451 Error: queue file write error (only when using postfix)

[ChristopherY]

Yes, I can run the autoinstaller, but QMail is not an included component for the Small Business Panel. This means that customers using the Small Business Panel simply have to 'deal with it' until the fix is released.

Chris

 

[IgorG]

The problem report has been frozen that I still haven't received any answer to my post

 

[ChristopherY]

I have several test systems setup, but they all come with QMail by default. You would need to switch the mail server to postfix, but it does have a domain associated with it for you to create e-mail accounts and test.

Would this be beneficial to you?

Chris

 

[BernhardE]

@IgorG: Since today I also get these error, and I have a Plesk 10.1 environment with Debian Lenny. I can provide access to it, but it is a live server.

 

[Thomas Gross]

Is there anything happening or is Parallels not able to fix this postfix queue problem?

I know that if the handlers (greylisting, spamfiltering) are taken out by changing the postfix master.cf to e.g.

smtp inet n - - - - smtpd -o smtpd_proxy_filter=127.0.0.1:10026

(instead of 10025)

the errors disappear. This is not acceptable though for us as a big hosting company.

Please find a solution with your queue handlers asap.

Regards,
Thomas

 

[IgorG]

Is there anything happening or is Parallels not able to fix this postfix queue problem?

I know that if the handlers (greylisting, spamfiltering) are taken out by changing the postfix master.cf to e.g.

smtp inet n - - - - smtpd -o smtpd_proxy_filter=127.0.0.1:10026

(instead of 10025)

the errors disappear. This is not acceptable though for us as a big hosting company.

Please find a solution with your queue handlers asap.

Regards,
Thomas

Did you read it?
http://kb.odin.com/en/9357

 

[BernhardE]

So there is still no real solution? Do you still need access to a server? Problem also occurs here, Plesk 10.1.1.

 

[Kewler]

Still happening on Parallels Small Business Panel v10.2.0_build102100709.06 os_CentOS

CPU AuthenticAMD, AMD Athlon(tm) 64 X2 Dual Core Processor 4200+
Version Parallels Small Business Panel v10.2.0_build102100709.06 os_CentOS 5
OS Linux 2.6.18-194.26.1.el5

Has anyone wondered why the /var/log/maillog.x files are all empty?

This error:

In: DATA
Out: 354 End data with <CR><LF>.<CR><LF>
Out: 451 4.3.0 Error: queue file write error


seems (from just my gut playing with postfix) like its failing to pipe its output.

I would expect to see something in the maillog, maillog.1 file

 

[Kewler]

/usr/local/psa/var/log/maillog is 10Mb and compressed to maillog.processed.1.gz ... daily - seems I should look at crontab, but it would be nice if the Parallels people would tell us how the logging system works.

 

[philippe henneau]

re: spoofing?

we have the same issue but it doesn't seem so bad as it seems to happen only with incoming Spam's, a protection must provoke a disconnect before the incoming has been completed uploaded to our server. A 2nd observation is the 'wave behavior': we get dozens of such messages in a minutes and then silence, must be related to 'spam floods''. The queue file error is related to the disconnect. But what does provoke the disconnect?

----

Out: 451 4.3.0 Error: queue file write error
Session aborted, reason: lost connection

 

[Kewler]

Even if Spamasassin is off it still barks.

No Spamassasin as it totally crushes the server, emails get lost - my installation is international and it can't handle the character sets. The result is a bunch of bounces, and undelivered mail. It appears to happen to large email attachments, and will cause outlook to keep trying to deliver as it disconnects. Its a really annoying solution. Not happy with this Parallels stuff, should have done a clean install, but no, I looked for a short-cut....what a headache.

 

[tech01]

Switch MTA

We ended up switching MTA to qmail.

Although we have noticed that sometimes the MTAs with plesk do not fully install after an upgrade and it is necessary to switch MTA from one to the other and back again. We had always had problems with postfix which sucks because it is a really nice MTA when it works.

Here is the KB:
http://kb.odin.com/5801

-Matt

 

[tech01]

On another note... Google Apps is still offering free accounts with up to 10 e-mail accounts including unlimited aliases. We switched and are slowly moving our customers over to take advantage of other integrated apps. Takes the load off of our server and gives them more functionality.

 

[philippe henneau]

another observation: we have the issue since we upgraded from Virtuozzo 4 to 4.6. Not only with plesk 10.x, we have as example that container running plesk 9.3 since 1 year, never had an issue with postifx. Since the upgrade to VC 4.6, that container keep posting issue with postifx.

Probably related to ipv6 support, we will restrict inet_protocols to ipv4 to see

Regards

 

[HaukeW]

We have the same problem since two days. It seems to be caused by SPam Mails...

--


In: RCPT TO: <[email protected]>
Out: 550 5.1.1 <[email protected]>: Recipient address rejected:
User unknown in virtual mailbox table
In: DATA
Out: 354 End data with <CR><LF>.<CR><LF>
Out: 451 4.3.0 Error: queue file write error

Session aborted, reason: lost connection

 

[philippe henneau]

Log

see attached log.

better start form the end: the following caught my attention:

Jul 24 17:57:58 prod5 postfix/smtpd[5448]: > 127.0.0.1:10025: .
Jul 24 17:58:01 prod5 postfix/smtpd[5601]: proxymap stream disconnect

 

[Daniel H]

since today i have the same problem - i had amavisd installed... but before i upgrade from plesk 10.2 to 10.3 i don't have this problem!!!!

