- dovecot - dovecot.org

Doveadm and iterate query
by Bogusław Juza 25 Jul '26

25 Jul '26

Hi! I have a problem with Dovecot on Debian 13: Dovecot 2.4.1, Debian Trixie > doveadm mailbox path -A INBOX >/dev/null doveadm(email@domain): Error: auth-master: userdb list: User listing returned failure doveadm: Error: cmd mailbox path: Failed to iterate through some users userdb sql { iterate_query = SELECT `email` AS `user` FROM `users` } In syslog file: dovecot: auth: Error: auth-worker: Aborted LIST request for *: Shutting down I have over 30k users in sql DB, doveadm with -A option is dying few second after start with this error, printing about 10% of users. I need much more time (FE command "doveadm expunge -A mailbox Spam savedbefore 3w" could run over 30 minutes). What I'm doing wrong? Bogusław Juza

5 12

Reporting a Patch needed: Fix-anvil-busy-loop-on-EPOLLHUP-when-a-multiplex-cha.patch
by David Lanning 24 Jul '26

24 Jul '26

The attached patch spec is what we had to deploy to cover the matter. The attached patch spec is what we had to deploy to cover the matter.

2 1

Strange characters in log file
by Doug Hardie 21 Jul '26

21 Jul '26

I am occasionally finding some strange characters in the log file. I have captured on entry that vi cannot handle. It simply won't go past this record. I suspect the issue is around 0347 in the record below. Any ideas how to prevent this from happening as it makes it difficult to find specific records in the log. vi is unable to delete the line with this issue. mail# head -1 imaplog.test | od -t x1c 0000000 4a 75 6c 20 31 37 20 30 32 3a 33 34 3a 35 35 20 J u l 1 7 0 2 : 3 4 : 5 5 0000020 6d 61 69 6c 20 64 6f 76 65 63 6f 74 5b 35 32 39 m a i l d o v e c o t [ 5 2 9 0000040 30 5d 3a 20 69 6d 61 70 28 64 6f 75 67 29 5b 31 0 ] : i m a p ( d o u g ) [ 1 0000060 30 2e 30 2e 31 2e 32 33 32 5d 3c 4e 74 79 78 52 0 . 0 . 1 . 2 3 2 ] < N t y x R 0000100 4d 74 57 4e 64 38 4b 41 41 48 6f 3e 3a 20 64 65 M t W N d 8 K A A H o > : d e 0000120 6c 65 74 65 3a 20 62 6f 78 3d 44 65 6c 65 74 65 l e t e : b o x = D e l e t e 0000140 64 20 4d 65 73 73 61 67 65 73 2c 20 75 69 64 3d d M e s s a g e s , u i d = 0000160 35 38 32 32 31 33 2c 20 6d 73 67 69 64 3d 3c 39 5 8 2 2 1 3 , m s g i d = < 9 0000200 31 30 35 30 36 39 31 2e 31 34 37 36 39 34 36 2e 1 0 5 0 6 9 1 . 1 4 7 6 9 4 6 . 0000220 31 37 38 33 35 38 37 39 33 35 37 36 36 40 6d 61 1 7 8 3 5 8 7 9 3 5 7 6 6 @ m a 0000240 72 6b 65 74 6f 2d 70 6c 61 74 66 6f 72 6d 2d 2d r k e t o - p l a t f o r m - - 0000260 65 6d 61 69 6c 2d 73 65 6e 64 65 72 2d 70 72 6f e m a i l - s e n d e r - p r o 0000300 64 2d 62 61 74 63 68 2d 63 6c 75 2e 2e 2e 2c 20 d - b a t c h - c l u . . . , 0000320 73 69 7a 65 3d 35 35 39 39 30 2c 20 66 72 6f 6d s i z e = 5 5 9 9 0 , f r o m 0000340 3d 57 4f 52 4c 44 e2 4d 2d 5e 40 4d 2d 5e 59 73 = W O R L D 342 M - ^ @ M - ^ Y s 0000360 20 54 69 6d 20 4c 61 6d 65 72 20 3c 65 64 69 74 T i m L a m e r < e d i t 0000400 6f 72 40 77 6e 67 2e 6f 72 67 3e 0a o r @ w n g . o r g > \n 0000414 -- Doug I am occasionally finding some strange characters in the log file. I have captured on entry that vi cannot handle. It simply won't go past this record. I suspect the issue is around 0347 in the record below. Any ideas how to prevent this from happening as it makes it difficult to find specific records in the log. vi is unable to delete the line with this issue. mail# head -1 imaplog.test | od -t x1c 0000000 4a 75 6c 20 31 37 20 30 32 3a 33 34 3a 35 35 20 J u l 1 7 0 2 : 3 4 : 5 5 0000020 6d 61 69 6c 20 64 6f 76 65 63 6f 74 5b 35 32 39 m a i l d o v e c o t [ 5 2 9 0000040 30 5d 3a 20 69 6d 61 70 28 64 6f 75 67 29 5b 31 0 ] : i m a p ( d o u g ) [ 1 0000060 30 2e 30 2e 31 2e 32 33 32 5d 3c 4e 74 79 78 52 0 . 0 . 1 . 2 3 2 ] < N t y x R 0000100 4d 74 57 4e 64 38 4b 41 41 48 6f 3e 3a 20 64 65 M t W N d 8 K A A H o > : d e 0000120 6c 65 74 65 3a 20 62 6f 78 3d 44 65 6c 65 74 65 l e t e : b o x = D e l e t e 0000140 64 20 4d 65 73 73 61 67 65 73 2c 20 75 69 64 3d d M e s s a g e s , u i d = 0000160 35 38 32 32 31 33 2c 20 6d 73 67 69 64 3d 3c 39 5 8 2 2 1 3 , m s g i d = < 9 0000200 31 30 35 30 36 39 31 2e 31 34 37 36 39 34 36 2e 1 0 5 0 6 9 1 . 1 4 7 6 9 4 6 . 0000220 31 37 38 33 35 38 37 39 33 35 37 36 36 40 6d 61 1 7 8 3 5 8 7 9 3 5 7 6 6 @ m a 0000240 72 6b 65 74 6f 2d 70 6c 61 74 66 6f 72 6d 2d 2d r k e t o - p l a t f o r m - - 0000260 65 6d 61 69 6c 2d 73 65 6e 64 65 72 2d 70 72 6f e m a i l - s e n d e r - p r o 0000300 64 2d 62 61 74 63 68 2d 63 6c 75 2e 2e 2e 2c 20 d - b a t c h - c l u . . . , 0000320 73 69 7a 65 3d 35 35 39 39 30 2c 20 66 72 6f 6d s i z e = 5 5 9 9 0 , f r o m 0000340 3d 57 4f 52 4c 44 e2 4d 2d 5e 40 4d 2d 5e 59 73 = W O R L D 342 M - ^ @ M - ^ Y s 0000360 20 54 69 6d 20 4c 61 6d 65 72 20 3c 65 64 69 74 T i m L a m e r < e d i t 0000400 6f 72 40 77 6e 67 2e 6f 72 67 3e 0a o r @ w n g . o r g > \n 0000414 -- Doug

