- dovecot - dovecot.org

[Dovecot] Are host names a secret?
by Timo Sirainen 24 Jul '09

24 Jul '09

Some time ago I added the ability for IMAP clients to fetch messages' GUIDs using FETCH X-GUID command. Because of a bug it wasn't working in 1.2.0 or 1.2.1, but I've fixed it now. But now I'm starting to wonder: With Maildirs the GUIDs are the maildir base filenames, which contain host names. Is it a bad idea to expose them to users?

8 10

[Dovecot] Maildir from offlineimap, INBOX folder created
by Ng Oon-Ee 24 Jul '09

24 Jul '09

Hi, I'm using dovecot to serve up a Maildir created/served by offlineimap. The layout is typical (as far as I know) of maildirs, with the root folder having cur, tmp, new, and subfolders being a folder within the root folder. Since I'd previously used Evolution to access this maildir, my subfolders were actual subfolders (separator /). Anyway, it all works spiffingly well, except that I can't see my inbox at all. What happens is that attempting to subscribe to Inbox will create an INBOX folder in the root folder, which is (of course) empty, so I see an empty folder. For completeness, my dovecot -n... # 1.1.17: /etc/dovecot/dovecot.conf # OS: Linux 2.6.30-ARCH x86_64 ext3 protocols: imap listen: localhost ssl_disable: yes disable_plaintext_auth: no login_dir: /var/run/dovecot/login login_executable: /usr/lib/dovecot/imap-login mail_location: maildir:/home/data/Profiles/OfflineIMAP/root:LAYOUT=fs auth default: passdb: driver: pam userdb: driver: passwd Thanks for the help.

3 11

[Dovecot] Commercial Support
by Tommy Ng 23 Jul '09

23 Jul '09

Hi, my company would like to subscribe the commercial support for the product, may I have more information on this? (e.g. cost, support hours, SLA). Thanks & Regards MJK Yahoo!香港提供網上安全攻略，教你如何防範黑客! 請前往 http://hk.promo.yahoo.com/security/ 了解更多!

2 1

[Dovecot] Problem with sieve scripts including sieve scripts
by Mauricio Tavares 23 Jul '09

