Postfix only works for a few minutes. that is Mail is recieved for a short time and then Postfix stops recieving. restarting Postfix just does the same thing. This is a SLES 9 installation running the latest Postfix, amavisd and Spamassassin. I have no clue where to start looking at this problem. During this time qpopper still works. Thanks
Fri, 10 Dec 2004, by suse-list@justsuse.com:
Postfix only works for a few minutes. that is Mail is recieved for a short time and then Postfix stops recieving. restarting Postfix just does the same thing.
This is a SLES 9 installation running the latest Postfix, amavisd and Spamassassin. I have no clue where to start looking at this problem. During this time qpopper still works.
The logs would be a good place to start. Postfix doesn't just stop for the heck of it. http://www.postfix.org/DEBUG_README.html #Look for obvious signs of trouble #Postfix logs all failed and successful deliveries to a logfile. The #file is usually called /var/log/maillog or /var/log/mail; the exact #pathname is defined in the /etc/syslog.conf file. #When Postfix does not receive or deliver mail, the first order of #business is to look for errors that prevent Postfix from working #properly: #% egrep '(warning|error|fatal|panic):' /some/log/file | more #Note: the most important message is near the BEGINNING of the #output. Error messages that come later are less useful. Usually it's something like flaky hardware, full disk(s) or a borked *.cf file. Qpopper of course has nothing to do with this. Theo -- Theo v. Werkhoven Registered Linux user# 99872 http://counter.li.org ICBM 52 13 27N , 4 29 45E. + ICQ: 277217131 SUSE 9.2 + Jabber: gurp@nedlinux.nl Kernel 2.6.8 + MSN: twe-msn@ferrets4me.xs4all.nl See headers for PGP/GPG info. +
thanks for the help, I myself think it is a flacky main.cf but I can not find where the problem is, Is sysconfig good to use? #% egrep '(warning|error|fatal|panic):' /some/log/file | more A small sample of the log file: Dec 10 15:12:20 mail amavis[7557]: (07557-07) warning - MIME::Parser error: unexpected end of preamble Dec 10 15:17:07 mail postfix/smtpd[7568]: warning: smtpd_peer_init: 147.202.38.253: address not listed for hostname vertexhost.com Dec 10 15:17:58 mail postfix/smtpd[7818]: warning: smtpd_peer_init: 69.6.66.235: address not listed for hostname m7.bigposition.com Dec 10 15:20:51 mail postfix/smtpd[7818]: warning: smtpd_peer_init: 65.17.116.9: address not listed for hostname archie.mmcity.net Dec 10 15:21:08 mail postfix/smtpd[7818]: warning: smtpd_peer_init: 66.63.165.62: hostname mx9.numberrest.com verification failed: Name or service not known Dec 10 15:22:12 mail postfix/smtpd[7568]: warning: smtpd_peer_init: 209.172.104.131: hostname mail.regpage.com verification failed: Name or service not known Dec 10 15:22:13 mail postfix/smtpd[7818]: warning: smtpd_peer_init: 66.63.174.184: hostname mx1.reversebreathless.com verification failed: Name or service not known Dec 10 15:22:48 mail postfix/smtpd[8048]: warning: smtpd_peer_init: 69.6.66.235: address not listed for hostname m7.bigposition.com Dec 10 15:27:14 mail postfix/smtpd[7568]: warning: smtpd_peer_init: 202.180.87.216: hostname p216.pool-87.cpiak1-l1.callplus.net.nz verification failed: Name or service not known Dec 10 15:28:55 mail postfix/smtpd[8048]: warning: smtpd_peer_init: 66.63.174.205: hostname mx8.mercyamount.com verification failed: Name or service not known Dec 