- dovecot - dovecot.org

[Dovecot] Dovecot 1.1.1 killed with SIGABRT
by Andreas M. Kirchwitz 08 Aug '08

08 Aug '08

Hi folks, with Dovecot 1.1.1, I've had two "crashes" (SIGABRT) recently. Dovecot 1.1.1 has been compiled from source and is running on Fedora 8 Linux 32 Bit (all patches) with custom OpenSSL 0.9.8h. [ output of "dovecot -n" ] # 1.1.1: /usr/local/dovecot/etc/dovecot.conf ssl_cert_file: /usr/local/dovecot/etc/dovecot.crt ssl_key_file: /usr/local/dovecot/etc/dovecot.key login_dir: /usr/local/dovecot/var/run/dovecot/login login_executable: /usr/local/dovecot/libexec/dovecot/imap-login mail_location: mbox:~/Mail:INBOX=/var/spool/mail/%u auth default: mechanisms: plain login digest-md5 cram-md5 passdb: driver: passwd-file args: /usr/local/dovecot/etc/dovecot.passwd userdb: driver: passwd-file args: /usr/local/dovecot/etc/dovecot.passwd First, the "small" problem with Pine 4.64: Jul 20 03:17:04 linux dovecot: imap-login: Login: user=<user1>, method=CRAM-MD5, rip=138.201.2.8, lip=138.201.2.2, TLS Jul 20 03:17:53 linux dovecot: IMAP(user1): UIDVALIDITY changed (1215137127 -> 1216072777) in mbox file /home/user1/Mail/Trash Jul 20 03:17:53 linux dovecot: Panic: IMAP(user1): file mail-index-transaction.c: line 548 (mail_index_transaction_get_next_uid): assertion failed: (recs[count-1].uid >= next_uid) Jul 20 03:17:53 linux dovecot: IMAP(user1): Raw backtrace: imap [0x80cf2f0] -> imap [0x80cf34a] -> imap [0x80cec5c] -> imap [0x80ad621] -> imap [0x80ae56b] -> imap(mail_cache_get_first_new_seq+0x12) [0x80a1d82] -> imap(mail_cache_field_want_add+0x8d) [0x80a555d] -> imap(index_mail_parse_header_init+0x1cc) [0x80964ac] -> imap(index_mail_cache_parse_init+0x4e) [0x809696e] -> imap(mbox_save_init+0x457) [0x807d4a7] -> imap(mailbox_save_init+0x51) [0x809fd71] -> imap(mail_storage_copy+0xc1) [0x809df31] -> imap(cmd_copy+0x1d2) [0x805b1f2] -> imap(cmd_uid+0x59) [0x805f489] -> imap [0x805fe2c] -> imap [0x805fed5] -> imap [0x8060695] -> imap(client_input+0x5e) [0x80608ae] -> imap(io_loop_handler_run+0x100) [0x80d6c30] -> imap(io_loop_run+0x28) [0x80d5dc8] -> imap(main+0x4a1) [0x80683d1] -> /lib/libc.so.6(__libc_start_main+0xe0) [0x149390] -> imap [0x805a211] Jul 20 03:17:53 linux dovecot: child 19853 (imap) killed with signal 6 The user didn't configure his Pine properly. He used IMAP to access his INBOX, but he forgot to configure this for all other folders as well. Now, Pine modified /home/user1/Mail/Trash directly, and when Dovecot stumbled upon this file it killed itself with signal 6. The problem is not reproducible. After this single failure, Dovecot doesn't seem to have any problems with the folder. However, instead of