23 Jul '09

I have the following global cmusieve scripts defined in /etc/dovecot/sieve, which is owned by the user deliver is run (virtual) as defined in the lda session of dovecot.conf: global-spam.script ====================================== require ["fileinto"]; # Move spam to spam folder if header :contains "X-Spam-Level" "********************" { fileinto "Spam"; # Stop here so that we do not reply on spam stop; } elsif header :contains "X-Spam-Flag" "Yes" { fileinto "Spam"; stop; } global-default.script ==================================== require ["include"]; include :global "global-spam.script"; i.e. all the default script does is call the spam script. I have created the compiled versions using sievec, making sure they are still owned by virtual: mail sieve # ls global-default.script global-spam.script global-vacation.script global-default.scriptc global-spam.scriptc global-vacation.scriptc mail sieve # In the plugins session of dovecot.conf I have defined the path for global scripts (so I can do the include :global thingie) and the path for the default script (just in case user does not have a custom script in the mail directory): plugin { sieve_global_dir = /etc/dovecot/sieve/ global_script_path = /etc/dovecot/sieve/global-default.script } So I send me some certifiable spam from my collection of test spam. This is what I got from the log: Jul 22 14:49:01 mail MailScanner[15047]: New Batch: Scanning 1 messages, 3206 bytes Jul 22 14:49:01 mail MailScanner[15047]: SpamAssassin cache hit for message C04F79F002F.784A8 Jul 22 14:49:01 mail MailScanner[15047]: Spam Checks: Found 1 spam messages Jul 22 14:49:02 mail MailScanner[15047]: Virus and Content Scanning: Starting Jul 22 14:49:02 mail MailScanner[15047]: Requeue: C04F79F002F.784A8 to 304C09F0012 Jul 22 14:49:02 mail postfix/qmgr[14869]: 304C09F0012: from=<raubvogel(a)gmail.com>, size=2518, nrcpt=1 (queue active) Jul 22 14:49:02 mail MailScanner[15047]: Uninfected: Delivered 1 messages Jul 22 14:49:02 mail deliver(raub(a)domain.com): Loading modules from directory: /usr/lib/dovecot/lda Jul 22 14:49:02 mail deliver(raub(a)domain.com): Module loaded: /usr/lib/dovecot/lda/lib90_cmusieve_plugin.so Jul 22 14:49:02 mail dovecot: auth(default): master in: USER 1 raub(a)domain.com service=deliver Jul 22 14:49:02 mail dovecot: auth(default): master out: USER 1 raub(a)domain.com uid=1500 gid=1500 home=/var/spool/vmail/domain.com/raub Jul 22 14:49:02 mail deliver(raub(a)domain.com): auth input: raub(a)domain.com Jul 22 14:49:02 mail deliver(raub(a)domain.com): auth input: uid=1500 Jul 22 14:49:02 mail deliver(raub(a)domain.com): auth input: gid=1500 Jul 22 14:49:02 mail deliver(raub(a)domain.com): auth input: home=/var/spool/vmail/domain.com/raub Jul 22 14:49:02 mail deliver(raub(a)domain.com): maildir: data=/var/spool/vmail/domain.com/raub Jul 22 14:49:02 mail deliver(raub(a)domain.com): maildir++: root=/var/spool/vmail/domain.com/raub, index=, control=, inbox=/var/spool/vmail/domain.com/raub Jul 22 14:49:02 mail deliver(raub(a)domain.com): cmusieve: /var/spool/vmail/domain.com/raub/.dovecot.sieve doesn't exist Jul 22 14:49:02 mail deliver(raub(a)domain.com): cmusieve: Using sieve path: /etc/dovecot/sieve/global-default.script Jul 22 14:49:02 mail deliver(raub(a)domain.com): cmusieve: Executing script /etc/dovecot/sieve/global-default.scriptc Jul 22 14:49:02 mail deliver(raub(a)domain.com): msgid=<000d01ca078a$d13122a0$6400a8c0@chameleonsh>: saved mail to INBOX Jul 22 14:49:02 mail postfix/pipe[15137]: 304C09F0012: to=<raub(a)domain.com>, relay=dovecot, delay=3.4, delays=3.3/0/0/0.1, dsn=2.0.0, status=sent (delivered via dovecot service) From what I gathered, mailscanner saw it was spam and tagged it. When the mail got back, postfix gave it to dovecot deliver, which then passed to cmusieve. Now, cmusieve saw user did not have a script defined, Jul 22 14:49:02 mail deliver(raub(a)domain.com): cmusieve: /var/spool/vmail/domain.com/raub/.dovecot.sieve doesn't exist So, it used the default script as it should, Jul 22 14:49:02 mail deliver(raub(a)domain.com): cmusieve: Using sieve path: /etc/dovecot/sieve/global-default.script Jul 22 14:49:02 mail deliver(raub(a)domain.com): cmusieve: Executing script /etc/dovecot/sieve/global-default.scriptc But, instead of calling the spam script, it simply delivered the spam mail to the inbox, Jul 22 14:49:02 mail deliver(raub(a)domain.com): msgid=<000d01ca078a$d13122a0$6400a8c0@chameleonsh>: saved mail to INBOX Why?