10 15:28:55 mail postfix/smtpd[8048]: warning: smtpd_peer_init: 66.63.174.200: hostname mx3.mercyamount.com verification failed: Name or service not known Dec 10 15:28:55 mail postfix/smtpd[8048]: warning: smtpd_peer_init: 66.63.174.205: hostname mx8.mercyamount.com verification failed: Name or service not known Dec 10 15:28:58 mail postfix/smtpd[8048]: warning: smtpd_peer_init: 200.12.44.90: hostname ip-44-90.guate.net.gt verification failed: Name or service not known Dec 10 15:29:01 mail postfix/smtpd[8048]: warning: smtpd_peer_init: 66.63.174.199: hostname mx2.mercyamount.com verification failed: Name or service not known Dec 10 15:29:12 mail postfix/smtpd[8048]: warning: smtpd_peer_init: 69.6.66.235: address not listed for hostname m7.bigposition.com Dec 10 15:32:55 mail postfix/smtpd[8449]: warning: smtpd_peer_init: 67.131.19.226: hostname 67-131-19-226.dia.cust.qwest.net verification failed: Name or service not known Dec 10 15:34:19 mail postfix/smtpd[8530]: warning: smtpd_peer_init: 213.37.24.208: hostname 208.red-213-37-24.user.auna.net verification failed: Name or service not known Dec 10 15:35:49 mail postfix/smtpd[8449]: warning: smtpd_peer_init: 208.210.222.116: hostname net.fla.lrp.com verification failed: Name or service not known Dec 10 15:39:23 mail postfix/smtpd[8530]: warning: smtpd_peer_init: 66.63.165.49: hostname mx8.indoorstv.com verification failed: Name or service not known Dec 10 15:39:23 mail postfix/smtpd[8530]: warning: smtpd_peer_init: 164.42.133.94: hostname host133-94.inter.edu verification failed: Name or service not known At 04:40 PM 12/10/2004, Theo v. Werkhoven wrote:
Fri, 10 Dec 2004, by suse-list@justsuse.com:
Postfix only works for a few minutes. that is Mail is recieved for a short time and then Postfix stops recieving. restarting Postfix just does the same thing.
This is a SLES 9 installation running the latest Postfix, amavisd and Spamassassin. I have no clue where to start looking at this problem. During this time qpopper still works.
The logs would be a good place to start. Postfix doesn't just stop for the heck of it. http://www.postfix.org/DEBUG_README.html #Look for obvious signs of trouble
#Postfix logs all failed and successful deliveries to a logfile. The #file is usually called /var/log/maillog or /var/log/mail; the exact #pathname is defined in the /etc/syslog.conf file.
#When Postfix does not receive or deliver mail, the first order of #business is to look for errors that prevent Postfix from working #properly:
#% egrep '(warning|error|fatal|panic):' /some/log/file | more
#Note: the most important message is near the BEGINNING of the #output. Error messages that come later are less useful.
Usually it's something like flaky hardware, full disk(s) or a borked *.cf file. Qpopper of course has nothing to do with this.
Theo -- Theo v. Werkhoven Registered Linux user# 99872 http://counter.li.org ICBM 52 13 27N , 4 29 45E. + ICQ: 277217131 SUSE 9.2 + Jabber: gurp@nedlinux.nl Kernel 2.6.8 + MSN: twe-msn@ferrets4me.xs4all.nl See headers for PGP/GPG info. +
-- Check the headers for your unsubscription address For additional commands send e-mail to suse-linux-e-help@suse.com Also check the archives at http://lists.suse.com Please read the FAQs: suse-linux-e-faq@suse.com
Fri, 10 Dec 2004, by suse-list@justsuse.com:
thanks for the help, I myself think it is a flacky main.cf but I can not find where the problem is, Is sysconfig good to use?
