- dovecot - dovecot.org

dovecot 2.2.13 core dump (signal 11)
by Tom Lieuallen 07 Jul '14

07 Jul '14

We just upgraded from 2.1.16 to 2.2.13. One user is experiencing problems when using the shared namespace. Specifically it happens when opening/reading a message and when deleting messages. These functions work for other users. So far, this appears to be only affecting one user. The shared folder is a local ext4 file system and utilizes acls # 2.2.13: /private/dovecot/etc/dovecot/dovecot.conf doveconf: Warning: service auth { client_limit=1000 } is lower than required under max. load (15360) doveconf: Warning: service anvil { client_limit=1000 } is lower than required under max. load (10243) # OS: Linux 2.6.32-358.11.1.el6.x86_64 x86_64 Red Hat Enterprise Linux Server release 6.5 (Santiago) default_process_limit = 5120 first_valid_uid = 100 listen = * mail_location = mbox:~/mail:INBOX=/var/mail/%u:INDEX=/a2/imap-index/%u mail_plugins = quota acl namespace { hidden = yes inbox = no list = children location = maildir:/a1/dove-shared:INDEX=/a2/imap-index/dove-shared/%u prefix = sharedimap/ separator = / type = shared } namespace inbox { inbox = yes location = mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { special_use = \Trash } prefix = } passdb { args = /private/dovecot/etc/dovecot/dovecot-ldap.conf.ext driver = ldap } passdb { args = scheme=CRYPT username_format=%u /private/dovecot/etc/passwd driver = passwd-file } plugin { acl = vfile quota = fs:INBOX:mount=/a1 quota2 = fs:Home quota:mount=%h } protocols = imap lmtp service imap-login { service_count = 1 } service imap { process_limit = 2048 } ssl_cert = </private/ssl/certs/mail.engr.oregonstate.edu.pem ssl_key = </private/ssl/certs/mail.engr.oregonstate.edu.pem userdb { args = /private/dovecot/etc/dovecot/dovecot-ldap.conf.ext driver = ldap } userdb { args = username_format=%u /private/dovecot/etc/passwd driver = passwd-file } protocol imap { mail_max_userip_connections = 20 } Here's the backtrace in gdb. Core was generated by `dovecot/imap'. Program terminated with signal 11, Segmentation fault. #0 acl_mailbox_right_lookup (box=0x12e6730, right_idx=7) at acl-mailbox.c:40 40 if (abox->skip_acl_checks) Missing separate debuginfos, use: debuginfo-install glibc-2.12-1.132.el6_5.2.x86_64 sssd-client-1.9.2-129.el6.x86_64 (gdb) bt full #0 acl_mailbox_right_lookup (box=0x12e6730, right_idx=7) at acl-mailbox.c:40 abox = 0x0 alist = <value optimized out> ret = <value optimized out> #1 0x00007f4c11604d06 in acl_copy_has_rights (ctx=0x12f9810, mail=0x12f5030) at acl-mailbox.c:409 destbox = 0x12edf90 save_right = <value optimized out> #2 acl_copy (ctx=0x12f9810, mail=0x12f5030) at acl-mailbox.c:430 t = <value optimized out> abox = 0x12ee5f8 #3 0x00007f4c11b741ed in mailbox_copy (_ctx=<value optimized out>, mail=0x12f5030) at mail-storage.c:2149 ctx = 0x12f9810 t = 0x12f32d0 keywords = 0x12f9b80 pvt_flags = 0 backend_mail = 0x12f5030 ret = <value optimized out> __FUNCTION__ = "mailbox_copy" #4 0x00007f4c11b742dd in mailbox_move (_ctx=<value optimized out>, mail=0x12f5030) at mail-storage.c:2170 ctx = 0x12f9810 #5 0x000000000040c587 in fetch_and_copy (cmd=0x12e3f10, move=true) at cmd-copy.c:67 search_ctx = 0x12f4db0 src_trans = 0x12f4090 srcset_ctx = {str = 0x12bf328, first_uid = 0, last_uid = 4294967295} ret = 1 save_ctx = 0x0 mail = 0x12f5030 copy_count = 1 src_uidset = 0x12bf328 #6 cmd_copy_full (cmd=0x12e3f10, move=true) at cmd-copy.c:123 client = 0x12e3330 dest_storage = <value optimized out> destbox = 0x12edf90 t = 0x12f32d0 src_trans = <value optimized out> search_args = 0x12ecf80 messageset = 0x12cac10 "112850,112871,112903:112905,112919" mailbox = 0x12cac38 "Trash" src_uidset = <value optimized out> sync_flags = 0 imap_flags = 0 changes = {pool = 0x12e6730, uid_validity = 19807984, saved_uids = { (I'm not sure how much more info is needed; let me know if more...) thank you Tom Lieuallen