2 2

[Dovecot] 1.1.8 upgrade to 1.2
by Guy 23 Jul '09

23 Jul '09

Hi, I'm hoping to upgrade to 1.2 in the next couple of days. I remember seeing something mentioning some sort of config change, but unfortunately I don't remember where I saw it nor what the config option was. Are there any gotchas I need to watch out for in the change to 1.2 from 1.1.8? I've included dovecot -n below. Thanks Guy #dovecot -n # 1.1.8: /etc/dovecot/dovecot.conf # OS: Linux 2.6.24-23-server x86_64 Ubuntu 8.04.2 protocols: imap imaps pop3 pop3s listen(default): *:143 listen(imap): *:143 listen(pop3): *:110 ssl_listen(default): *:993 ssl_listen(imap): *:993 ssl_listen(pop3): *:995 ssl_cert_file: /etc/ssl/certs/imapd.pem ssl_key_file: /etc/ssl/private/imapd.pem disable_plaintext_auth: no verbose_ssl: yes login_dir: /var/run/dovecot/login login_executable(default): /usr/libexec/dovecot/imap-login login_executable(imap): /usr/libexec/dovecot/imap-login login_executable(pop3): /usr/libexec/dovecot/pop3-login login_greeting: Dovecot ready on mink.domain.org. login_process_per_connection: no login_processes_count: 5 login_max_processes_count: 1024 login_max_connections: 1024 max_mail_processes: 1024 mail_max_userip_connections: 20 verbose_proctitle: yes mail_location: maildir:%h/Maildir/ mail_debug: yes mail_full_filesystem_access: yes mmap_disable: yes dotlock_use_excl: no mail_nfs_storage: yes mail_nfs_index: yes lock_method: dotlock mail_executable(default): /usr/libexec/dovecot/imap mail_executable(imap): /usr/libexec/dovecot/imap mail_executable(pop3): /usr/libexec/dovecot/pop3 mail_plugins(default): imap_quota quota mail_plugins(imap): imap_quota quota mail_plugins(pop3): quota mail_log_max_lines_per_sec: 30 imap_client_workarounds(default): outlook-idle delay-newmail netscape-eoh imap_client_workarounds(imap): outlook-idle delay-newmail netscape-eoh imap_client_workarounds(pop3): pop3_uidl_format(default): %08Xu%08Xv pop3_uidl_format(imap): %08Xu%08Xv pop3_uidl_format(pop3): %08Xv%08Xu pop3_client_workarounds(default): pop3_client_workarounds(imap): pop3_client_workarounds(pop3): outlook-no-nuls oe-ns-eoh namespace: type: private separator: / inbox: yes list: yes subscriptions: yes namespace: type: private separator: / prefix: mail/ location: maildir:%h/Maildir/ hidden: yes subscriptions: yes auth default: cache_size: 2048 cache_ttl: 300 cache_negative_ttl: 1 username_chars: abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ01234567890.-_@' master_user_separator: * debug: yes passdb: driver: passwd-file args: /etc/dovecot/dovecot-master.pwd master: yes passdb: driver: sql args: /etc/dovecot/dovecot-mysql.conf userdb: driver: sql args: /etc/dovecot/dovecot-mysql.conf plugin: quota: maildir quota_rule: *:storage=100M quota_rule2: Trash:ignore -- Don't just do something...sit there!

2 1

[Dovecot] flat file dict
by Sven Strickroth 23 Jul '09

23 Jul '09

Hi, I'm using version 1.2.1 with acl. How do I use acl_shared_dict? Especially with file:-dictionary. What's the layout of that file? Is it possible that the file is ignored with the current version? If the file doesn't exist or isn't readable, I don' get any error message. Best regards, Sven

2 3

[Dovecot] PostLogout on Dovecot
by Fcuk F 23 Jul '09

23 Jul '09

Hi Timo/All Is it possible to have postlogout script for dovecot, similar to as you can have post logging in? Please advise.. Thanks Neil