Good to use for what? As a relaxing piece of reading before sleep? Not really. As a tool to keep all config parameters in one place? I'd say yes. Show postconf -n
#% egrep '(warning|error|fatal|panic):' /some/log/file | more
A small sample of the log file: [..] Only resolver warnings here. Where's the 'error|fatal|panic' part? Like I said: Postfix doesn't 'just die'
Please trim the quotes and stop topposting. The least you can do when you ask for help is try to be as polite as possible.. Theo -- Theo v. Werkhoven Registered Linux user# 99872 http://counter.li.org ICBM 52 13 27N , 4 29 45E. + ICQ: 277217131 SUSE 9.2 + Jabber: gurp@nedlinux.nl Kernel 2.6.8 + MSN: twe-msn@ferrets4me.xs4all.nl See headers for PGP/GPG info. +
postconf -n, yes a very usfull tool. alias_maps = hash:/etc/aliases canonical_maps = hash:/etc/postfix/canonical command_directory = /usr/sbin config_directory = /etc/postfix content_filter = vscan: daemon_directory = /usr/lib/postfix defer_transports = disable_dns_lookups = no html_directory = /usr/share/doc/packages/postfix/html inet_interfaces = all mail_owner = postfix mail_spool_directory = /var/mail mailbox_command = mailbox_size_limit = 0 mailbox_transport = mailq_path = /usr/bin/mailq manpage_directory = /usr/share/man masquerade_classes = envelope_sender, header_sender, header_recipient masquerade_domains = masquerade_exceptions = root message_size_limit = 10240000 mydestination = fresno.edu, localhost, mail.fresno.edu mydomain = fresno.edu myhostname = mail.fresno.edu mynetworks = 127.0.0.0/8,/etc/postfix/mynetworks mynetworks_style = subnet myorigin = fresno.edu newaliases_path = /usr/bin/newaliases queue_directory = /var/spool/postfix readme_directory = /usr/share/doc/packages/postfix/README_FILES relay_domains = /etc/postfix/relay_domains relayhost = relocated_maps = hash:/etc/postfix/relocated sample_directory = /usr/share/doc/packages/postfix/samples sender_canonical_maps = hash:/etc/postfix/sender_canonical sendmail_path = /usr/sbin/sendmail setgid_group = maildrop smtp_sasl_auth_enable = no smtp_use_tls = no smtpd_client_restrictions = permit_mynetworks, reject_rbl_client cbl.abuseat.org, reject_rbl_client sbl-xbl.spamhaus.org, rej ect_rbl_client relays.ordb.org smtpd_helo_required = yes smtpd_helo_restrictions = smtpd_recipient_restrictions = permit_mynetworks, reject_unauth_destination smtpd_sasl_auth_enable = no smtpd_sender_restrictions = permit_mynetworks, check_sender_access hash:/etc/postfix/access_sender smtpd_use_tls = no strict_rfc821_envelopes = no transport_maps = hash:/etc/postfix/transport unknown_local_recipient_reject_code = 550 My setup I use the default SuSE setup with amavisd, clamd, spamassassin. No changes to either of those except to discard spam at 5 hits. Thanks again. At 06:14 PM 12/10/2004, Theo v. Werkhoven wrote:
Fri, 10 Dec 2004, by suse-list@justsuse.com:
thanks for the help, I myself think it is a flacky main.cf but I can not find where the problem is, Is sysconfig good to use?
Good to use for what? As a relaxing piece of reading before sleep? Not really. As a tool to keep all config parameters in one place? I'd say yes.
Show postconf -n
#% egrep '(warning|error|fatal|panic):' /some/log/file | more
A small sample of the log file: [..] Only resolver warnings here. Where's the 'error|fatal|panic' part? Like I said: Postfix doesn't 'just die'
Please trim the quotes and stop topposting. The least you can do when you ask for help is try to be as polite as possible..
Theo -- Theo v. Werkhoven Registered Linux user# 99872 http://counter.li.org ICBM 52 13 27N , 4 29 45E. + ICQ: 277217131 SUSE 9.2 + Jabber: gurp@nedlinux.nl Kernel 2.6.8 + MSN: twe-msn@ferrets4me.xs4all.nl See headers for PGP/GPG info. +
-- Check the headers for your unsubscription address For additional commands send e-mail to suse-linux-e-help@suse.com Also check the archives at http://lists.suse.com Please read the FAQs: suse-linux-e-faq@suse.com
Fri, 10 Dec 2004, by suse-list@justsuse.com:
postconf -n, yes a very usfull tool.