2 3

Multiple Instances with Director
by Cassidy Larson 07 Jul '14

07 Jul '14

I have four front end servers with two backend servers. Each front end server has three instances for different mail hosts (individual SSL certs, etc). I'm looking at adding a fourth instance on each front end server for the director process. Currently we just return the backend server ip via SQL and use the old master username/password to proxy the inbound POP/IMAP sessions to the correct backend server. The plan is for each instance to have connect to the same list of director_servers (one from each front end server). What's the trick to separate a director instance into its own setup and having multiple instances/servers connect to that director? Currently I have on my proxy instance: director_servers = 10.128.1.42:9090 (this is the inet_listener from the director instance) service director { unix_listener login/director { mode = 0666 } fifo_listener login/proxy-notify { mode = 0600 user = $default_login_user } unix_listener director-userdb { mode = 0600 } } And I get in logs: director: Fatal: No inet_listeners defined for director service (for standalone keep director_servers empty) Do I need an inet_listener on the front end instance if I want to forward all director queries to the director_servers list? Or is there some other trick to get a proxy front end to talk to a separate director instance(s)? Thanks, Cassidy

1 0

question on lmtp logged message
by Leonardo Rodrigues 07 Jul '14

07 Jul '14

Hi, I have dovecot running for IMAP4/POP3 and also local delivery through LMTP. It's working just fine, absolutely no problem on that, setup is fine. Anyway, sometimes LMTP seems to not be able to deliver some messages and keep them on postfixqueue. And on the next or third try, the message gets delivered successfully. The logged message, however, is not helping me identify what is happening.Example: (error - message was expunged) Jun 25 11:49:39 correio postfix/lmtp[21835]: ADB0A1AC05108: to=<vanilson.parreira(a)domain.com.br>, relay=correio.domain.com.br[private/dovecot-lmtp], conn_use=6, delay=13, delays=0.07/0/0/13, dsn=4.2.0, status=deferred (host correio.domain.com.br[private/dovecot-lmtp] said: 451 4.2.0 <vanilson.parreira(a)domain.com.br> Message was expunged (received-date) (in reply to end of DATA command)) (successfull delivery) Jun 25 12:34:42 correio postfix/lmtp[6411]: ADB0A1AC05108: to=<vanilson.parreira(a)domain.com.br>, relay=correio.domain.com.br[private/dovecot-lmtp], delay=2716, delays=2714/0/0/2.3, dsn=2.0.0, status=sent (250 2.0.0 <vanilson.parreira(a)domain.com.br> U7pOLAHsqlPvMgAAHvf8vg Saved) During these two log entries, absolutely nothing was changed, no configuration, absolutely nothing. Everything is local, i mean, no NFS involved. So, finally, question is: what does the 'message was expunged' message given from LMTP means ??? This is happening quite often but, as i mentionted, sooner or later all messages are getting delivered. It's working despite the delay this is causing. [root@correio log]# dovecot --version 2.2.13 -- Atenciosamente / Sincerily, Leonardo Rodrigues Solutti Tecnologia http://www.solutti.com.br Minha armadilha de SPAM, NÃO mandem email gertrudes(a)solutti.com.br My SPAMTRAP, do not email it

3 4

dbox alternate storage and samba
by Francesco 07 Jul '14

07 Jul '14

Hello everyone, i was just wondering if it's possible to use a samba share as an alternate storage for dbox format. I tried creating a samba share, but when i try to move the email to the alternate storage using doveadm altmove, no errors are printed but no mails are moved. Everything works if the alternate storage is simply another directory on the server hard drive. am i mssing something? thanks in advance.

2 3

rmdir() EBUSY warnings on 2.2.12
by Giel van Schijndel 07 Jul '14

07 Jul '14

Hi, I'm getting a lot of syslog spam on FreeBSD 10 from nfs_flush_file_handle_cache_dir where, apparently it tries to rmdir() the mountpoint of my mail storage. > Jul 4 00:04:44 mailhost dovecot: imap(user(a)example.com): Error: nfs_flush_file_handle_cache_dir: rmdir(/var/mail/vhosts/example.com) failed: Device busy We're using ZFS and have /var, /var/mail, /var/mail/vhosts, /var/mail/vhosts/example.com and /var/mail/vhosts/example.com/user as separate ZFS datasets (to enable per-user and per-domain snapshotting). So apart from NFS flushing being unnecessary (ZFS != NFS) I'm not really sure why using rmdir() for another directory than the Maildir specified in mail_location is a good idea. I.e. given that Maildirs shouldn't ever be emtpy (cur,new,tmp) anyway. -- Met vriendelijke groet, With kind regards, Giel van Schijndel -- > A programmer started to cuss > Because getting to sleep was a fuss > As he lay there in bed Looping 'round in his head > was: while(!asleep()) sheep++; > -- piercings -- http://bash.org/?845468

