New E-mail/Address creation causes disruption to Postfix

[Nuxwin]

I'll check that. I'm eat now

---

Edit:

I've checked and it's ok :

[joximu]

Sorry - was not here...

+1 :-)

/J

[edeis]

Thanks Guys. Glad I could help find this bug.

Is the patch safe to implement on a production system?
http://isp-control.net/ispcp/changeset/3196

Will it stop the following errors?
Aug 11 10:44:58 drizzle postfix/smtpd[15981]: warning: premature end-of-input on private/rewrite socket while reading input attribute name
Aug 11 10:44:58 drizzle postfix/smtpd[15981]: warning: problem talking to service rewrite: Connection reset by peer

or will they still come up because of postfix reload the hash table?
Aug 11 10:44:58 drizzle postfix/trivial-rewrite[15919]: table hash:/etc/postfix/ispcp/transport(0,lock|no_regsub|fold_fix) has changed -- restarting



ps. How does one ever get reputation points on this forum?

[joximu]

Hi edeis

you can just comment the line with "sys_command_rs("$cmd stop");" (you see it in red, in the patch)... then you have the same result.

Since postfix will then reload the hash tables by itself there won't be errors because of reloading... (or worse: restarting)


This was a smaller bug - no mails were lost because of this - only some entries in the logs. So it's more an improvement now :-)
There are lot's of places where such improvements are welcome :-)

To the reputation points... if your comments here are good then you will get some good points...
Just keep on helping here!

/J

[edeis]

Cool. Thanks Joximu.

---

---

If we are removing the command to stop postfix, should we also remove the command to start it too?

Code:

if ($main::changed_mail_cnt > 0 || $main::changed_dmn_cnt > 0 ||
                $main::changed_sub_cnt > 0 || $main::changed_als_cnt > 0) {

                if ($main::cfg{'CMD_MTA'} ne 'no') {

                        $cmd = $main::cfg{'CMD_MTA'};

                       sys_command_rs("$cmd start");

                }

I also noticed a descripancy when checking to stop the services when new domains, etc are added and when they are started again. They are started when mail_cnt, dmn_cnt, sub_cnt, and als_cnt are changed. But stopped only when mail_cnt, dmn_cnt, sub_cnt are changed.

[Nuxwin]

(08-14-2010 05:27 AM)edeis Wrote:  Cool. Thanks Joximu.

---

---

If we are removing the command to stop postfix, should we also remove the command to start it too?

hello ;

No the start statement will do nothing if postfix is already running.

[edeis]

Had some errors after a client added an autoresponder. Wondering if this is the same problem?

ispcp admin log:
add mail autoresponder: domainname.com

Oct 18 10:00:11 drizzle postfix/smtpd[5287]: table hash:/etc/postfix/ispcp/mailboxes(0,lock|fold_fix) has changed -- restarting
Oct 18 10:00:11 drizzle postfix/smtpd[1684]: table hash:/etc/postfix/ispcp/mailboxes(0,lock|fold_fix) has changed -- restarting
Oct 18 10:00:12 drizzle postfix/smtpd[7617]: connect from unknown[116.37.119.27]
Oct 18 10:00:12 drizzle dovecot: dovecot: Killed with signal 15 (by pid=7625 uid=0 code=kill)
Oct 18 10:00:12 drizzle dovecot: pop3-login: Disconnected: Shutting down (disconnected while authenticating): method=PLAIN, rip=96.50.231.238, lip=192.168.1.121
Oct 18 10:00:12 drizzle postfix/smtpd[7641]: warning: SASL: Connect to private/auth failed: No such file or directory
Oct 18 10:00:12 drizzle postfix/smtpd[7641]: fatal: no SASL authentication mechanisms
Oct 18 10:00:12 drizzle postfix/postfix-script: fatal: the Postfix mail system is already running
Oct 18 10:00:12 drizzle dovecot: Dovecot v1.2.11 starting up (core dumps disabled)
Oct 18 10:00:12 drizzle dovecot: auth-worker(default): mysql: Connected to localhost (ispcp)
Oct 18 10:00:12 drizzle last message repeated 2 times
Oct 18 10:00:12 drizzle postfix/trivial-rewrite[18201]: table hash:/etc/postfix/ispcp/transport(0,lock|no_regsub|fold_fix) has changed -- restarting
Oct 18 10:00:13 drizzle postfix/smtpd[7617]: lost connection after DATA from unknown[116.37.119.27]
Oct 18 10:00:13 drizzle postfix/smtpd[7617]: disconnect from unknown[116.37.119.27]
Oct 18 10:00:13 drizzle postfix/master[25691]: warning: process /usr/libexec/postfix/smtpd pid 7641 exit status 1
Oct 18 10:00:13 drizzle postfix/master[25691]: warning: /usr/libexec/postfix/smtpd: bad command startup -- throttling

---

More info:

Turns out that the autoresponder has broke.

User has reported that e-mails to the address that has the auto-responder enable returns a Undelivered Mail return to sender:
I'm sorry to have to inform you that your message could not
be delivered to one or more recipients. It's attached below.

For further assistance, please send mail to <postmaster>

If you do so, please include this problem report. You can
delete your own text from the attached returned message.

The mail system

<USER@ispcp-arpl.DOMAINNAME.com> (expanded from
<USER@DOMAINNAME.com>): Command died with status 2:
"/var/www/ispcp/engine/messenger/ispcp-arpl-msgr". Command output: Can't
redirect stdout at /var/www/ispcp/engine/messenger/ispcp-arpl-msgr line
530.

Reporting-MTA: dns; myserver.com
X-Postfix-Queue-ID: 4C15F5550A50
X-Postfix-Sender: rfc822; sendersemail@domain.com
Arrival-Date: Mon, 18 Oct 2010 11:20:15 -0700 (PDT)

Final-Recipient: rfc822; USER@ispcp-arpl.DOMAIN.com
Original-Recipient: rfc822;USER@DOMAIN.com
Action: failed
Status: 5.3.0
Diagnostic-Code: x-unix; Can't redirect stdout at
/var/www/ispcp/engine/messenger/ispcp-arpl-msgr line 530.

[edeis]

Nevermind. Found the solution.

Reference, similiar thread:
http://isp-control.net/forum/thread-9222...l#pid85007

Steps to correct:
# set-engine-permissions.sh
Setting Engine Permissions: ..find: /var/log/ispcp/ispcp-arpl-msgr: No such file or directory
find: /var/log/ispcp/ispcp-arpl-msgr: No such file or directory
find: /var/log/ispcp/ispcp-arpl-msgr: No such file or directory
done

# mkdir /var/log/ispcp/ispcp-arpl-msgr

# set-engine-permissions.sh
Setting Engine Permissions: .. done

Autoresponder now works!.