this panic operation, Dovecot maybe could issue a warning or error response to the client so that the user (hopefully) gets a clue about what's going wrong. Second, the "big" problem with Thunderbird 2.0.0.14: Jul 12 01:04:06 linux dovecot: imap-login: Login: user=<user2>, method=CRAM-MD5, rip=138.201.2.4, lip=138.201.2.2, TLS Jul 12 01:04:45 linux dovecot: Panic: IMAP(user2): file index-sync.c: line 39 (index_mailbox_set_recent_uid): assertion failed: (seq_range_exists(&ibox->recent_flags, uid)) Jul 12 01:04:45 linux dovecot: IMAP(user2): Raw backtrace: imap [0x80cf2f0] -> imap [0x80cf34a] -> imap [0x80cec5c] -> imap [0x809ccba] -> imap [0x8080d10] -> imap(mbox_sync+0x522) [0x80817d2] -> imap [0x807a334] -> imap(index_transaction_commit+0x4e) [0x809d99e] -> imap(cmd_copy+0x35f) [0x805b37f] -> imap(cmd_uid+0x59) [0x805f489] -> imap [0x805fe2c] -> imap [0x805fed5] -> imap [0x8060695] -> imap(client_input+0x5e) [0x80608ae] -> imap(io_loop_handler_run+0x100) [0x80d6c30] -> imap(io_loop_run+0x28) [0x80d5dc8] -> imap(main+0x4a1) [0x80683d1] -> /lib/libc.so.6(__libc_start_main+0xe0) [0x149390] -> imap [0x805a211] Jul 12 01:04:45 linux dovecot: child 25672 (imap) killed with signal 6 [happens over and over in an endless loop until Thunderbird is killed] This has been a nasty one. Different user, different mail client. INBOX with 350 messages in total, 3 of them new. Started Thunderbird, index page is displayed fine, then clicked on one of the 3 new messages to read it. For no apparent reason, Dovecot dies with signal 6. Thunderbird reconnected automatically, and then Dovecot dies again. Nice endless loop ... Stopping and restarting the complete Dovecot subsystem with the help of the init.d script didn't help. Finally the crashes stopped after removing the cache files in /home/fubar/Mail/.imap/INBOX. It's obvious that the cache files got corrupted somehow. There was plenty of disk space free. And the cache files have been freshly created from scratch with Dovecot 1.1.0. Other folders were not affected. Except Dovecot, nobody ever touches these cache files. Well, the good news is, it stopped crashing after removing the cache files. The bad news is, without the cache files (which I removed) the problem cannot be reproduced. So, there's nothing more to offer than these two backtraces. Both crashes are not related to each other and happened to different users with different clients. Maybe the backtraces and my descriptions help to find what's going wrong or how to make Dovecot behave more robustly. If there's any additional information needed, please let me know. Sorry that I didn't save the corrupted cache files. Greetings, Andreas