2 6

Fw: LMTP segfaulting constantly
by Toney Mareo 07 Jul '14

07 Jul '14

Hello I have the dmesg loaded with lmtp crash messages. My OS: No LSB modules are available. Distributor ID: Debian Description: Debian GNU/Linux 7.4 (wheezy) Release: 7.4 Codename: wheezy My postfix version: i postfix 2.9.6-2 amd64 High-performance mail transport agent ii postfix-doc 2.9.6-2 all Documentation for Postfix ii postfix-mysql 2.9.6-2 amd64 MySQL map support for Postfix ii postfix-pcre 2.9.6-2 amd64 PCRE map support for Postfix ii dovecot-antispam 2.0+20130822-21 amd64 Dovecot plugins for training spam filters ii dovecot-core 1:2.2.10-11 amd64 secure POP3/IMAP server - core files ii dovecot-imapd 1:2.2.10-11 amd64 secure POP3/IMAP server - IMAP daemon ii dovecot-lmtpd 1:2.2.10-11 amd64 secure POP3/IMAP server - LMTP server ii dovecot-managesieved 1:2.2.10-11 amd64 secure POP3/IMAP server - ManageSieve server ii dovecot-mysql 1:2.2.10-11 amd64 secure POP3/IMAP server - MySQL support ii dovecot-pop3d 1:2.2.10-11 amd64 secure POP3/IMAP server - POP3 daemon ii dovecot-sieve 1:2.2.10-11 amd64 secure POP3/IMAP server - Sieve filters support Wed Jun 18 13:09:14 2014] lmtp[17453]: segfault at 238 ip 00007f560c3c8ce4 sp 00007fff15b0a440 error 4 in libdovecot-sieve.so.0.0.0[7f560c3a2000+82000] [Wed Jun 18 15:56:22 2014] lmtp[17325]: segfault at 238 ip 00007f2b359aace4 sp 00007fffc3a63e10 error 4 in libdovecot-sieve.so.0.0.0[7f2b35984000+82000] [Mon Jun 23 06:56:44 2014] lmtp[1168]: segfault at 238 ip 00007f5ec34d2ce4 sp 00007fff76545070 error 4 in libdovecot-sieve.so.0.0.0[7f5ec34ac000+82000] [Mon Jun 23 09:47:01 2014] lmtp[20612]: segfault at 238 ip 00007f5fb3830ce4 sp 00007fffea15dc60 error 4 in libdovecot-sieve.so.0.0.0[7f5fb380a000+82000] [Mon Jun 23 10:19:31 2014] lmtp[24839]: segfault at 238 ip 00007f8098ddbce4 sp 00007fffff3f8af0 error 4 in libdovecot-sieve.so.0.0.0[7f8098db5000+82000] [Mon Jun 23 14:45:27 2014] lmtp[28080]: segfault at 238 ip 00007fc2ea901ce4 sp 00007fff528f62d0 error 4 in libdovecot-sieve.so.0.0.0[7fc2ea8db000+82000] [Tue Jun 24 13:58:35 2014] lmtp[31028]: segfault at 238 ip 00007f2c4c493ce4 sp 00007fff5ab22080 error 4 in libdovecot-sieve.so.0.0.0[7f2c4c46d000+82000] [Tue Jun 24 14:50:51 2014] lmtp[9487]: segfault at 238 ip 00007f384ca7ece4 sp 00007fff7a02c520 error 4 in libdovecot-sieve.so.0.0.0[7f384ca58000+82000] [Tue Jun 24 14:55:19 2014] lmtp[10541]: segfault at 238 ip 00007f640e9e3ce4 sp 00007fff9e538060 error 4 in libdovecot-sieve.so.0.0.0[7f640e9bd000+82000] [Tue Jun 24 20:37:47 2014] lmtp[5914]: segfault at 238 ip 00007f7869e11ce4 sp 00007fff962eaaf0 error 4 in libdovecot-sieve.so.0.0.0[7f7869deb000+82000] [Wed Jun 25 09:28:43 2014] lmtp[1855]: segfault at 238 ip 00007ff5c473ace4 sp 00007fff521f08e0 error 4 in libdovecot-sieve.so.0.0.0[7ff5c4714000+82000] [Thu Jun 26 08:11:30 2014] lmtp[4735]: segfault at 238 ip 00007ffe9017fce4 sp 00007fffbe67ed90 error 4 in libdovecot-sieve.so.0.0.0[7ffe90159000+82000] [Thu Jun 26 11:49:57 2014] lmtp[17082]: segfault at 238 ip 00007f5ce9e20ce4 sp 00007fff234152e0 error 4 in libdovecot-sieve.so.0.0.0[7f5ce9dfa000+82000] [Thu Jun 26 15:34:27 2014] lmtp[682]: segfault at 238 ip 00007ffa9c191ce4 sp 00007fff0c6ee870 error 4 in libdovecot-sieve.so.0.0.0[7ffa9c16b000+82000] [Thu Jun 26 