- dovecot - dovecot.org

HTTP API vs. doveadm
by Felipe Gasper 24 Aug '18

24 Aug '18

To follow up on an earlier thread: is there any functional advantage of the new HTTP API over the doveadm API? How does the HTTP API deal with a query like “mailbox list”, which (as best I can tell) could return non-UTF8 text? -FG

2 1

Is the Doveadm HTTP API considered stable for production use?
by James Beck 23 Aug '18

23 Aug '18

Hi, I'm running 2.2.34 in production (installed from Debian stretch backports) and want to rework some scripts. Can the HTTP API be considered stable in 2.2.34 please? The wiki says it is "considered experimental in v2.2.22" so I thought I'd check before writing API calls rather than a wrapper around doveadm! -- James Beck Andrews & Arnold Ltd Tel: 03333 400999

5 8

allow_nets based on RBL
by Tom Sommer 23 Aug '18

23 Aug '18

This was brought up in 2014, and left without conclusion, so I thought it would be time to bump it :) I would love a way to do allow_nets based on an RBL check, could this be added to the feature-list? https://wiki2.dovecot.org/PasswordDatabase/ExtraFields/AllowNets Thanks -- Tom

2 1

lmtp Panic Buffer write out of range
by Gabriele Nencioni 23 Aug '18

23 Aug '18

Hi all, as described here: https://www.dovecot.org/pipermail/dovecot/2018-July/112173.html we are experiencing the same error on dovecot version 2.3.2.1 while it never occurs on an old version as 2.2.15 It followings the error logs: On an upgraded dovecot backend: Aug 21 12:03:51 backend20 dovecot: lmtp(test1(a)internalinboundcm.eu)<SONkAYfje1veGgAAu8+/vw>: Panic: Buffer write out of range (789 + 18446744073709551606) /var/log/dovecot_lmtp_20180821.12.log-Aug 21 12:03:51 monti-backend20 dovecot: lmtp(test1(a)internalinboundcm.eu)<SONkAYfje1veGgAAu8+/vw>: Error: Raw backtrace: /usr/lib/dovecot/libdovecot.so.0(+0xcb851) [0x7f78e6f0a851] -> /usr/lib/dovecot/libdovecot.so.0(+0xcb8e9) [0x7f78e6f0a8e9] -> /usr/lib/dovecot/libdovecot.so.0(i_fatal+0) [0x7f78e6e78851] -> /usr/lib/dovecot/libdovecot.so.0(buffer_write+0x131) [0x7f78e6f04871] -> /usr/lib/dovecot/libdovecot-sieve.so.0(rfc2822_header_append+0xcf) [0x7f78e531613f] -> /usr/lib/dovecot/libdovecot-sieve.so.0(+0x7d35b) [0x7f78e531335b] -> /usr/lib/dovecot/libdovecot-sieve.so.0(edit_mail_header_add+0x1b) [0x7f78e531525b] -> /usr/lib/dovecot/libdovecot-sieve.so.0(+0x73014) [0x7f78e5309014] -> /usr/lib/dovecot/libdovecot-sieve.so.0(sieve_interpreter_continue+0x81) [0x7f78e52d2a11] -> /usr/lib/dovecot/libdovecot-sieve.so.0(sieve_interpreter_run+0x2b) [0x7f78e52d2c7b] -> /usr/lib/dovecot/libdovecot-sieve.so.0(+0x51000) [0x7f78e52e7000] -> /usr/lib/dovecot/libdovecot-sieve.so.0(sieve_multiscript_run+0x3a) [0x7f78e52e7eba] -> /usr/lib/dovecot/modules/lib90_sieve_plugin.so(+0x38f0) [0x7f78e554e8f0] -> /usr/lib/dovecot/libdovecot-lda.so.0(mail_deliver+0xf0) [0x7f78e750da30] -> dovecot/lmtp [81.88.49.172 DATA](lmtp_local_data+0x4f1) [0x5582554451d1] -> dovecot/lmtp [81.88.49.172 DATA](cmd_data_continue+0x243) [0x558255443ee3] -> /usr/lib/dovecot/libdovecot.so.0(+0x4c402) [0x7f78e6e8b402] -> /usr/lib/dovecot/libdovecot.so.0(io_loop_call_io+0x69) [0x7f78e6f228a9] -> /usr/lib/dovecot/libdovecot.so.0(io_loop_handler_run_internal+0x109) [0x7f78e6f24199] -> /usr/lib/dovecot/libdovecot.so.0(io_loop_handler_run+0x52) [0x7f78e6f229b2] -> /usr/lib/dovecot/libdovecot.so.0(io_loop_run+0x38) [0x7f78e6f22bc8] -> /usr/lib/dovecot/libdovecot.so.0(master_service_run+0x13) [0x7f78e6e9cff3] -> dovecot/lmtp [81.88.49.172 DATA](main+0x240) [0x558255442f70] -> /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf1) [0x7f78e6ac02e1] -> dovecot/lmtp [81.88.49.172 DATA](_start+0x2a) [0x5582554430ba] while once the user test1(a)internalinboundcm.eu has been moved on an old dovecot backend, with the "doveadm director move" command the message has been properly delivered at next retry attempt: Aug 21 12:32:31 backend01 dovecot: lmtp(test1(a)internalinboundcm.eu)<Zv32Lzbqe1tSLQAAF5Yx/g>: copy from <lmtp DATA>: box=Spam, uid=1, msgid=<ceb54c05728586392119651424858d85(a)l2.losclientes3.com> Aug 21 12:32:31 backend01 dovecot: lmtp(test1(a)internalinboundcm.eu)<Zv32Lzbqe1tSLQAAF5Yx/g>: Zv32Lzbqe1tSLQAAF5Yx/g: sieve: msgid=<ceb54c05728586392119651424858d85(a)l2.losclientes3.com>: stored mail into mailbox 'Spam' As you can see the message was detected as spam Let me know if you need more info or details (such us the original message or the sieve filter that trigger the problem) Thanks in advance Regards -- Gabriele Nencioni System Administrator eml gabriele.nencioni(a)register.it

