- dovecot - dovecot.org

[Dovecot] Stale imap processes
by Michael Bond 01 Sep '06

01 Sep '06

I read through the thread on Stale imap process and would like to chime in that I too am having the problem. The only imap user on the system is using Mail.app 2.1. The other users are using pop with thunderbird. After quiting mail.app and stopping dovecot I can still see many imap sessions opened. Over the course of the day a good number of imap process will be open and no one will be able to connect to the server via imap or pop. Typically I have to restart dovecot twice a day with the following: /etc/init.d/dovecot stop killall imap /etc/init.d/dovecot start I've narrowed it down to an imap issue. If i don't connect with imap the other users don't report any issues connecting with pop. I'm using mbox as my mail format on centos 4. I know that the default version on centos4 is .99, but it appears that people are still having issues into 1.0rc2. Could it have something to do with using mbox format? Mike. mbond(a)the-forgotten.org

1 0

[Dovecot] deliver LDA and INBOX location
by David Lee 01 Sep '06

01 Sep '06

(OS: Fedora Core 5; dovecot: 1.0 rc7) On a typical UNIX-like OS, the INBOXes are in "/var/spool/mail/" using the user identifier: so user 'fred' has INBOX "/var/spool/mail/fred". We have a well-established different convention which subdivides this, based on the last two digits of the uid: "/var/spool/mail/12/fred" (for fred's uid as something ending "[...]12"). This is working nicely with dovecot, using the new rc7 functionality: default_mail_env = mbox:%h:INBOX=/var/spool/mail/%-2.02i/%u:[...] I am now trying to get sendmail to use dovecot's "deliver" as the LDA, (following "http://wiki.dovecot.org/LDA") as a simple drop-in replacement for the normal LDA. (Actually our migration starting point is a UW-IMAP background using its 'tmail' as LDA.) 1. It seems that using "auth default {...socket listen {...}...}" is mandatory. Can someone confirm? This feels like an extra complication: is it strictly necessary in a NIS/PAM set-up? (Fair enough if so.) 2. The main problem is that, according to the log file, the delivery attempt is being made to an INBOX "/var/spool/mail//fred". That is, it seems to be ignoring the "default_mail_env" setting. Have I missed something in the wiki, or is this a bug of some sort? -- : David Lee I.T. Service : : Senior Systems Programmer Computer Centre : : Durham University : : http://www.dur.ac.uk/t.d.lee/ South Road : : Durham DH1 3LE : : Phone: +44 191 334 2752 U.K. :

1 1

[Dovecot] No tcp wrappers, other ideas to help stop brute force attacks?
by David Rees 01 Sep '06

01 Sep '06

I'm looking for a way to deny access to dovecot from certain IP addresses, basically to help prevent brute force attacks on the server. Right now I'm using denyhosts which scans /var/log/secure for authentication failures which then can add an entry to /etc/hosts.deny, but since dovecot doesn't have tcp wrappers support, that doesn't do anything. It doesn't look like I can run dovecot run xinetd. Any other ideas to help protect dovecot from brute force attacks? I don't think pam can help, can it? Otherwise I need to figure out a way to have denyhosts trigger iptables rules or something, or maybe there's another application that will work? -Dave

9 15

[Dovecot] one example, just one example ...
by OpenMacNews 01 Sep '06

01 Sep '06

... of a correctly formatted QUOTA spec in/for a static userdb. That's all I'm askin' for! i've tried a bunch of variants. the most recent two: user@domain.com:{PLAIN}testpass:::::: quota=maildir:storage=4096 user@domain.com:{PLAIN}testpass:::::: maildir:storage=4096 *none* (so far ...) work to override the spec in dovecot.conf. no errors in the logs; rather, simply: dovecot: Aug 31 08:54:25 Info: auth(default): passwd-file /var/MailServer/Data/USERS/imap_user_file: Read 2 users nobody? really, nobody does this successfully? richard

2 2

[Dovecot] [Fwd: dovecot-1.0_rc7 crashes (Login process died too early) with repeated deletes from thunderbird]
by Tim Schafer 01 Sep '06