2 4

AW: AW: [EXT] Re: doveadm backup / purge problems since migration from 2.3 to 2.4
by MK 21 Jul '26

21 Jul '26

>>> On 09/07/2026 13:30 EEST MK <dovecot-ml(a)mk.de> wrote: >>> >>> >>> >> On 09/07/2026 11:58 EEST MK via dovecot <dovecot(a)dovecot.org>> wrote: >>> >> >>> >> >>> >> Hi everyone, >>> >> >>> >> I have a two-server setup where all user data from the actively used server is periodically transferred to the second server using "doveadm >backup". For this, I currently use a script that caches a list of all users (containing only the usernames) via "doveadm user '*' " and then >executes the command "doveadm backup -u [username] tcp:1.2.3.4" (target host) for each user. I use a similarly structured script for a daily >"doveadm purge", where a purge is performed per user instead of a backup. >>> >> >>> >> Under version 2.3, I always ran both commands using the "-A" flag (doveadm purge -A and doveadm backup -A tcp:1.2.3.4). Since migrating >from Dovecot 2.3 to 2.4, this no longer works. When running the purge, the process always aborts after a certain number of users with the >following error message: >>> >> >>> >> doveadm(john(a)doe.de): Error: auth-master: userdb list: User listing returned failure >>> >> >>> >>> >>> >Try adding >>> > >>> >log_debug=category=auth >>> > >>> >and see if it can shed light one this. >>> >>> I'll add it to the config and get back next time the problem occurs. >>> >>> >> It's not the same User allways but after nearly the same time or amount of users the script has processed. >>> >> Since I couldn't identify the exact cause of the error and suspect it might be related to changes in how the SQL iterate query is >integrated in the 2.4 configuration (I use MySQL as the backend) I switched to the script described above. >>> >> >>> >> Then since is used the new script I have a non-reproducible issue. After the initial data transfer (when the target is still empty) the >amount of data is equal to the source. Then after one of the next transfers the data is twice as high. It is random as to which subsequent >transfer triggers this. When comparing the quotas on the source and target server, the output is identical on both. Even running `quota recalc` >makes no difference. If I check the user directories, I can see that the `storage` folder contains nearly double the number of files and used >disk space per user. >>> >> >>> >> My second problem is that when I run the purge script on the second server (target), it sometimes deletes all emails for all users, also >non-reproducible. Afterwards, all files under "storage" are gone, and the quota shows 0 emails. I did not have this issue with the exactly same >purge script on the first server which is the source for the backup. >>> >> >>> >>> >What does your purge script do? >>> >>> I have attached it so you can take a look. >>> Is it a known issue that the -A option doesn't work as well as it did in 2.3? >>> >>> >> >>And you are sure it's the purge doing it, and not running doveadm sync? > >If you mean delete everything, yes. There is nothing else running there. No other scripts, nothing other is accessing Dovecot. >A few times I have seen it live and the other times I can see in my logging that when the script starts the space is sinking and sinking near to >zero until the script ends. I'm logging the disk space every minute. > >It’s clear that these are two different problems right? (purge deletes everything and backup writes to mutch data) > >What do I have to search for in the logfiles if the problem oncours? What is interesting for you? The debug makes realy mutch log entries. Update to this. Since I activated the debugging on 09/07/2026 the problem did not occur again, really strange. I can't explain it sice I haven't changed anything other than enabling debugging. Oliver