3 6

Re: [Dovecot] Auth message
by Pavel Shirov 08 Aug '08

08 Aug '08

Further testing show that such message could occur simply by providing random username/password and termination session afterwards (not sending QUIT).

1 0

[Dovecot] Dovecot auto vacation with sieve doesn't work
by CJ Keist 08 Aug '08

08 Aug '08

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Sorry if this comes in twice, I sent this first from my test account. - --------- Guys, I'm running out of hair to pull out ;). Can anyone out there say that this does work?? To me this is looking like a bug and I'm not sure whether it's the sieve plugin or the dovecot deliver program. I have also had no luck the the "reject" sending any reply back to the sender. My setup: OS: Solaris 10 sparc platform postfix: 2.5.2 dovecot: 1.1.2 sieve plugin: 1.1.5 I'm going back forth between procmail and sieve and having no luck. procmail will send the reply to the sender but doesn't deliver the message to the recipient. Sieve will deliver the message to the recipient but doesn't send the reply to the sender. Now if I can just combine the two, I'll be in business. Postfix mail command: mailbox_command = /userM/mail-services/dovecot/libexec/dovecot/deliver - -d $USER* * Dovecot -n (I've made some changes from my previous posting namely with the master_auth settings) # 1.1.2: /userM/mail-services/dovecot/etc/dovecot.conf Warning: fd limit 256 is lower than what Dovecot can use under full load (more than 3072). Either grow the limit or change login_max_processes_count and max_mail_processes settings base_dir: /userM/mail-services/dovecot/var/run/dovecot/ protocols: imap imaps pop3 pop3s ssl_cert_file: /userM/mail-services/dovecot/etc/ssl/dovecot.pem ssl_key_file: /userM/mail-services/dovecot/etc/ssl/privkey.pem ssl_parameters_regenerate: 0 ssl_cipher_list: MEDIUM:!LOW disable_plaintext_auth: no login_dir: /userM/mail-services/dovecot/var/run/dovecot/login login_executable(default): /userM/mail-services/dovecot/libexec/dovecot/imap-login login_executable(imap): /userM/mail-services/dovecot/libexec/dovecot/imap-login login_executable(pop3): /userM/mail-services/dovecot/libexec/dovecot/pop3-login login_greeting: Cyrus ready. login_max_processes_count: 1024 max_mail_processes: 1024 mail_location: mbox:~/:INBOX=/userM/mail/%u mail_debug: yes mmap_disable: yes mail_nfs_index: yes mail_executable(default): /userM/mail-services/dovecot/libexec/dovecot/imap mail_executable(imap): /userM/mail-services/dovecot/libexec/dovecot/imap mail_executable(pop3): /userM/mail-services/dovecot/libexec/dovecot/pop3 mail_plugin_dir(default): /userM/mail-services/dovecot/lib/dovecot/imap mail_plugin_dir(imap): /userM/mail-services/dovecot/lib/dovecot/imap mail_plugin_dir(pop3): /userM/mail-services/dovecot/lib/dovecot/pop3 auth default: mechanisms: plain login verbose: yes debug: yes passdb: driver: pam passdb: driver: passwd args: blocking=yes userdb: driver: passwd args: blocking=yes socket: type: listen client: path: /userM/mail-services/postfix/var/private/auth mode: 432 user: postfix group: postfix master: path: /userM/mail-services/dovecot/var/run/dovecot/auth-master mode: 384 user: postfix group: postfix My .dovecot.sieve file: require ["vacation"]; if header :matches "subject" "*" { vacation :days 30 :addresses "cj.keist(a)colostate.edu" :subject "Automatic response to: ${1}" "I'm away -- send mail to foo in my absence"; } I made some modifications to the vacation reply base on the following URL: http://ietfreport.isoc.org/idref/draft-ietf-sieve-vacation/ In my testing I know the sieve plugin is reading my .dovecot.sieve script and creating the .dovecot.sievec, but that is as for as it goes. The .dovecot.lda-dupes file is not being created (or any .dovecot.sieve.err file) and no reply is even being attempted from the logs: Aug 6 11:30:51 mail2.engr.colostate.edu dovecot: [ID 107833 mail.info] auth(def ault): new auth connection: pid=27187 Aug 6 11:30:51 mail2.engr.colostate.edu postfix/smtpd[27187]: [ID 197553 mail.i nfo] connect from