2 3

Re: Possible to set quota in size and number of messages
by J Doe 21 Aug '18

21 Aug '18

Hi Reio and list, It appears placing it at the end of the quota rules and then manually recalculating the quotas via doveadm has solved the problem! Configuration: plugin { quota = dict:User quota::file:%h/dovecot-quota quota_rule = *:storage=10G quota_rule1 = INBOX/Inbox:storage=+30M quota_rule2 = INBOX/Trash:storage=+30M quota_rule3 = INBOX/Deleted Messages:storage=+30M quota_rule4 = *:messages=100000 } Recalculate quotas: $ sudo doveadm quota recalc -u <account> Restart Dovecot and Apple Mail. Thanks, - J

1 0

Possible to set quota in size and number of messages
by J Doe 21 Aug '18

21 Aug '18

Hello, I am using Dovecot 2.2.33.2 and am attempting to set quotas on mailboxes. I currently have the following: plugin { quota = dict:User quota::file:%h/dovecot-quota quota_rule = *:storage=10G quota_rule1 = INBOX/Inbox:storage=+30M quota_rule2 = INBOX/Trash:storage=+30M quota_rule3 = INBOX/Deleted Messages:storage=+30M } If I check the quota via Apple Mail version 11.5 (3445.9.1) on Mac OS X 10.13.6 by selecting “Get Account Info” it displays the correct quota size in bytes but also states “0 messages” (in that it states I can store 0 messages). Is there a way to signify to an IMAP client that the quota should be _EITHER_ 10 GB or 100,000 messages, just so a user doesn’t see that they can store a total of 0 messages ? My initial thought was to alter the previous quota settings by adding: quota_rule4 = INBOX:messages=100000 … and restarting Dovecot and Apple Mail, but upon checking the quota’s on the user’s account, I still see that a user can store 10 GB or 0 messages. Thanks, - J