It is, but nothing in here suggests why Postfix would stop. Like I've asked you previously: stop topposting and trim the quotes. I'm not going to do that for you, so this thread stops until you read, understand and comply to http://www.rfc.net/rfc1855.html Theo -- Theo v. Werkhoven Registered Linux user# 99872 http://counter.li.org ICBM 52 13 27N , 4 29 45E. + ICQ: 277217131 SUSE 9.2 + Jabber: gurp@nedlinux.nl Kernel 2.6.8 + MSN: twe-msn@ferrets4me.xs4all.nl See headers for PGP/GPG info. +
The Saturday 2004-12-11 at 01:40 +0100, Theo v. Werkhoven wrote:
The logs would be a good place to start. Postfix doesn't just stop for the heck of it.
Mine does... has been doing so for months, since I updated to suse 9.1.
Dec 10 12:43:15 nimrodel fetchmail[6645]: 65 messages for * at pop.tiscali.es (339598 octets).
Dec 10 12:43:15 nimrodel fetchmail[6645]: POP3> LIST 1
Dec 10 12:43:15 nimrodel fetchmail[6645]: POP3< +OK 1 2766
Dec 10 12:43:15 nimrodel fetchmail[6645]: POP3> RETR 1
Dec 10 12:43:15 nimrodel fetchmail[6645]: POP3< +OK 2766 bytes
Dec 10 12:43:15 nimrodel fetchmail[6645]: reading message *@pop.tiscali.es:1 of 65 (2766 octets)
Dec 10 12:43:31 nimrodel ip-up.local.doit: -> FidoNet polled.
Dec 10 12:43:55 nimrodel ip-up.local.doit: -> FidoNet tossed and linked
Dec 10 12:44:25 nimrodel fetchmail[6645]: smtp listener protocol error
Dec 10 12:45:35 nimrodel fetchmail[6645]: smtp listener protocol error
Dec 10 12:45:35 nimrodel fetchmail[6645]: SMTP connect to localhost failed
Notice that fetchmail gives up after 70 seconds trying to connect to
postfix locally. Also, if I happen to try sending an email from Pine at
that exact moment, it also times out. When I notice this, I restart
postfix (I'm using a modem, so I'm normally looking). If I don't notice,
the next round of fetchmail works.
Previous to this timeout, mails seem to work. This usually happen on the
very first email I fetch from tiscali (and not from the two other servers
I fetch before tiscali) on the first fetchmail run after my first
connection of the day - but sometimes also happens later on the day.
Another one, previous day. I'll copy much more details from the logs. You
see that there is a send that works;
Dec 10 00:57:51 nimrodel fetchmail[13587]: POP3< +OK 57 messages
Dec 10 00:57:52 nimrodel postfix/smtp[13590]: connect to *.com[*.*.*.65]: server dropped connection without sending the initial SMTP greeting (port 25)
Dec 10 00:57:53 nimrodel postfix/smtp[13590]: 0E5BE20C3D: to=<*@*.*.com>, relay=gandalf.*.com[*.*.*.243], delay=12989, status=sent (250 Ok: queued as 608566787D)
And a minute before, not shown, fetchmail gets email from several other
servers, sucessfully. But a few seconds later, that same first fetchmail
run for this connection of server tiscali, fails:
Dec 10 00:57:53 nimrodel postfix/qmgr[5362]: 0E5BE20C3D: removed
Dec 10 00:57:54 nimrodel fetchmail[13587]: POP3> STAT
Dec 10 00:57:54 nimrodel fetchmail[13587]: POP3< +OK 57 177537
Dec 10 00:57:54 nimrodel fetchmail[13587]: 57 messages for * at pop.tiscali.es (177537 octets).
