Re(2): Re(2): [Exim] Exim goes crazy

Page principale
Supprimer ce message
Répondre à ce message
Auteur: Armin Hornetz
Date:  
À: ph10
CC: exim-users
Nouveaux-sujets: [Exim] Messages frozen UID's not equal to a new users /var/spool/mail/{username}
Sujet: Re(2): Re(2): [Exim] Exim goes crazy
ph10@??? writes:
>On Fri, 8 Dec 2000, Armin Hornetz wrote:
>
>> The problem vanished after restarting exim, but came back the other day.
>
>I presume you mean restarting the daemon. I suppose it could mean the
>daemon's memory is getting corrupted.


Yes, restarting the daemon
>
>
>> The first thought was, that too many exim instances at the same time
>might
>> confuse some shared memory or overload some string-allocations. But the
>> logs show that there was no traffic for some minute and then the first
>> mail after the pause could not be processed because of the corrupt path.
>
>What happened to the message afterwards? Did the next queue run manage
>to process it successfully, or does it continue to give the error? If
>the error continues, it is extremely weird, because there's a new exec
>of Exim for a queue run, i.e. a fresh copy of the configuration is read.


I thought the original process (being listening on the smtp port) is
forking itself for each incoming message. So an error would stay until the
daemon ist restarted.

As I understand, there have been no active queuerunners at the moment. See
the following log lines:

2000-12-06 14:33:08 143egz-0002qO-00 Completed
2000-12-06 14:34:49 143eb5-0002pn-00 <= gerwig@???
H=fc.newsaktuell.ch [195.112.88.227] P=esmtp S=5231798 id=fc.0087cdb70
2000-12-06 14:35:55 143eb5-0002pn-00 => sdacs@??? R=lookuphost
T=remote_smtp H=mx0.gmx.net [213.165.64.100]
2000-12-06 14:35:55 143eb5-0002pn-00 Completed
# These were the last complete messages
2000-12-06 14:36:32 Start queue run: pid=10939
2000-12-06 14:36:32 End queue run: pid=10939
# The queue is empty, so the queue run had nothing to do
2000-12-06 14:38:19 143em2-0002qS-00 <= SYSTEM@???
H=dns-in01.sda-ats.ch [193.73.128.100] P=esmtp S=1567 id=009F431A9C6CA52C.
2000-12-06 14:38:19 143em2-0002qS-00 system_aliases director: failed to
open /et#/aliases for linear search: No such file or directo
2000-12-06 14:38:19 143em2-0002qS-00 == ots@???
D=system_aliases defer (-18): failed to open /et#/aliases for linear sear
# And suddenly this message does not work.
2000-12-06 14:38:40 143emO-0002qU-00 <= SYSTEM@???
H=dns-in01.sda-ats.ch [193.73.128.100] P=esmtp S=2027 id=009F431AA90E082F.
2000-12-06 14:38:40 143emO-0002qU-00 system_aliases director: failed to
open /et#/aliases for linear search: No such file or directo
2000-12-06 14:38:40 143emO-0002qU-00 == ots@???
D=system_aliases defer (-18): failed to open /et#/aliases for linear sear
2000-12-06 14:39:04 143emm-0002qX-00 <= SYSTEM@???
H=dns-in01.sda-ats.ch [193.73.128.100] P=esmtp S=10639 id=009F431AB7A48341
2000-12-06 14:39:04 143emm-0002qX-00 system_aliases director: failed to
open /et#/aliases for linear search: No such file or directo
2000-12-06 14:39:04 143emm-0002qX-00 == ots@???
D=system_aliases defer (-18): failed to open /et#/aliases for linear sear
2000-12-06 14:39:09 143eme-0002qW-00 <= SYSTEM@???
H=dns-in02.sda-ats.ch [193.73.128.101] P=esmtp S=4291 id=009F431A92F477E0.
2000-12-06 14:39:09 143eme-0002qW-00 system_aliases director: failed to
open /et#/aliases for linear search: No such file or directo
2000-12-06 14:39:09 143eme-0002qW-00 == ots@???
D=system_aliases defer (-18): failed to open /et#/aliases for linear sear
2000-12-06 14:39:12 143emt-0002qa-00 <= SYSTEM@???
H=dns-in01.sda-ats.ch [193.73.128.100] P=esmtp S=2313 id=009F431ABC08EC92.
2000-12-06 14:39:12 143emt-0002qa-00 system_aliases director: failed to
open /et#/aliases for linear search: No such file or directo
2000-12-06 14:39:12 143emt-0002qa-00 == ots@???
D=system_aliases defer (-18): failed to open /et#/aliases for linear sear
2000-12-06 14:39:34 143enF-0002qc-00 <= SYSTEM@???
H=dns-in02.sda-ats.ch [193.73.128.101] P=esmtp S=4751 id=009F431A956306E0.
2000-12-06 14:39:34 143enF-0002qc-00 system_aliases director: failed to
open /et#/aliases for linear search: No such file or directo
2000-12-06 14:39:34 143enF-0002qc-00 == ots@???
D=system_aliases defer (-18): failed to open /et#/aliases for linear sear
2000-12-06 14:40:02 system_aliases director: failed to open /et#/aliases
for linear search: No such file or directory
2000-12-06 14:40:02 warning: temporarily unable to resolve sender address:
accepted unverified <lardon@???> H=fc.newsaktu
2000-12-06 14:40:03 143enj-0002qe-00 <= lardon@???
H=fc.newsaktuell.ch [195.112.88.227] P=esmtp S=1199 id=fc.0087cdb70001
2000-12-06 14:40:03 143enj-0002qe-00 => tuenker@???
R=domainlist T=remote_smtp H=mta.newsaktuell.de [195.222.232.26]
2000-12-06 14:40:03 143enj-0002qe-00 -> nebreda@???
R=domainlist T=remote_smtp H=mta.newsaktuell.de [195.222.232.26]
2000-12-06 14:40:03 143enj-0002qe-00 Completed
>

# And another thing: Outgoing messages (that is created on the exim server
machine) were fine. I guess that is because they use a fresh sendmail/exim
call everytime a message is created. Only incoming and passing messages
had the /et# problem

>Hmm. That's not quite true. It is only true if you have set an exim uid
>and gid so that the daemon is not running as root. Have you set an exim
>uid and gid?


we used the majordomo id for all instances, which usually works fine. See
the config below:

system_aliases:
driver = aliasfile
file_transport = address_file
pipe_transport = address_pipe
file = /etc/aliases
search_type = lsearch
user = mdom
...
address_pipe:
driver = pipe
return_output
...
address_file:
driver = appendfile
...

regards

Armin