01 Sep '06

I've run this test on every permutation of poll and notify: can NOT reproduce the crash: --with-ioloop=select --with-notify=none --with-ioloop=select --with-notify=dnotify --with-ioloop=select --with-notify=inotify --with-ioloop=poll --with-notify=none --with-ioloop=poll --with-notify=dnotify --with-ioloop=poll --with-notify=inotify CAN reproduce the crash: --with-ioloop=epoll --with-notify=none --with-ioloop=epoll --with-notify=dnotify --with-ioloop=epoll --with-notify=inotify So it would seem that epoll is related some how. As IDLE only seems to work well with epoll, I'm highly interested in resolving this crash. My Linux kernel is version 2.6.14.2 glibc 2.3.6 If any other info would help, please let me know. If any other test combinations would be helpful, I will try to do them ASAP. Regards, Tim Schafer -------- Original Message -------- Subject: [Dovecot] dovecot-1.0_rc7 crashes (Login process died too early) with repeated deletes from thunderbird Date: Thu, 24 Aug 2006 15:18:56 -0700 From: Tim Schafer <tim_schafer(a)shipfsp.com> To: Dovecot Mailing List <dovecot(a)dovecot.org> I can consistently reproduce a crash in dovecot: Select 20 or more messages and copy them to a test folder. Go into that test folder and SHIFT-DEL the messages rapidly (holding down SHIFT-DEL works best) Just rapidly DELing them (even holding down DEL) doesn't do it (unless it's the Trash folder) I'm using the following namespace (in case that's relevant) namespace private { separator = . prefix = INBOX. inbox = yes } The /var/log/syslog for the session reads: Aug 24 14:58:54 hostname dovecot: Dovecot v1.0.rc7 starting up Aug 24 14:58:55 hostname slapd[7243]: conn=147 fd=9 ACCEPT from IP=127.0.0.1:41074 (IP=0.0.0.0:389) Aug 24 14:58:55 hostname slapd[7243]: conn=147 op=0 BIND dn="" method=128 Aug 24 14:58:55 hostname slapd[7243]: conn=147 op=0 RESULT tag=97 err=0 text= Aug 24 14:59:09 hostname dovecot: auth(default): client in: AUTH^I1^IPLAIN^Iservice=IMAP^Ilip=192.168.0.7^Irip=192.168.0.100 Aug 24 14:59:09 hostname dovecot: auth(default): client out: CONT^I1^I Aug 24 14:59:09 hostname dovecot: auth(default): client in: CONT^I1^IAHRdfmkerjejfkdeie Aug 24 14:59:09 hostname dovecot: auth(default): client out: OK^I1^Iuser=username Aug 24 14:59:09 hostname dovecot: auth(default): master in: REQUEST^I1^I10081^I1 Aug 24 14:59:09 hostname dovecot: auth(default): master out: USER^I1^Iusername^Isystem_user=username^Iuid=1000^Igid=100^Ihome=/home/username Aug 24 14:59:09 hostname dovecot: imap-login: Login: user=<username>, method=plain, rip=192.168.0.100, lip=192.168.0.7 Aug 24 14:59:10 hostname dovecot: auth(default): client in: AUTH^I1^IPLAIN^Iservice=IMAP^Ilip=192.168.0.7^Irip=192.168.0.100 Aug 24 14:59:10 hostname dovecot: auth(default): client out: CONT^I1^I Aug 24 14:59:10 hostname dovecot: auth(default): client in: CONT^I1^IAHRdfmkerjejfkdeie Aug 24 14:59:10 hostname dovecot: auth(default): client out: OK^I1^Iuser=username Aug 24 14:59:10 hostname dovecot: auth(default): master in: REQUEST^I2^I10082^I1 Aug 24 14:59:10 hostname dovecot: auth(default): master out: USER^I2^Iusername^Isystem_user=username^Iuid=1000^Igid=100^Ihome=/home/username Aug 24 14:59:10 hostname dovecot: imap-login: Login: user=<username>, method=plain, rip=192.168.0.100, lip=192.168.0.7 Aug 24 14:59:12 hostname dovecot: auth(default): client in: AUTH^I1^IPLAIN^Iservice=IMAP^Ilip=192.168.0.7^Irip=192.168.0.100 Aug 24 14:59:12 hostname dovecot: auth(default): client out: CONT^I1^I Aug 24 14:59:12 hostname dovecot: auth(default): client in: CONT^I1^IAHRdfmkerjejfkdeie Aug 24 14:59:12 hostname dovecot: auth(default): client out: OK^I1^Iuser=username Aug 24 14:59:12 hostname dovecot: auth(default): master in: REQUEST^I3^I10085^I1 Aug 24 14:59:12 hostname dovecot: auth(default): master out: USER^I3^Iusername^Isystem_user=username^Iuid=1000^Igid=100^Ihome=/home/username Aug 24 14:59:12 hostname dovecot: imap-login: Login: user=<username>, method=plain, rip=192.168.0.100, lip=192.168.0.7 Aug 24 14:59:14 hostname dovecot: IMAP(username): Disconnected Aug 24 14:59:14 hostname dovecot: imap-login: Disconnected: rip=192.168.0.100, lip=192.168.0.7 Aug 24 14:59:14 hostname last message repeated 2 times Aug 24 14:59:14 hostname dovecot: Login process died too early - shutting down Aug 24 14:59:14 hostname dovecot: imap-login: Disconnected: rip=192.168.0.100, lip=192.168.0.7 Aug 24 14:59:14 hostname slapd[7243]: conn=147 op=1 UNBIND Aug 24 14:59:14 hostname slapd[7243]: conn=147 fd=9 closed -- Regards, Tim Schafer