Dec 10 00:57:54 nimrodel fetchmail[13587]: POP3> LIST 1
Dec 10 00:57:54 nimrodel fetchmail[13587]: POP3< +OK 1 2460
Dec 10 00:57:54 nimrodel fetchmail[13587]: POP3> RETR 1
Dec 10 00:57:55 nimrodel fetchmail[13587]: POP3< +OK 2460 bytes
Dec 10 00:57:55 nimrodel fetchmail[13587]: reading message *@pop.tiscali.es:1 of 57 (2460 octets)
Dec 10 00:58:30 nimrodel ip-up.local.doit: -> FidoNet polled.
Dec 10 00:58:39 nimrodel ip-up.local.doit: -> FidoNet tossed and linked
Dec 10 00:59:05 nimrodel fetchmail[13587]: smtp listener protocol error
70 seconds timeout of fetchmail (70 seconds is my seting, because at 85"
I have the pppd daemon drop the connection).
Dec 10 00:59:25 nimrodel amavis[4748]: (04748-04-3) Requesting a process rundown after 10 tasks
Dec 10 01:00:15 nimrodel fetchmail[13587]: smtp listener protocol error
Dec 10 01:00:15 nimrodel fetchmail[13587]: SMTP connect to localhost failed
Connection to local smtp fails, and fetchmail aborts this server.
Dec 10 01:00:15 nimrodel fetchmail[13587]: POP3> QUIT
Dec 10 01:00:15 nimrodel fetchmail[13587]: POP3< * Laurent Renard <*@*.be> [12-09-04 14:37]:
Dec 10 01:00:15 nimrodel fetchmail[13587]: SMTP transaction error while fetching from pop.tiscali.es
Fetchmail complains.
Dec 10 01:00:15 nimrodel fetchmail[13587]: 6.2.5 querying pop.tiscali.es (protocol POP3) at Fri Dec 10 01:00:15 2004: poll completed
Dec 10 01:00:15 nimrodel fetchmail[13587]: Query status=10 (SMTP)
A bit later, after polling some more servers (empty), I see:
Dec 10 01:00:25 nimrodel fetchmail[13587]: No mail for *@* at *susemania.org
...
Dec 10 01:00:26 nimrodel fetchmail[13587]: 6.2.5 querying *susemania.org (protocol IMAP) at Fri Dec 10 01:00:26 2004:
Dec 10 01:00:26 nimrodel fetchmail[13587]: SMTP> QUIT
Dec 10 01:00:26 nimrodel postfix/smtpd[13606]: disconnect from localhost[127.0.0.1]
Dec 10 01:00:26 nimrodel fetchmail[13587]: SMTP< 221 Bye
Dec 10 01:00:26 nimrodel fetchmail[13587]: SMTP> QUIT
Dec 10 01:00:26 nimrodel postfix/smtpd[13606]: connect from localhost[127.0.0.1]
Dec 10 01:00:26 nimrodel postfix/smtpd[13606]: lost connection after CONNECT from localhost[127.0.0.1]
Dec 10 01:00:26 nimrodel postfix/smtpd[13606]: disconnect from localhost[127.0.0.1]
Dec 10 01:00:26 nimrodel postfix/smtpd[13606]: connect from localhost[127.0.0.1]
Dec 10 01:00:26 nimrodel postfix/smtpd[13606]: lost connection after CONNECT from localhost[127.0.0.1]
Dec 10 01:00:26 nimrodel postfix/smtpd[13606]: disconnect from localhost[127.0.0.1]
Dec 10 01:00:26 nimrodel postfix/smtpd[13617]: disconnect from localhost[127.0.0.1]
I don't know what are these connections that postfix is dropping here. PID
13606 was used previously for the first email fetched from the first
server of this run, sucessfully, it seems. PID 13617 was used by the
second. Could these two connections be pending for about 2'30" and hog
postfix? I wonder, because around 8 emails were sucessfully downloaded
after those two, and before it stops at tiscali - it only happens at the
first email of tiscali, on the first run of the connection, and not every
connection.