akron.colostate.edu[129.82.103.160] Aug 6 11:30:51 mail2.engr.colostate.edu postfix/smtpd[27187]: [ID 197553 mail.i nfo] setting up TLS connection from akron.colostate.edu[129.82.103.160] Aug 6 11:30:51 mail2.engr.colostate.edu postfix/smtpd[27187]: [ID 197553 mail.i nfo] Anonymous TLS connection established from akron.colostate.edu[129.82.103.16 0]: TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits) Aug 6 11:30:52 mail2.engr.colostate.edu postfix/smtpd[27187]: [ID 197553 mail.i nfo] 16E2059984: client=akron.colostate.edu[129.82.103.160] Aug 6 11:30:52 mail2.engr.colostate.edu postfix/cleanup[27192]: [ID 197553 mail .info] 16E2059984: message-id=<3b676ae5f90fadad863930dcc61db3f8.squirrel@webmail .engr.colostate.edu> Aug 6 11:30:52 mail2.engr.colostate.edu postfix/smtpd[27187]: [ID 197553 mail.i nfo] disconnect from akron.colostate.edu[129.82.103.160] Aug 6 11:30:52 mail2.engr.colostate.edu postfix/qmgr[18021]: [ID 197553 mail.in fo] 16E2059984: from=<joeblow(a)engr.colostate.edu>, size=1877, nrcpt=1 (queue act ive) Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 315030 mail.info] Lo ading modules from directory: /userM/mail-services/dovecot/lib/dovecot/lda Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 619685 mail.info] Mo dule loaded: /userM/mail-services/dovecot/lib/dovecot/lda/lib90_cmusieve_plugin. so Aug 6 11:30:52 mail2.engr.colostate.edu dovecot: [ID 107833 mail.info] auth(def ault): master in: USER 1 cjay service=deliver Aug 6 11:30:52 mail2.engr.colostate.edu dovecot: [ID 107833 mail.info] auth-wor ker(default): passwd(cjay): lookup Aug 6 11:30:52 mail2.engr.colostate.edu dovecot: [ID 107833 mail.info] auth(def ault): master out: USER 1 cjay system_user=cjay uid=10946 gid=32 home=/top/admin/ENS/people/cjay/home Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 868608 mail.info] au th input: cjay Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 868608 mail.info] au th input: system_user=cjay Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 868608 mail.info] au th input: uid=10946 Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 868608 mail.info] au th input: gid=32 Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 868608 mail.info] au th input: home=/top/admin/ENS/people/cjay/home Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 848615 mail.info] mb ox: data=/top/admin/ENS/people/cjay/home/:INBOX=/userM/mail/cjay Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 885103 mail.info] fs : root=/top/admin/ENS/people/cjay/home, index=, control=, inbox=/userM/mail/cjay Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 223526 mail.info] cm usieve: Using sieve path: /top/admin/ENS/people/cjay/home/.dovecot.sieve Aug 6 11:30:52 mail2.engr.colostate.edu deliver(cjay): [ID 507699 mail.info] cm usieve: Executing script /top/admin/ENS/people/cjay/home/.dovecot.sievec Aug 6 11:30:53 mail2.engr.colostate.edu deliver(cjay): [ID 702911 mail.info] ms gid=<3b676ae5f90fadad863930dcc61db3f8.squirrel(a)webmail.engr.colostate.edu>: save d mail to INBOX Aug 6 11:30:53 mail2.engr.colostate.edu postfix/local[27193]: [ID 197553 mail.i nfo] 16E2059984: to=<cjay(a)engr2.colostate.edu>, relay=local, delay=1.2, delays=0 .32/0.08/0/0.82, dsn=2.0.0, status=sent (delivered to command: /userM/mail-servi ces/dovecot/libexec/dovecot/deliver -d $USER) Aug 6 11:30:53 mail2.engr.colostate.edu postfix/qmgr[18021]: [ID 197553 mail.in fo] 16E2059984: removed Aug 6 11:34:12 mail2.engr.colostate.edu postfix/anvil[27190]: [ID 197553 mail.i nfo] statistics: max connection rate 1/60s for (smtp:129.82.103.160) at Aug 6 1 1:30:51 Aug 6 11:34:12 mail2.engr.colostate.edu postfix/anvil[27190]: [ID 197553 mail.i nfo] statistics: max connection count 1 for (smtp:129.82.103.160) at Aug 6 11:3 0:51 Aug 6 11:34:12 mail2.engr.colostate.edu postfix/anvil[27190]: [ID 197553 mail.i nfo] statistics: max cache size 1 at Aug 6 11:30:51 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFImeYWA29OFr7C6jcRAkj+AJ0eKINCZmRDWh8gSjbW86Ga6DPWlgCeL/1M rVvbvn6al/3fLep60h2Lxg4= =veQm -----END PGP SIGNATURE-----