1 0

lazy expunge folder delete bug
by Jan-Pieter Cornet 21 Aug '18

21 Aug '18

There's a bug in "folder delete" for lazy expunge, type "1 namespace", as descibed on https://wiki2.dovecot.org/Plugins/Lazyexpunge When trying to delete a mailbox that still has messages in it, but that has no EXPUNGED/<name> counterpart, the process hangs after the imap "DELETE" command, and the following appears in the log file after a 60s timeout: Jun 22 15:48:15 userimap6 dovecot: imap(xtra30): Error: Couldn't create mailbox list lock /var/mail/.8d1/index/4/03/xtra30/index/mailboxes.lock: file_create_locked(/var/mail/.8d1/index/4/03/xtra30/index/mailboxes.lock) failed: flock(/var/mail/.8d1/index/4/03/xtra30/index/mailboxes.lock, write-lock) failed: Timed out after 60 seconds (BUG: lock is held by our own process) Jun 22 15:48:15 userimap6 dovecot: imap(xtra30): Error: lazy_expunge: Couldn't open expunge mailbox: Failed to create mailbox EXPUNGED/Test: Internal error occurred. Refer to server log for more information. [2018-06-22 15:47:15] Jun 22 15:48:15 userimap6 dovecot: imap(xtra30): Error: Lazy-expunge transaction failed: Internal error occurred. Refer to server log for more information. [2018-06-22 15:47:15] very quick summage of settings: ----8<----- mail_plugins = $mail_plugins lazy_expunge mail_location = mdbox:/var/mail/.8d1/mail/4/03/xtra30:INDEX=/var/mail/.8d1/index/4/03/xtra30/index ### from userdb namespace { inbox = yes list = yes prefix = separator = / } namespace expunged { hidden = yes inbox = no list = no prefix = EXPUNGED/ separator = / location = mdbox:/var/mail/.8d1/mail/4/03/xtra30:INDEX=/var/mail/.8d1/index/4/03/xtra30/index:MAILBOXDIR=expunged:LISTINDEX=expunged.list.index:SUBSCRIPTIONS=expunged.subscriptions ### also from userdb } plugin { lazy_expunge = EXPUNGED/ } ----8<----- (full doveconf available on request) State of the account before delete: mailbox "Test" exists, 20 messages in it: x LIST * * * LIST (\HasNoChildren \UnMarked \Trash) "/" Trash * LIST (\HasNoChildren \Marked) "/" Test * LIST (\HasNoChildren \UnMarked \Junk) "/" Spam * LIST (\HasNoChildren \UnMarked) "/" Sent * LIST (\HasNoChildren \UnMarked) "/" Drafts * LIST (\HasNoChildren) "/" INBOX x OK List completed (0.009 + 0.000 + 0.008 secs). x LIST EXPUNGED/* * x OK List completed (0.001 + 0.000 secs). x DELETE Test x NO [SERVERBUG] Internal error occurred. Refer to server log for more information. [2018-06-22 15:47:15] (60.064 + 0.000 + 60.063 secs). ... resulting in the above log lines. However, if you first delete one message (causing EXPUNGED/Test to be created), and then remove the folder, it works fine: x SELECT Test * FLAGS (\Answered \Flagged \Deleted \Seen \Draft) * OK [PERMANENTFLAGS (\Answered \Flagged \Deleted \Seen \Draft \*)] Flags permitted. * 20 EXISTS * 2 RECENT * OK [UNSEEN 1] First unseen. * OK [UIDVALIDITY 1352910749] UIDs valid * OK [UIDNEXT 21] Predicted next UID * OK [HIGHESTMODSEQ 21] Highest x OK [READ-WRITE] Select completed (0.010 + 0.000 + 0.009 secs). x FETCH 1:* (FLAGS) * 1 FETCH (FLAGS ()) [...boring...] * 20 FETCH (FLAGS (\Recent)) x OK Fetch completed (0.002 + 0.000 + 0.001 secs). x STORE 1 +FLAGS (\Deleted) * 1 FETCH (FLAGS (\Deleted)) x OK Store completed (0.011 + 0.000 + 0.010 secs). x EXPUNGE * 1 EXPUNGE x OK Expunge completed (0.131 + 0.000 + 0.130 secs). X LIST * * * LIST (\HasNoChildren \UnMarked \Trash) "/" Trash * LIST (\HasNoChildren \UnMarked) "/" Test * LIST (\HasNoChildren \UnMarked \Junk) "/" Spam * LIST (\HasNoChildren \UnMarked) "/" Sent * LIST (\HasNoChildren \UnMarked) "/" Drafts * LIST (\HasNoChildren) "/" INBOX X OK List completed (0.005 + 0.000 + 0.004 secs). x LIST EXPUNGED/* * * LIST (\HasNoChildren \Marked) "/" EXPUNGED/Test x OK List completed (0.002 + 0.000 + 0.001 secs). x SELECT INBOX * OK [CLOSED] Previous mailbox closed. * FLAGS (\Answered \Flagged \Deleted \Seen \Draft) [...INBOX opening msgs...] x OK [READ-WRITE] Select completed (0.005 + 0.000 + 0.004 secs). x DELETE Test x OK Delete completed (0.129 + 0.000 + 0.128 secs). x LIST * * * LIST (\HasNoChildren \UnMarked \Trash) "/" Trash * LIST (\HasNoChildren \UnMarked \Junk) "/" Spam * LIST (\HasNoChildren \UnMarked) "/" Sent * LIST (\HasNoChildren \UnMarked) "/" Drafts * LIST (\HasNoChildren) "/" INBOX x OK List completed (0.005 + 0.000 + 0.004 secs). x LIST EXPUNGED/* * * LIST (\HasNoChildren \Marked) "/" EXPUNGED/Test x OK List completed (0.003 + 0.000 + 0.002 secs). Seems the delete opeation locks the mailbox list, and then the expunge create hits the same lock. Is this something we can fix by changing settings? Eg use another location for the expunge lock? -- Jan-Pieter Cornet <johnpc(a)xs4all.net> Systeembeheer XS4ALL Internet bv www.xs4all.nl