Dec 10 01:00:26 nimrodel fetchmail[13587]: SMTP< 221 Bye
Dec 10 01:00:26 nimrodel fetchmail[13587]: normal termination, status 0
Dec 10 01:00:26 nimrodel ip-up.local.doit: -> Mail fetched - second poll in 30"
and now it works - without me resetting postfix this time:
Dec 10 01:01:34 nimrodel fetchmail[13695]: 58 messages for * at pop.tiscali.es (180498 octets).
Dec 10 01:01:34 nimrodel fetchmail[13695]: POP3> LIST 1
Dec 10 01:01:34 nimrodel fetchmail[13695]: POP3< +OK 1 2460
Dec 10 01:01:34 nimrodel fetchmail[13695]: POP3> RETR 1
Dec 10 01:01:34 nimrodel fetchmail[13695]: POP3< +OK 2460 bytes
Dec 10 01:01:34 nimrodel fetchmail[13695]: reading message *@pop.tiscali.es:1 of 58 (2460 octets)
Dec 10 01:01:34 nimrodel postfix/smtpd[13606]: connect from localhost[127.0.0.1]
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP< 220 nimrodel.valinor ESMTP Postfix
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP> EHLO localhost
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP< 250-nimrodel.valinor
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP< 250-PIPELINING
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP< 250-SIZE 10240000
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP< 250-VRFY
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP< 250-ETRN
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP< 250 8BITMIME
Dec 10 01:01:34 nimrodel fetchmail[13695]: SMTP> MAIL FROM:
Sat, 11 Dec 2004, by robin1.listas@tiscali.es:
The Saturday 2004-12-11 at 01:40 +0100, Theo v. Werkhoven wrote:
The logs would be a good place to start. Postfix doesn't just stop for the heck of it.
Mine does... has been doing so for months, since I updated to suse 9.1.
I don't see any evidence here that shows it's Postfix that borkes.
Dec 10 12:44:25 nimrodel fetchmail[6645]: smtp listener protocol error Dec 10 12:45:35 nimrodel fetchmail[6645]: smtp listener protocol error Dec 10 12:45:35 nimrodel fetchmail[6645]: SMTP connect to localhost failed [..]
Previous to this timeout, mails seem to work. This usually happen on the very first email I fetch from tiscali (and not from the two other servers I fetch before tiscali) on the first fetchmail run after my first connection of the day - but sometimes also happens later on the day.
See if you can make Postfix show more debug info main.cf debug_peer_list = 127.0.0.1 [..]
second. Could these two connections be pending for about 2'30" and hog postfix? I wonder, because around 8 emails were sucessfully downloaded after those two, and before it stops at tiscali - it only happens at the first email of tiscali, on the first run of the connection, and not every connection.
Only a good look in your logs with debug level up a few steps would show why Postfix and Fetchmail can't get along. Also; are mails stuck in any queue during this time-out? 'mailq' [snip more logs]
So... yes, postfix sometimes does stall for some non obvious reasons. The configuration was the same I had in suse 8.2, I think (I upgraded, not reinstalled).
I worked with SuSE 9.1 standard Postfix for quite some time without a problem, but I don't use POP or IMAP, only SMTP. You could try the Postfix rpm from Guillermo to see if it is the SuSE build or not. ftp://ftp.oxixares.com/pub/rpms/SuSE_9.1/ Theo -- Theo v. Werkhoven Registered Linux user# 99872 http://counter.li.org ICBM 52 13 27N , 4 29 45E. + ICQ: 277217131 SUSE 9.2 + Jabber: gurp@nedlinux.nl Kernel 2.6.8 + MSN: twe-msn@ferrets4me.xs4all.nl See headers for PGP/GPG info. +
The Saturday 2004-12-11 at 14:41 +0100, Theo v. Werkhoven wrote:
Mine does... has been doing so for months, since I updated to suse 9.1.