3 5

[Dovecot] Expire Plugin - LDA with Vpopmail errors
by Ron Culler 08 Aug '08

08 Aug '08

I am trying to setup Dovecot to use the expire plugin and wanted to use the LDA so I can ensure the messages are indexed as they are delivered to the mailboxes. This system is only accessed via a webmail interface Using dovecot-1.1.2 and Vpopmail with Qmail Dovecot.conf file base_dir = /usr/local/var/run/dovecot/ protocols = imap disable_plaintext_auth = no shutdown_clients = yes log_path = /var/log/dovecott info_log_path = /var/log/dovecott log_timestamp = "" ssl_disable = yes login_dir = /usr/local/var/run/dovecot/login login_user = dovecot login_process_per_connection = yes login_processes_count = 3 login_max_processes_count = 128 mail_location = maildir:~/Maildir namespace private { separator = . prefix = INBOX. inbox = yes } verbose_proctitle = yes first_valid_uid = 89 #last_valid_uid = 0 maildir_copy_with_hardlinks = yes maildir_copy_preserve_filename = no protocol imap { listen = 127.0.0.1:4143 imap_max_line_length = 65536 login_greeting_capability = yes mail_plugins = expire } protocol lda { mail_plugins = expire log_path = /var/log/dovecot-deliver.log info_log_path = /var/log/dovecot-deliver.log } plugin { # Trash and its children 7d, Spam 30d expire = Trash 7 expire_dict = proxy::expire } auth default { mechanisms = plain passdb vpopmail { args = } userdb vpopmail { } user = vpopmail count = 1 ssl_require_client_cert = no } I see errors in the /var/log/dovecot-deliver.log deliver (email@addrss) Error: Can't connect to auth server at /usr/local/var/run/dovecot//auth-master: no such file or directory. What do I need to do to create this? Ron Culler

2 1

[Dovecot] virtual domains and SSL certificates
by Kacper Wysocki 07 Aug '08

07 Aug '08

Hi all, I have dovecot 1.1.0 setup to access vpopmail accounts for several virtual domains. Dovecot IMAP is accessed through several virtual domains as well, ie mail.foo.com an d mail.bar.com The problem is that the configuration file specifies only one certificate file for dovecot, which means only one Common Name, which means one cannot provide one server cert that will match mail.foo.com AND mail.bar.com, and either mary(a)foo.com or bob(a)bar.com will get a "Security Error: Domain Name Mismatch" in their mail client when connecting through IMAPS. How can I avoid this domain name mismatch error? Regards, Kacper Wysocki

4 4

[Dovecot] RESOLVED: dovecot-imap displays empty domainalias mailboxen
by Kacper Wysocki 07 Aug '08

07 Aug '08

On Thu, Jul 31, 2008 at 5:21 PM, Kacper Wysocki <kacperw(a)online.no> wrote: > I'm using dovecot 1.0.13 as the IMAPS frontend for a vpopmail/qmail > virtual domain setup. Dovecot is set up as follows: > passdb vpopmail{ > } > userdb vpopmail{ > } > > I've added a domain alias for the main domain with vaddaliasdomain and > email to post@aliasdomain is delivered to post@maindomain and can be > checked with pop. > > However, when I try to check IMAP through the account post@aliasdomain > I am confronted with an empty mailbox. If I log in using > post@maindomain I get the expected contents. > > Why can't my users log in using user@aliasdomain to check their emails? Update: dovecot can't find the mail directory because of mail_location = maildir:/var/vpopmail/domains/%d/%n/Maildir and dovecot doesn't know anything about domain aliases. ln -s ~vpopmail/domains/maindomain ~vpopmail/domains/aliasdomain fixes the problem for dovecot. -K