16:43:17 2014] lmtp[15603]: segfault at 238 ip 00007f3994435ce4 sp 00007fff32b93500 error 4 in libdovecot-sieve.so.0.0.0[7f399440f000+82000] [Mon Jun 30 10:19:43 2014] lmtp[473]: segfault at 238 ip 00007fe0e4706ce4 sp 00007fff97b648b0 error 4 in libdovecot-sieve.so.0.0.0[7fe0e46e0000+82000] [Tue Jul 1 07:06:12 2014] lmtp[28516]: segfault at 238 ip 00007f62fbcfcce4 sp 00007fffdd76fd80 error 4 in libdovecot-sieve.so.0.0.0[7f62fbcd6000+82000] [Tue Jul 1 12:43:01 2014] lmtp[32438]: segfault at 238 ip 00007fc4bdda6ce4 sp 00007fff6f527820 error 4 in libdovecot-sieve.so.0.0.0[7fc4bdd80000+82000] [Wed Jul 2 13:21:20 2014] lmtp[28053]: segfault at 238 ip 00007f112a1d4ce4 sp 00007fff0706e350 error 4 in libdovecot-sieve.so.0.0.0[7f112a1ae000+82000] [Wed Jul 2 16:19:42 2014] lmtp[30986]: segfault at 238 ip 00007f7e0ad15ce4 sp 00007fff68c1b710 error 4 in libdovecot-sieve.so.0.0.0[7f7e0acef000+82000] [Wed Jul 2 16:36:18 2014] lmtp[2223]: segfault at 238 ip 00007f69ecc40ce4 sp 00007fff2e8783d0 error 4 in libdovecot-sieve.so.0.0.0[7f69ecc1a000+82000] [Thu Jul 3 09:14:59 2014] lmtp[30956]: segfault at 238 ip 00007f86f1dccce4 sp 00007fffae17fc70 error 4 in libdovecot-sieve.so.0.0.0[7f86f1da6000+82000] [Thu Jul 3 09:38:18 2014] lmtp[3631]: segfault at 238 ip 00007fae2d901ce4 sp 00007fff3f7a9270 error 4 in libdovecot-sieve.so.0.0.0[7fae2d8db000+82000] [Thu Jul 3 10:18:09 2014] lmtp[12435]: segfault at 238 ip 00007f49343bfce4 sp 00007fff29e64670 error 4 in libdovecot-sieve.so.0.0.0[7f4934399000+82000] [Fri Jul 4 11:34:27 2014] lmtp[22575]: segfault at 238 ip 00007f9e52493ce4 sp 00007fff765e9fe0 error 4 in libdovecot-sieve.so.0.0.0[7f9e5246d000+82000] [Fri Jul 4 11:49:37 2014] lmtp[26586]: segfault at 238 ip 00007f387638fce4 sp 00007fff624bf970 error 4 in libdovecot-sieve.so.0.0.0[7f3876369000+82000] [Sun Jul 6 13:49:15 2014] lmtp[15451]: segfault at 238 ip 00007fa435865ce4 sp 00007fff436a28f0 error 4 in libdovecot-sieve.so.0.0.0[7fa43583f000+82000] The server itself is working, but it's annoying seeing these messages in the logs. Sometimes when it's overloaded it refuses to authenticate users (i don't think this has to do with it). Any ideas what causes it? I have my debian wheezy updated to the latest. Thanks

1 0

[help]how to refresh pigeonhole sieve vacation days data
by wuwei 07 Jul '14

07 Jul '14

hello everybody: I'm using dovecot with pigeonhole sieve Plugin.everything goes well.but when i use vacation module,i found a problem: I write a sieve with a vacation module in, and put it to the sieve location ,used *PUTSCRIPT* comand with *ManageSieve* and *set it ACTIVE*.It's work!I use A email send mail to him and receive an auto-reply.then ,I modify the sieve only change the days num from 7 to 1.Then I use B email send mail to him and receive an auto-reply too. The next day,I send two mails too.But only B receive the auto-reply.I thought that I would receive two mails too. so i think that A email's data in sieve maybe still 7 days.Then I restart the dovecot,but the problem still exsit. Is anyone know how to refresh the data or is their anyother way to solve the problem? thank you! -- View this message in context: http://dovecot.2317879.n4.nabble.com/help-how-to-refresh-pigeonhole-sieve-v… Sent from the Dovecot mailing list archive at Nabble.com.