I don't see any evidence here that shows it's Postfix that borkes.
Well... fetchmail complains of no connection, and Pine as well, if I happen to try to send in that precise minute. I would say that Postfix is involved, but of course, it could be amavis, or even the network stack or firewall.
See if you can make Postfix show more debug info
main.cf debug_peer_list = 127.0.0.1
Yes, I thought of that. But that would throw tons of debug info: it doesn't happen every day, and only on the first email from that provider, an on the first run. That is, once a day at most. I'd have to be ready to change it back fast. I'll think it over... might be worth trying, for knowledge sake.
Only a good look in your logs with debug level up a few steps would show why Postfix and Fetchmail can't get along. Also; are mails stuck in any queue during this time-out? 'mailq'
I'm not sure, but I think not. I'll have to wait till the next time it happens.
[snip more logs]
So... yes, postfix sometimes does stall for some non obvious reasons. The configuration was the same I had in suse 8.2, I think (I upgraded, not reinstalled).
I worked with SuSE 9.1 standard Postfix for quite some time without a problem, but I don't use POP or IMAP, only SMTP. You could try the Postfix rpm from Guillermo to see if it is the SuSE build or not. ftp://ftp.oxixares.com/pub/rpms/SuSE_9.1/
Yes; but I prefer to wait for SuSE 9.2, which I will install before the month ends, I hope. -- Cheers, Carlos Robinson
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 [long email, contains logs excerpts] The Saturday 2004-12-11 at 14:41 +0100, Theo v. Werkhoven wrote:
See if you can make Postfix show more debug info
Ok, I did that today. But I don't see anything else clear. First, it downloaded one email from two diferent servers, without a problem. It also sent a pending email to the list. Then it started to download the first email from tiscali... and stuck. Here goes the log, trimmed a bit (** replaces addreses names, date and hostname form every line removed). Dec 12 20:13:36 nimrodel fetchmail[8596]: 6.2.5 querying pop.tiscali.es (protocol POP3) at Sun Dec 12 20:13:36 2004: poll started 13:37 fetchmail[8596]: POP3< +OK POP3 server ready (6.7.018) <3E7BA****1D16@netmail.tiscali.es> 13:37 fetchmail[8596]: POP3> CAPA 13:37 fetchmail[8596]: POP3< +OK Capability list follows 13:37 fetchmail[8596]: POP3< TOP 13:37 fetchmail[8596]: POP3< RESP-CODES 13:37 fetchmail[8596]: POP3< USER 13:37 fetchmail[8596]: POP3< SASL CRAM-MD5 DIGEST-MD5 PLAIN 13:37 fetchmail[8596]: POP3< PIPELINING 13:37 fetchmail[8596]: POP3< UIDL 13:37 fetchmail[8596]: POP3< IMPLEMENTATION InScribe-6.7.018 13:37 fetchmail[8596]: POP3< AUTH-RESP-CODE 13:37 fetchmail[8596]: POP3< . 13:37 fetchmail[8596]: POP3> AUTH CRAM-MD5 13:37 fetchmail[8596]: POP3< + PE***Pg== 13:37 fetchmail[8596]: POP3> Y2***zOQ== 13:39 fetchmail[8596]: POP3< +OK 80 messages 13:42 fetchmail[8596]: POP3> STAT 13:42 fetchmail[8596]: POP3< +OK 80 375755 13:42 fetchmail[8596]: 80 messages for ** at pop.tiscali.es (375755 octets). 13:42 fetchmail[8596]: POP3> LIST 1 13:42 fetchmail[8596]: POP3< +OK 1 3010 13:42 fetchmail[8596]: POP3> RETR 1 13:42 fetchmail[8596]: POP3< +OK 3010 bytes 13:42 fetchmail[8596]: reading message **@pop.tiscali.es:1 of 80 (3010 octets) 14:04 ip-up.local.doit: -> FidoNet polled. 