1 0

Re: [Dovecot] Dovecot 1.1 Quota Plugin Configuration?
by David Reid 07 Aug '08

07 Aug '08

Charles Marcus wrote: > On 8/7/2008 8:39 AM, David Reid wrote: >> With 1.0 I had the quota plugin set to allow some accounts to have no >> quota by setting their size (in bytes) to 0 (via mysql). Following the >> upgrade to 1.1 this no longer seems to work and has resulted in a lot of >> annoyed people having messages bumped :-( >> >> The config I have is as follows... >> >> quota = maildir >> quota_rule = *:storage=15M >> quota_rule2 = Trash:ignore >> quota_rule3 = INBOX.Trash:ignore >> quota_rule4 = INBOX.SpamTrain:ignore >> quota_rule5 = SpamTrain:ignore >> >> In the sql.conf I get the quota for each mailbox using this line >> >> user_query = SELECT '/var/vmail/%d/%n' as home, >> 'maildir:/var/vmail/%d/%n' as mail, xxx AS uid, xxx AS gid, >> concat('maildir:storage=', mb_quota) AS quota FROM >> mailboxes WHERE mb_mbname = '%u' AND mb_active = '1' >> >> With this setup when I have a mb_quota set to 0 it simply bounces all >> messages for the mailboxes that should have no quota. >> >> Any ideas how I need to fix this? > > You could start by not posting the same question twice in a 24 hour > period, especially when your question has already been answered. Apologies, but as the problem affected the mailbox I posted from I missed the reply and hence the answer. > > You are trying to use the same mysql query from 1.0 for 1.1 and it won't > work. > > http://wiki.dovecot.org/Quota/1.1 > Knowing the sql needed to be adjusted I have made the changes. The wiki page says that the uid and guid can be set in the config file, but doesn't give any indication of the lines to use. Maybe the information could be added? Thanks for your (duplicate) reply. david

1 0

[Dovecot] Dovecot 1.1 Quota Plugin Configuration?
by David Reid 07 Aug '08

07 Aug '08

With 1.0 I had the quota plugin set to allow some accounts to have no quota by setting their size (in bytes) to 0 (via mysql). Following the upgrade to 1.1 this no longer seems to work and has resulted in a lot of annoyed people having messages bumped :-( The config I have is as follows... quota = maildir quota_rule = *:storage=15M quota_rule2 = Trash:ignore quota_rule3 = INBOX.Trash:ignore quota_rule4 = INBOX.SpamTrain:ignore quota_rule5 = SpamTrain:ignore In the sql.conf I get the quota for each mailbox using this line user_query = SELECT '/var/vmail/%d/%n' as home, 'maildir:/var/vmail/%d/%n' as mail, xxx AS uid, xxx AS gid, concat('maildir:storage=', mb_quota) AS quota FROM mailboxes WHERE mb_mbname = '%u' AND mb_active = '1' With this setup when I have a mb_quota set to 0 it simply bounces all messages for the mailboxes that should have no quota. Any ideas how I need to fix this? david

1 0

[Dovecot] Experience moving mailboxes from Dovecot 0.99.14 to Dovecot 1.07 => Improvement possible
by Benoît DESLOGES 07 Aug '08