1 0

Sieve vnd.dovecot.config extension does not work when mail is delivered to lmtp inet_listener but works when delivered to lmtp unix_listener
by DK 20 Jul '26

20 Jul '26

Hello. I have a strange problem with sieve scripts vnd.dovecot.config extension processing on newest dovecot 2.3 and 2.4. When I deliver mail via lmtp unix_listener sieve script can get sieve_env_* variable with no problem. However when I deliver mail via lmtp inet_listener sieve script can not get the same sieve_env_* variable. I deliver mails from postfix to lmtp inet_listener directly (not via haproxy). The same postfix is used when testing lmtp unix_listener mail delivery which works. Any clues what could be the cause of such problem? DK

2 3

How to use sieve_env_ variables in 2.4.4?
by inaugurator＠ya.ru 16 Jul '26

16 Jul '26

Hi Dear Devs! In the previous version 2.3.x I used variables with the **sieve_env_** prefix to access them later in Sieve scripts, for example `sieve_env_hello = hello` changed to `"${env.vnd.dovecot.config.hello}"`. But now I am getting an error ``` doveconf: Fatal: Error in configuration file /etc/dovecot/dovecot.conf line 11: sieve_env_hello: Unknown setting: sieve_env_hello ``` Thank you in advance!

2 2

dsync does not like sieve
by Stéphane Veyret 14 Jul '26

14 Jul '26

Hi, I’m using Dovecot for many years. My installation is configured to manage spam using sieve. In the same time, I have a backup server, connecting to the main server every night in order to backup emails (note that the main server is not pushing its data, but the backup server is pulling them). Dovecot v2.4 made me re-write the configuration files, so maybe I missed something, but since then, I cannot make both these service work in the same time. With my current configuration, the backup server is terminating in error with the following logs: dsync-remote(example(a)example.com)<HbkSGoKQG2o8+gEAGKlOPA>: Error: brain S: Exporting mailbox INBOX failed: Mailbox attribute iteration failed: Internal error occurred. Refer to server log for more information. [2026-05-31 03:36:03] (and there are no more logs anywhere). Because this message <https://dovecot.org/mailman3/archives/list/dovecot@dovecot.org/thread/MGCGR…> was suggesting there could be a problem with sieve, I tried to modify the configuration. And indeed, when changing sieve configuration, the backup is working again. But then, my spam filter isn’t anymore. Here are the relevant parts of the configuration for managing sieve. Can someone help me saying what could be wrong? protocols { sieve = yes imap = yes lmtp = yes } sieve_extensions { fileinto = yes reject = yes envelope = yes encoded-character = yes vacation = yes subaddress = yes comparator-i;ascii-numeric = yes relational = yes regex = yes imap4flags = yes copy = yes include = yes body = yes variables = yes enotify = yes environment = yes mailbox = yes date = yes index = yes ihave = yes duplicate = yes mime = yes foreverypart = yes extracttext = yes vnd.dovecot.pipe = yes vnd.dovecot.environment = yes } sieve_pipe_bin_dir = /etc/dovecot/sieve sieve_pipe_socket_dir = sieve-pipe sieve_plugins { sieve_imapsieve = yes sieve_extprograms = yes } service managesieve-login { inet_listener sieve { port = 4190 } } sieve_script personal { } sieve_script default { path = /etc/dovecot/sieve/spam.sieve } protocol imap { mail_plugins { acl = yes quota = yes imap_acl = yes imap_quota = yes imap_sieve = yes } } protocol lmtp { mail_plugins { acl = yes quota = yes sieve = yes } } namespace user { inbox = yes prefix = separator = / type = private mailbox * { sieve_script reportspam { cause = flag path = /etc/dovecot/sieve/report-spam.sieve type = before } } } Thank you. Stéphane -- Bien cordialement, / Plej kore, Stéphane Veyret Hi, I'm using Dovecot for many years. My installation is configured to manage spam using sieve. In the same time, I have a backup