14:53 fetchmail[8596]: smtp listener protocol error 15:11 ip-up.local.doit: -> FidoNet tossed and linked 15:11 postfix/lmtp[8613]: idle timeout -- exiting 15:11 postfix/lmtp[8613]: > 127.0.0.1[127.0.0.1]: QUIT 15:11 postfix/lmtp[8613]: < 127.0.0.1[127.0.0.1]: 221 2.0.0 [127.0.0.1] (amavisd) closing transmission channel 16:03 fetchmail[8596]: smtp listener protocol error 16:03 fetchmail[8596]: SMTP connect to localhost failed 16:03 fetchmail[8596]: POP3> QUIT 16:03 fetchmail[8596]: POP3< Carlos E. R. wrote: 16:03 fetchmail[8596]: SMTP transaction error while fetching from pop.tiscali.es 16:03 fetchmail[8596]: 6.2.5 querying pop.tiscali.es (protocol POP3) at Sun 16:03 2004: poll completed 16:03 fetchmail[8596]: Query status=10 (SMTP) You see that the error information is not much: fetchmail: starts download, tries to stablish connection to local smtp , 70" timeout: and fails: fetchmail: smtp listener protocol error postfix/lmtp: postfix/lmtp: postfix/lmtp: idle timeout -- exiting postfix/lmtp: > 127.0.0.1[127.0.0.1]: QUIT postfix/lmtp: < 127.0.0.1[127.0.0.1]: 221 2.0.0 [127.0.0.1] (amavisd) closing transmission channel fetchmail: smtp listener protocol error fetchmail: SMTP connect to localhost failed The only important thing, crucial IMO, is that "postfix/smtpd" reports nothing, and it should. lmtp is not active yet, till the email is downloaded and passed to amavis, so it closes for the time being. A minute later, that same email is downloaded cleanly (long log, for comparison): [I moved this to attached gziped text file - I don't know if it will pass, but in plain email it did not (probably overlarge), although suse server claimed it did as AE95611A347C. No bounce :-/ ] I'll leave the log there, it continues. But by now fetchmail is already downloading the second email, and postfix is passing the first to amavis-new.
Only a good look in your logs with debug level up a few steps would show why Postfix and Fetchmail can't get along. Also; are mails stuck in any queue during this time-out? 'mailq'
None, empty queue. I hope you can see something I don't :-) - -- Cheers, Carlos Robinson -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Made with pgp4pine 1.76 iD8DBQFBwzrFtTMYHG2NR9URAlcnAJ9TKkYOSndeMwKrPC2Bkh+G3zEO1ACfRkO3 GRfcaj9Wgajo0dIxYIiEHck= =Ufjw -----END PGP SIGNATURE-----
Yes that is what my post said too, are you using amavisd, spamassassin with your setup up? I was going to try to set my email system with no amavisd and spamassassin to see if that stops the error. It only seems to happen when there is a heavy load on the system - that is more load than it can handle. The system does about 20,000 emails a day which does not seam alot. We have actually two postfix servers. an smtp and mail system. the smtp server sends just fine, except for sending to the mail server where it backs up untill we restart postfix. The mail server just stops receiving email like I started with the posting. I am still confused why postfix would stop like it does. I am going to try the debug_peer_list = 127.0.0.1 like Theo suggests, I am really thinking of installing sendmail to see if that fixes my problem.
The Saturday 2004-12-11 at 10:54 -0800, suse-list wrote:
I am going to try the debug_peer_list = 127.0.0.1 like Theo suggests,
Be careful! If you get thousands of emails on that interface, you will get hundred of thousands of lines written to your log. On the other hand, increased debug info only happens for those SMTP transactions happening over the 127.0.0.1 interface; when using fetchmail, that means all. In you case, perhaps none. -- Cheers, Carlos Robinson
participants (3)
-
Carlos E. R.
-
suse-list
-
Theo v. Werkhoven