07 Aug '08

Hi, This small mail to share my observation about a recent move of mailboxes between two servers and ask about explanation and/or improvement about UIDL in dovecot. SV1 : Dovecot 0.99.14 / Red Hat Fedore Core 4 SV2 : Dovecot 1.07 / CentOS 5.2 Mailboxes in /var/spool/mail on the twoo servers. Test will be done with outlook express with option "leave message on server" checked. Goal is simply to move users mailboxes from SV1 to SV2 without re-downloading all messages. Try 1: - Stop dovecot on SV2 - Clear all dovecot indexes on SV2 - Rsync of my mailbox - Start dovecot on SV2 - Update pop setting in outlook and send/receive Result : => starting to download 3000 mails... After some debug, I discovered that UIDL where not of the same format => put pop3_uidl_format = %v.%u in dovecot.conf of SV2. Try 2: Same operations as Try1 Result : => starting to download 3000 mails... UIDL's where of the same format but values where not corresponding... Teleting on SV1 and asking UIDL, last value is xxxxxxxxxx.85878. On SV2, xxxxxxxxxx.85879 was the FIRST value of the list. Conclusion for the moment, Dovecot has a problem with the detection of existing UID of the moved mailbox. Comparaison of first header found from the two mailboxes show ... differences ! SV1 give the following : (...) X-UID: 70570 Content-Length: 1561 X-IMAPbase: xxxxxxxxxx 85845 $MDNSent X-Keywords: (...) SV2 give the following (after first POP attempt) : (...) Content-Length: 1561 X-IMAPbase: xxxxxxxxxx 0000089204 $MDNSent X-Keywords: X-UID: 85846 (...) Observation : X-UID: 85846 is not at the same place in the header. Try 3: - Stop dovecot on SV2 - Clear all dovecot indexes on SV2 - Rsync of my mailbox - Edit mailbox and move X-UID header after the X-Keywords header - Start dovecot on SV2 - Update pop setting in outlook and send/receive Result : => OK. Is that normal that dovecot is impacted by the position of the headers ? Maybe some improvement possible ? Maybe theyre is another method to make my migration ? Any observation or comment is welcome... -- Benoît Desloges Network Engineer

7 8

[Dovecot] Fwd: Re: sasl parameters missing (in postfix)
by Daniel Black 07 Aug '08

07 Aug '08

In response to my request for postfix to support dovecot auth arguments I got the forwarded reply. If someone gets around to this before me I won't be offended. ------------ Story is I deployed a webmail with certificate based authentication that substitutes a global master password (http://wiki.dovecot.org/Authentication/MasterUsers) when the certificate matches. The webmail accesses the inbox by imap and reuses the password for smtp through postfix. I configured dovecot sasl authentication to allow a particular global password to be allowed from one IP address of the webmail server. Unfortuanately it seems as though postfix doesn't pass rip= (remote ip) or the other AUTH parameters of the protocol (http://dovecot.org/doc/auth-protocol.txt) Is adding these parameters to postfix's sasl authentication a useful feature request? ---------- Forwarded Message ---------- Subject: Re: sasl parameters missing Date: Thu, 7 Aug 2008 From: Wietse Venema <wietse(a)porcupine.org> To: Daniel Black <daniel.subs(a)internode.on.net> Daniel Black: > Thanks Wietse, > > On Tue, 5 Aug 2008 09:30:44 am Wietse Venema wrote: > > Postfix passes the information in the SMTP client's AUTH command. > > This is how I got the Dovecot extension from Timo. If someone is > > willing to monitor his docs for changes, > > it seems fairly stable. Going off the doc/auth-protocol.txt changelog > Nov 12 2006 lport/rport was added. > Aug 07 2005 changed valid-client-cert to ssl-valid-cert > Oct 22 2004 original documentation > > Current implementation of the authentication server in dovecot seems to ignore > parameters it doesn't understand. > > > then they are welcome to do so. I won't. > > On the basis of this apparent stability and compatibility would you consider > accepting a patch? Yes. No promise, though, that it will be adopted. One consideration is that Postfix does not talk directly to Dovecot, but instead talks to an abstraction layer that is used for both Cyrus SASL and for Dovecot. Obviously, that XSASL abstraction layer must not be made specific to the underlying Cyrus SASL or Dovecot implementation. The solution therefore is not to extend XSASL functions with one extra argument for each Dovecot feature. Apart from being Dovecot-specific, functions with many parameters are difficult to update correctly; compilers can't always tell that two arguments should be swapped. I solved the problem of many-parameter functions by using macros such as TLS_SERVER_START(). This gives more assurance that data is passed correctly, and it less likely to break due to human maintainer error. Wietse ------------------------------------------------------- -- Daniel Black -- Proudly a Gentoo Linux User. Gnu-PG/PGP signed and encrypted email preferred http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x76677097 GPG Signature D934 5397 A84A 6366 9687 9EB2 861A 4ABA 7667 7097

1 0