1 0

[Dovecot] Released Sieve v0.1.9 for Dovecot v1.2.1
by Stephan Bosch 23 Jul '09

23 Jul '09

Hello Dovecot users, Relatively many bugs and problems were reported recently in a short period of time. Apparently, now that Dovecot v1.2 is finally stable, the new Sieve plugin is also starting to get used much more. No known bugs remain. Let's hope I didn't spawn any new ones. Apart from the bugfixes I made the deprecated notify plugin implementation compatible with CMUSieve. You can use it, but keep in mind that the (rather stupid) denotify command is not (yet) implemented and that the $text$ substitution does not work (yes, there is a world of differences compared to enotify). Changelog Sieve v0.1.9: * Removed the unfinished sieve-filter tool from the default build. It is now only built when the --with-unfinished-features switch is supplied during configure. + Started building support for the ereject version of the reject action, which has a preference to use an SMTP/LMTP protocol error in stead of a bounce message. This is to be used to make the Sieve plugin honour Deliver's -e parameter. This is not yet finished and not built by default. + Improved 'Permission denied' error messages just like Dovecot does, precisely specifying what permission is missing to access or create a file. + Added additional headers to the list of allowed headers for the address test. The restrictive nature of the address test is not always appropriate. Still thinking of a better, less restrictive implementation. + Made the deprecated notify extension compatible with the old CMUSieve plugin. However, the denotify command and the $text$ substitution are not yet supported. + Made the discard action log a message to avoid confusion about disappearing messages. - Fixed behavior of fileinto when delivering into a namespace prefix. It now uses silent delivery into INBOX as fallback. - Fixed logging of folder namespace prefix upon delivery into a prefixed namespace. Formerly it only logged the bare folder name. - Fixed a potential segfault in the argument validation. It didn't surface because no command could have a :tag followed by an associated parameter as last argument. - Fixed segfault bug occuring in envelope test when performed on null (<>) envelope path. The fix involves a rather large restructuring of the code to make sure envelope addresses are properly handled everywhere (bug reported by Nikita Koshikov) - Envelope: fixed bug in application of address parts; failure to obtain the part would cause inappropriate match success (bug reported by Ron Lee) - Fixed extension conflict checks during validation. It could sometimes produce useless errormessages. This is currently only used by the deprecated extensions. - Forgot to remove old explicit storage library dependency (patch by Arkadiusz Miskiewicz). - Fixed compiler warnings on certain platforms regarding the use fwrite for outgoing message construction -- The releases are available as follows: http://www.rename-it.nl/dovecot/1.2/dovecot-1.2-sieve-0.1.9.tar.gz http://www.rename-it.nl/dovecot/1.2/dovecot-1.2-sieve-0.1.9.tar.gz.sig -- Have fun testing the new releases and don't hesitate to notify me when there are problems. Regards, -- Stephan Bosch stephan(a)rename-it.nl

3 2

[Dovecot] Compiling Dovecot
by Jason Silkey 23 Jul '09

23 Jul '09

I'm attempting to compile Dovecot 1.2.1 from source with support for ManageSieve, with little success. ./configure seems to complete successfully, but make fails with the message: make: *** No targets. Stop. I checked the Makefile, and it's empty. I run into the same problem no matter the configure options, with or without the managesieve patch applied, and even with earlier versions (I also tried 1.2.0 and 1.1.17 just to see if it would work, but no go). I've had no problems compiling other software, so I'm at a bit of a loss here. Any thoughts?

2 1

[Dovecot] PAM_USER falsely assumed immutable
by Aaron Richton 22 Jul '09

22 Jul '09

In 1.2.1 there's: passdb-pam.c:230 status = pam_get_item(pamh, PAM_USER, &item); passdb-pam.c:237 auth_request_set_field(request, "user", item, NULL); so "item" is PAM_USER, which is then checked by auth_request_set_field: 1022 if (strcmp(request->user, value) != 0) { 1023 auth_request_log_debug(request, "auth", 1024 "username changed %s -> %s", 1025 request->user, value); that it hasn't changed. You're not allowed to assume that PAM_USER doesn't change. See, for example, http://www.kernel.org/pub/linux/libs/pam/Linux-PAM-html/adg-security-user-i… to say nothing of the fact that Rutgers has PAM modules that do exactly that (change PAM_USER). This check needs to be relaxed (or, perhaps if you want a config directive for DontComplyWithThePAMSpec = true, you can have a tunable). Can this be as simple as ditching the call to auth_request_set_field, or is there concern over interactions between PAM and other auth features?

2 2