1 1

[Dovecot] MANAGESIEVE patch v2
by Stephan Bosch 01 Sep '06

01 Sep '06

Hello Timo, Dispite the heat out here, I managed to come up with a new version of the managesieve patch for dovecot-1.0rc2: - Fixed the bug (missing CRLF) in the authenticate command - Modified the sieve storage library making the interface much less crude. - The scripts put on the server using the putscript command are now checked before they are accepted. - The reported SIEVE capability is now directly read from the sieve implementation (in this case cmu), listing much more than "FILEINTO VACATION". - Fixed some minor bugs in the putscript command Considering the dovecot-sieve plugin, I noticed that you obviously do not want to include the cmu sieve code in the main dovecot source tree. Unfortunately I did not see any other way but to include this library in my patch to make it support checking the uploaded scripts (for now). There are a few alternatives to solve this issue. We could combine de dovecot-sieve plugin with the managesieve patch yielding a single large sieve plugin. However, this would require some means to let the master know about the new managesieve process type. This could be achieved with some more plugin magic. Another way to solve this problem is by truly only excluding the cmu source from the main dovecot tree. If the interface of the sieve implementation is abstracted somehow, one could build a plugin interface for sieve implementations. This sieve interface would be common to dovecot-deliver and the managesieve daemon. This way, it does not matter how the sieve language is implemented (dovecot could even implement its own and still support cmu and others). I saw a hint to this solution in your original dovecot-lda plugin (i.e. sieve-mailutils.c etc), so you must have been playing with this idea as well. I took the liberty to define an example interface for the sieve implementations in the src/lib-sieve directory. It was concieved very quickly and it is far from complete, but it should support the current features of the delivery process. It is only tested for the managesieve putscript verification (no script was ever run though this interface). I am also not quite happy with the error reporting scheme of this interface. This needs more work and if you don't like it you can of course discard it entirely and define your own. Currently, I am also testing other sieve implementations. I am experimenting with libsieve2 (libsieve.sourceforge.net) and with building a wrapper around 'exim -bf'. This last solution is of course not useful for dovecot-lda, but I think the managesieve daemon is not only useful for dovecot itself. Support for these implementations is not yet included in this patch. The patch can be downloaded at: http://sinas.rename-it.nl/~sirius/dovecot-1.0.rc2-MANAGESIEVE-v2.diff.gz (Simply apply it (-p0) to a clean 1.0rc2 source tree and run autogen.sh etc.) For testing, I have tried other client implementations than horde-ingo, but somehow everyone is very focused on the timsieve daemon and they tend to require its protocol specifics. The only other implementation I got working is smartsieve. The others either crashed or produced protocol errors (i.e. sieveshell). I hope I can get more client implementations to work to ensure the incompatibilies are not my fault. I am very eager to know what you think about the sieve implementation ideas and what you think of the new managesieve patch. Regards, -- Stephan Bosch stephan(a)rename-it.nl