1 0

Assertion virtual-storage.c
by Thomas Leuxner 06 Jul '14

06 Jul '14

# 2.2.13 (dffc6dd2be0d): /etc/dovecot/dovecot.conf # OS: Linux 3.14-0.bpo.1-amd64 x86_64 Debian 7.5 Latest HG has issues with Virtual Mailboxes. I was unable to get it dump a core. Jul 5 07:46:51 nihlus dovecot: imap(tlx(a)leuxner.net): Panic: file virtual-storage.c: line 326 (virtual_backend_box_open): assertion failed: (!bbox->box->opened) Jul 5 07:46:51 nihlus dovecot: imap(tlx(a)leuxner.net): Error: Raw backtrace: /usr/lib/dovecot/libdovecot.so.0(+0x6ce3f) [0x7fe9e733ee3f] -> /usr/lib/dovecot/libdovecot.so.0(+0x6ce9e) [0x7fe9e733ee9e] -> /usr/lib/dovecot/libdovecot.so.0(i_fatal+0) [0x7fe9e72f796e] -> /usr/lib/dovecot/modules/lib20_virtual_plugin.so(+0xa1d4) [0x7fe9e5ae01d4] -> /usr/lib/dovecot/modules/lib20_virtual_plugin.so(+0xa237) [0x7fe9e5ae0237] -> dovecot/imap(cmd_idle+0x45) [0x410595] -> dovecot/imap(command_exec+0x3c) [0x41849c] -> dovecot/imap() [0x4174e0] -> dovecot/imap() [0x417574] -> dovecot/imap(client_handle_input+0x115) [0x417855] -> dovecot/imap(client_input+0x72) [0x417c02] -> /usr/lib/dovecot/libdovecot.so.0(io_loop_call_io+0x4e) [0x7fe9e735002e] -> /usr/lib/dovecot/libdovecot.so.0(io_loop_handler_run_internal+0xd7) [0x7fe9e7351057] -> /usr/lib/dovecot/libdovecot.so.0(io_loop_handler_run+0x9) [0x7fe9e73500b9] -> /usr/lib/dovecot/libdovecot.so.0(io_loop_run+0x38) [0x7fe9e7350138] -> /usr/lib/dovecot/libdovecot.so.0(master_service_run+0x13) [0x7fe9e72fcdd3] -> dovecot/imap(main+0x298) [0x40c238] -> /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xfd) [0x7fe9e6f65ead] -> dovecot/imap() [0x40c39d] Jul 5 07:46:51 nihlus dovecot: imap(tlx(a)leuxner.net): Fatal: master: service(imap): child 37155 killed with signal 6 (core not dumped) Regards Thomas

2 2

Re: number of users
by Andreas Tauscher 05 Jul '14

05 Jul '14

> No, this will not do it. Any user coming from an HTTP connection will > appear very briefly on the Dovecot server and is unlikely to appear > on the "doveadm who" list. This is a result of the HTTP > connectionless protocol, nothing to do with Dovecot (ie > connects-carries a brief data > transmission/commands-disconnects-connects again after a > predetermined time). The connection time is very brief, even though > you can probably catch a few if you insist on typing doveadm who. If like from Nick Edwards suggested a imapproxy is used it will. imapproxy is keeping the connection open. doveadm who will give you a list of all actual logged in users. Without imapproxy you are right: Every page reload will cause a login and a logoff. I get some webmail users by "accident". To get an overview how many users used roudcube within a period check the database table users/last_login. Andreas

1 0

Disable user login for a few seconds ?
by Florent B 05 Jul '14

05 Jul '14

Hi all, I need to move e-mail files of my users from one file system to another. I use a MySQL backend on Dovecot 2.1. How can I do to disable temporary an account while moving files ? I have an "active" bit in my DB, but if I set it to 0, my MX servers are bouncing mails. How can I do to disable temporary an account ? In fact, I need dovecot not to do any change on files... I think I could change account password for a while, and reset it after migration, but it will disturb connected users. Maybe I could chmod 0 files on my old file system during migration ? How Dovecot will act ? Thank you :) Flo

4 4