1 1

iOS Mail Client Not Copying to Sent Folder
by Asai 21 Aug '18

21 Aug '18

Greetings, We have many iOS clients, and some report that they’re having problems with iOS mail copying messages to their Sent folder. This happens 2 - 3 times per day, and the phone just says Sending and the client never knows if it sends. Client says apparently email DOES go out, but there is no recording of it. This is usually on our LAN. We’re not sure if this is solved via Postfix or Dovecot, or maybe something else, so I’m posting this message on both forums as a start. Can anyone point us in the right direction here? Running Postfix 2.11.1 and Dovecot ee 2.2.18.2 see doveconf -n below: # 2.2.18.2 (866bffbafde7): /etc/dovecot/dovecot.conf # Pigeonhole version 0.4.8 (73d4f6cd676a) # OS: Linux 2.6.32-358.el6.x86_64 x86_64 CentOS release 6.4 (Final) ext4 auth_mechanisms = plain login auth_verbose = yes auth_verbose_passwords = plain deliver_log_format = msgid=%m: from=%f: %$ lock_method = dotlock mail_home = /vmail/%d/%n/home mail_location = maildir:/vmail/%d/%n mail_plugins = " notify" managesieve_notify_capability = mailto managesieve_sieve_capability = fileinto reject envelope encoded-character vacation subaddress comparator-i;ascii-numeric relational regex imap4flags copy include variables body enotify environment mailbox date index ihave duplicate namespace { inbox = yes location = prefix = separator = / type = private } passdb { args = /etc/dovecot/dovecot-mysql.conf driver = sql } plugin { mail_log_events = delete expunge copy mailbox_delete flag_change append mailbox_rename mail_log_fields = box msgid from flags sieve = /vmail/%d/%n/home/.dovecot.sieve sieve_before = /etc/dovecot/sieve/junk-spam.sieve sieve_default = /etc/dovecot/sieve/junk-spam.sieve sieve_dir = /vmail/%d/%n/home/sieve } protocols = imap lmtp sieve service auth { unix_listener /var/spool/postfix/private/auth { group = postfix mode = 0666 user = postfix } unix_listener auth-master { group = vmail mode = 0660 user = vmail } unix_listener auth-userdb { mode = 0660 user = vmail } user = root } service imap-login { process_limit = 200 process_min_avail = 16 service_count = 0 vsz_limit = 0 } service managesieve-login { inet_listener sieve { port = 4190 } process_min_avail = 0 service_count = 1 vsz_limit = 64 M } ssl_cert = </etc/postfix/ssl/triata.pem ssl_key = </etc/postfix/ssl/triata.key userdb { args = /etc/dovecot/dovecot-mysql.conf driver = sql } protocol lda { hostname = domain.net mail_plugin_dir = /usr/lib64/dovecot/ mail_plugins = sieve postmaster_address = postmaster(a)domain.net } protocol imap { imap_idle_notify_interval = 24 mins mail_max_userip_connections = 60 } protocol sieve { mail_max_userip_connections = 10 managesieve_implementation_string = Dovecot Pigeonhole managesieve_logout_format = bytes=%i/%o managesieve_max_line_length = 65536 }

1 0

Strange problem on lmtp with dovecot 2.3.2.1
by davide marchi 20 Aug '18

20 Aug '18

Hi to all i have compiled and installed dovecot 2.3.2.1 on debian 6 with success but now i'm facing a strange problem on communication from postfix installed on server A with dovecot installed on server B (which version is 2.3.2.1). All incoming emails delivered from esternal from server A to server B with LMTP is rejected with this message error "smtpgw postfix/lmtp[1772]: 6194CA00B3: to=<xxxxxxxxxxxx(a)mail.cgilfe.it>, relay=192.168.10.50[192.168.10.50]:24, delay=0.02, delays=0/0.01/0.01/0, dsn=5.5.4, status=undeliverable (host 192.168.10.50[192.168.10.50] said: 501 5.5.4 Invalid FROM: Missing domain (in reply to MAIL FROM command))" I cant identify what is the problem , for now i have bypassed lmtp using smtp but my goal is returning to lmtp that i have used since old versions of dovecot.

2 1

systemd also listening on imaps port
by Kai Schaetzl 20 Aug '18

20 Aug '18

I noticed that most of the time with dovecot 2.3 I get this output, showing systemd listening on the port as well: lsof -i TCP:993 systemd 1 root 38u IPv4 3388360574 0t0 TCP *:imaps (LISTEN) systemd 1 root 41u IPv6 3388360575 0t0 TCP *:imaps (LISTEN) dovecot 11607 root 5u IPv4 3388360574 0t0 TCP *:imaps (LISTEN) dovecot 11607 root 6r IPv6 3388360575 0t0 TCP *:imaps (LISTEN) Sometimes I don't get the extra systemd lines, but I can't reproduce it now. It's only for port 143 and 993. Other ports show only two dovecot processes. Operation seems to be normal. Is this how it is supposed to be with dovecot 2.3? This is Ubuntu 16.04.

2 2