# telnet localhost pop3
Trying 127.0.0.1...
Connected to mail.taylor-madeak.org (127.0.0.1).
Escape character is '^]'.
+OK Hello there.
user repairman@taylor-madeak.org
+OK Password required.
pass ********
+OK logged in.
quit
+OK Bye-bye.
Connection closed by foreign host.

So I'm getting it to authenticate me now. One problem down, another one crops up:

tp[22374]: connect to 127.0.0.1[127.0.0.1]: Connection refused (port 10024)
Feb 28 16:22:42 mail postfix/smtp[22496]: connect to 127.0.0.1[127.0.0.1]: Connection refused (port 10024)
Feb 28 16:22:42 mail postfix/smtp[22374]: 23C5B760AF: to=<helgathehorrible@shamelessolehussies.com>, relay=none, delay=44, delays=44/0.01/0/0, dsn=4.4.1, status=deferred (connect to 127.0.0.1[127.0.0.1]: Connection refused)
Feb 28 16:22:42 mail postfix/smtp[22496]: BAC04760AA: to=<yansimak@gmail.com>, relay=none, delay=679, delays=679/0.01/0/0, dsn=4.4.1, status=deferred (connect to 127.0.0.1[127.0.0.1]: Connection refused)
Feb 28 16:22:42 mail postfix/smtp[22374]: 23C5B760AF: to=<lori@shamelessolehussies.com>, relay=none, delay=44, delays=44/0.01/0/0, dsn=4.4.1, status=deferred (connect to 127.0.0.1[127.0.0.1]: Connection refused)
Feb 28 16:22:42 mail postfix/smtp[22496]: connect to 127.0.0.1[127.0.0.1]: Connection refused (port 10024)
Feb 28 16:22:42 mail postfix/smtp[22496]: 63D1F760AE: to=<fwog@taylor-madeak.org>, relay=none, delay=182, delays=182/0.01/0.05/0, dsn=4.4.1, status=deferred (connect to 127.0.0.1[127.0.0.1]: Connection refused)
Hm...looks like a simple firewall problem, but on the loopback? Well, easy to test: /etc/init.d/shorewall clear, then flush the mail queue. Oh look! The mail gets delivered out of the queue! May I suggest you add a note to the HOWTO about what ports need to be open for this to work? It's pretty straightforward to open TCP ports 10024 and 10024 using drakconf or even just editing /etc/shorewall/rules.drakx to add them.