2 1

[Dovecot] rc >= 6, libtool and zsh
by Thomas Hummel 31 Aug '06

31 Aug '06

Hello Timo, On FreeBSD 6.1-STABLE, rc7 doesn't build (rc6 didn't either) because of a broken "sed" in the genarated "libtool" script, at least with the "zsh" shell comming from the ports (zsh-4.3.2). This has occured starting from your move from Autoconf 2.59 to Autoconf 2.60 since for some reason, now "libtool" is run by zsh (instead of sh as in 2.59). I didn't investigate if it comes from : . "me/FreeBSD" : my particular version of zsh, or some patch that may be included in the port . "you" : changes in config.in, some problems in autotools you are using . "Debian" : the 1.5.22 Debian 1.5.22-4 patches of "ltmain.sh" used . "Libtool" : maybe some strange interaction with zsh . "Autoconf" : maybe some bug with the libtool script generation by configure . or some other reason At least, it works if I force CONFIG_SHELL to be /bin/sh -- Thomas Hummel | Institut Pasteur <hummel(a)pasteur.fr> | Pôle informatique - systèmes et réseau

1 0

[Dovecot] Dovecot as LDA with Postfix
by Gerard Seibert 31 Aug '06

31 Aug '06

I have been attempting to replace Procmail with Dovecot as the LDA for my Postfix mail server, but without success. Below is a truncated output from the /var/log/maillog. Aug 9 11:56:20 scorpio postfix/local[4338]: 88C3FC3D1: to=<gerard(a)localhost.seibercom.net>, orig_to=<gerard@localhost>, relay=local, delay=1119, delays=1118/0.71/0/0.11, dsn=4.3.0, status=deferred (temporary failure) Aug 9 11:56:20 scorpio postfix/local[4324]: 818AEBF6F: to=<gerard-gmail(a)localhost.seibercom.net>, orig_to=<gerard-gmail@localhost>, relay=local, delay=495, delays=494/0.06/0/0.81, dsn=4.3.0, status=deferred (temporary failure) Aug 9 11:57:37 scorpio postfix/postfix-script: stopping the Postfix mail system Aug 9 11:57:37 scorpio postfix/master[4233]: terminating on signal 15 Aug 9 12:00:09 scorpio postfix/postfix-script: starting the Postfix mail system Aug 9 12:00:09 scorpio postfix/master[4501]: daemon started -- version 2.4-2000727, configuration /usr/local/etc/postfix Aug 9 12:00:28 scorpio postfix/smtpd[4522]: connect from localhost[127.0.0.1] Aug 9 12:00:28 scorpio postfix/smtpd[4522]: A269EC46D: client=localhost[127.0.0.1] Aug 9 12:00:28 scorpio postfix/cleanup[4526]: A269EC46D: message-id=<44DA0115.40500(a)dlr.de> Aug 9 12:00:28 scorpio postfix/qmgr[4503]: A269EC46D: from=<samba-bounces+gerard=seibercom.net(a)lists.samba.org>, size=7308, nrcpt=1 (queue active) Aug 9 12:00:29 scorpio postfix/local[4527]: A269EC46D: to=<gerard(a)localhost.seibercom.net>, orig_to=<gerard@localhost>, relay=local, delay=0.58, delays=0.2/0.06/0/0.33, dsn=2.0.0, status=sent (delivered to command: /usr/local/bin/procmail) Aug 9 12:00:29 scorpio postfix/qmgr[4503]: A269EC46D: removed With Dovecot as the LDA, no mail is ever delivered. I keep getting this temporary failure message. As soon as I revert to Procmail everything works fine. I am simply trying to deliver to mbox format on a FreeBSD 6.1 STABLE system. Nothing exotic. I am using SASL2. Authentication is via saslauthd -a sasldb. Perhaps I do not have that configured correctly in Dovecot. Dovecot does work fine as a POP server though, so it is running at least partially correctly. I would appreciate any assistance. -- Gerard Seibert gerard(a)seibercom.net Democracy is a device that insures we shall be governed no better than we deserve. George Bernard Shaw