(now after the 2nd time today i enable the -v option in master.cf.... but for now i only can provide this

Transcript of session follows.

Out: 220 xxxxxxxxxxx ESMTP Postfix

In: HELO 93-79-105-142.sumy.volia.net
Out: 250 xxxxxxxxxxxxx
In: MAIL FROM: <[email protected]>
Out: 250 2.1.0 Ok
In: RCPT TO: <[email protected]>
Out: 250 2.1.5 Ok
In: DATA
Out: 354 End data with <CR><LF>.<CR><LF>
Out: 451 4.3.0 Error: queue file write error

Session aborted, reason: lost connection

For other details, see the local mail logfile

and here a part from the logfile:

Jul 24 19:16:14 shiva clamd[2603]: SelfCheck: Database status OK.
Jul 24 19:16:49 shiva postfix/cleanup[12081]: warning: 2F36524001D9: read timeout on cleanup socket
Jul 24 19:16:49 shiva postfix/anvil[12074]: statistics: max connection rate 2/60s for (smtp:89.136.157.158) at Jul 24 19:08:24
Jul 24 19:16:49 shiva postfix/anvil[12074]: statistics: max connection count 1 for (smtp:92.252.128.232) at Jul 24 19:07:11
Jul 24 19:16:49 shiva postfix/anvil[12074]: statistics: max cache size 3 at Jul 24 19:07:11
Jul 24 19:16:49 shiva postfix/smtpd[12069]: warning: timeout talking to proxy 127.0.0.1:10025
Jul 24 19:16:49 shiva postfix/smtpd[12069]: proxy-reject: END-OF-MESSAGE: 451 4.3.0 Error: queue file write error; from=<[email protected]> to=<[email protected]> proto=SMTP helo=<93-79-105-142.sumy.volia.net>
Jul 24 19:16:49 shiva postfix/smtpd[12080]: timeout after DATA (990 bytes) from localhost.localdomain[127.0.0.1]
Jul 24 19:16:49 shiva postfix/smtpd[12080]: disconnect from localhost.localdomain[127.0.0.1]
Jul 24 19:16:49 shiva before-queue[12077]: Lost connection.
Jul 24 19:16:49 shiva before-remote[12079]: Lost connection.
Jul 24 19:16:49 shiva postfix/cleanup[12081]: F14C424001D9: message-id=<[email protected]>
Jul 24 19:16:49 shiva postfix/qmgr[2768]: F14C424001D9: from=<[email protected]>, size=922, nrcpt=1 (queue active)
Jul 24 19:16:49 shiva postfix/smtpd[12069]: disconnect from 93-79-105-142.sumy.volia.net[93.79.105.142]
Jul 24 19:16:50 shiva postfix/local[15145]: warning: database /etc/postfix/relay.db is older than source file /etc/postfix/relay
Jul 24 19:16:50 shiva postfix/cleanup[12081]: 0279E24001DA: message-id=<[email protected]>
Jul 24 19:16:50 shiva postfix/qmgr[2768]: 0279E24001DA: from=<[email protected]>, size=1072, nrcpt=1 (queue active)
Jul 24 19:16:50 shiva postfix/local[15145]: F14C424001D9: to=<[email protected]>, orig_to=<postmaster>, relay=local, delay=0.02, delays=0/0.02/0/0, dsn=2.0.0, status=sent (forwarded as 0279E24001DA)
Jul 24 19:16:50 shiva postfix/qmgr[2768]: F14C424001D9: removed
Jul 24 19:16:50 shiva postfix/pipe[15149]: warning: database /etc/postfix/relay.db is older than source file /etc/postfix/relay
Jul 24 19:16:50 shiva postfix-local[15150]: postfix-local: [email protected], [email protected], dirname=/var/qmail/mailnames
Jul 24 19:16:50 shiva spamd[14691]: spamd: got connection over /tmp/spamd_full.sock
Jul 24 19:16:50 shiva spamd[14691]: spamd: using default config for [email protected]: /var/qmail/mailnames/euphective.at/daniel/.spamassassin/user_prefs
Jul 24 19:16:50 shiva spamd[14691]: spamd: processing message <[email protected]> for [email protected]:110
Jul 24 19:16:50 shiva spamd[14691]: spamd: clean message (-0.0/5.0) for [email protected]:110 in 0.8 seconds, 1154 bytes.
Jul 24 19:16:50 shiva spamd[14691]: spamd: result: . 0 - NO_RELAYS scantime=0.8,size=1154,[email protected],uid=110,required_score=5.0,rhost=localhost,raddr=127.0.0.1,rport=/tmp/spamd_full.sock,mid=<[email protected]>,autolearn=ham

 

[IgorG]

How to gather detailed information and find root cause of "queue file write error" Postfix problem.

 

[Daniel H]

yes, thanks - i have found this yesterday... now i'm waiting, that problem happened again, so i can give you more details... What I can say is, that problem only happened with spam...

 

[philippe henneau]

1 thing is sure: it happened suddenly after having upgraded virtuozzo 4.0 to 4.6. We have VPS running over 1 Yr on that box without any problem. After upgrading to PVC 4.6 we start getting this error from various plesk server (9.5 and 10.3)

Will support be able to fix this if we open a ticket?

Regards