5 13

[Dovecot] RC7: its issues or mine?
by David Lee 31 Aug '06

31 Aug '06

Background: I'm new to dovecot (although with many years Washington IMAP behind me). We're considering migrating from Washington IMAP to dovecot on the main service here, and have just started trying dovecot, using RC7. Washington, IMAP has the usual(-ish) "/var/spool/mail" shared area for the INBOX (trad. UNIX "From " format); a user's folders default to being in their home directory (same format). We are looking towards a "minimal effort" migration to dovecot, so wish to keep this layout. (In an ideal world, we would re-think, including "maildir" etc. But that's not an option at the moment.) I have encountered three issues (so far); I'm not sure whether their mine or something in dovecot (if the latter, whether bugs or features; whether RC7-specific issues; etc.). 1. The overall description of "default_mail_env" seems inadequate (in the conf file, "doc/variables.txt" and "doc/mail-storages.txt"). I'm trying: default_mail_env = mbox:~:INBOX=/var/spool/mail/%-2.02i/%u:INDEX=/tmp/indexes/%d/%n and that seems OK-ish. (The extra "%-2.02i" is because we subdivide the otherwise huge "/var/spool/mail/" directory using the user's uid-mod-100.) But the mere act of an IMAP "login" commands creates a directory (empty) with the liternal name of ~ (tilde). (An IMAP 'list "Mail" "*"' command successfully finds the usr folders in the 'Mail' subdirectory of the user's home directory.) It's almost as if the 'default_mail_env' is telling it to create (literal) '~' before realising that this is shorthand (as in C-shell) for home directory. 2. We have some Pine usage in our UNIX cluster. Historically this has taken advantage of the Pine "rsh mailmachine /etc/rimapd" ability to avoid the need for the password: pre-authentication etc. (Yes, we realise that 'rsh' has security issues.) But when I try making symlink "/etc/rimapd" point to "/usr/dovecot/sbin/dovecot" this fails: Error: Can't use SSL key file /etc/ssl/private/dovecot.pem: Permission denied We have: /etc/ssl/certs/dovecot.pem (mode 444) /etc/ssl/private/dovecot.pem (mode 400) for the sake of secure-IMAP (port 993). But why does this come into the equation at all for this (none-SSL) "rsh ... /etc/rimapd" usage? 3. I was developing and testing this here at work using an account that I mostly use from home using Outlook Express. I was very careful (I think!) only to use the read-only "examine INBOX" command (not "select INBOX"). When I went home and tried it as usual (connecting to our production Washington IMAP service reading that INBOX). But OE showed all the email (including previously read) as "unread" (closed envelope icon). It seems that dovecot has done something to the message headers (even under "examine") that has worried OE. Any thoughts? Finally any hints for NFS-based working? We have a farm of a few Fedora machines running the IMAP processes and the sendmail locally-delivery. Our current "/etc/fstab" NFS spec. for the INBOX area (on a NetApp) is: rw,noac,actimeo=0,vers=3,tcp,timeo=600,rsize=32768,wsize=32768,hard,intr,fg,nosuid Any changes? Issues? Thinks to consider? Etc. Many thanks. -- : David Lee I.T. Service : : Senior Systems Programmer Computer Centre : : Durham University : : http://www.dur.ac.uk/t.d.lee/ South Road : : Durham DH1 3LE : : Phone: +44 191 334 2752 U.K. :

5 8

[Dovecot] IMAPS + Outlook
by Mark Adams 31 Aug '06

31 Aug '06

Hi, We have been using Dovecot 1.0 to serve Maildir format over IMAPs to a client with an office of around 30 people, and a remote office of 10, all using Outlook 2003. While users in the base office (location of server) are not having issues with the server users in the remote office are often having problems with Outlook freezing, or not being able to connect. The dovecot log shows them logging in, but users report not being able to open any mail. We are having similar problems with 2 users that are often on 3G cards, in that the headers download but when they try to open the file it gives the 'must be online to download the message' error, even though they are still connected to their 3G. The net connection for the server is 2mb sdsl, which I have tested myself (using thunderbird on linux!) and achieved good throughput and reliable mailbox navigation. Is there any Outlook work arounds we could be using? (currently using idle timeout work around) Any help is very much appreciated. Thanks, Mark

5 5