server, connecting to the main server every night in order to backup emails (note that the main server is not pushing its data, but the backup server is pulling them). Dovecot v2.4 made me re-write the configuration files, so maybe I missed something, but since then, I cannot make both these service work in the same time. With my current configuration, the backup server is terminating in error with the following logs: dsync-remote([1]example(a)example.com)<HbkSGoKQG2o8+gEAGKlOPA>: Error: brain S: Exporting mailbox INBOX failed: Mailbox attribute iteration failed: Internal error occurred. Refer to server log for more information. [2026-05-31 03:36:03] (and there are no more logs anywhere). Because [2]this message was suggesting there could be a problem with sieve, I tried to modify the configuration. And indeed, when changing sieve configuration, the backup is working again. But then, my spam filter isn't anymore. Here are the relevant parts of the configuration for managing sieve. Can someone help me saying what could be wrong? protocols { sieve = yes imap = yes lmtp = yes } sieve_extensions { fileinto = yes reject = yes envelope = yes encoded-character = yes vacation = yes subaddress = yes comparator-i;ascii-numeric = yes relational = yes regex = yes imap4flags = yes copy = yes include = yes body = yes variables = yes enotify = yes environment = yes mailbox = yes date = yes index = yes ihave = yes duplicate = yes mime = yes foreverypart = yes extracttext = yes vnd.dovecot.pipe = yes vnd.dovecot.environment = yes } sieve_pipe_bin_dir = /etc/dovecot/sieve sieve_pipe_socket_dir = sieve-pipe sieve_plugins { sieve_imapsieve = yes sieve_extprograms = yes } service managesieve-login { inet_listener sieve { port = 4190 } } sieve_script personal { } sieve_script default { path = /etc/dovecot/sieve/spam.sieve } protocol imap { mail_plugins { acl = yes quota = yes imap_acl = yes imap_quota = yes imap_sieve = yes } } protocol lmtp { mail_plugins { acl = yes quota = yes sieve = yes } } namespace user { inbox = yes prefix = separator = / type = private mailbox * { sieve_script reportspam { cause = flag path = /etc/dovecot/sieve/report-spam.sieve type = before } } } Thank you. Stephane -- Bien cordialement, / Plej kore, Stephane Veyret References Visible links 1. mailto:example@example.com 2. https://dovecot.org/mailman3/archives/list/dovecot@dovecot.org/thread/MGCGR…

3 8

crash in sieve-filter with 2.4.4
by Noah Meyerhans 12 Jul '26

12 Jul '26

Forwarded from https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1139275 A certain poison pill message seems to crash sieve-filter. A reproducer script, email message, and sieve script are at https://salsa.debian.org/noahm/dovecot/-/tree/sieve-crash/debian/tests/siev… The problem is not reproducible with 2.4.1, so it seems to be a regression introduced since then. Coredump details follow: root@dovecot-test:/home/noahm/dovecot# coredumpctl info | cat PID: 110911 (sieve-filter) TID: 110911 (sieve-filter) UID: 65534 (nobody) GID: 65534 (nogroup) Signal: 6 (ABRT) Timestamp: Mon 2026-06-08 19:16:37 UTC (29s ago) Command Line: sieve-filter -u dep8 -e -W -D -C -v /srv/dovecot-dep8/test.sieve INBOX Executable: /usr/bin/sieve-filter Control Group: /user.slice/user-1000.slice/session-5.scope Unit: session-5.scope Slice: user-1000.slice Session: 5 Owner UID: 1000 (noahm) Boot ID: 46469969e01043559d913ef8ab625856 Machine ID: a497a098bd094c1c8f7c8ed02aa21853 Hostname: dovecot-test Storage: /var/lib/systemd/coredump/core.sieve-filter.65534.46469969e01043559d913ef8ab625856.110911.1780946197000000.zst (present) Size on Disk: 253.5K Message: Process 110911 (sieve-filter) of user 65534 dumped core. Module libgcc_s.so.1 from deb gcc-16-16.1.0-1.amd64 Module libstdc++.so.6 from deb gcc-16-16.1.0-1.amd64 Stack trace of thread 110911: #0 0x00007fe06b09fcfc n/a (libc.so.6 + 0x97cfc) #1 0x00007fe06b048d02 raise (libc.so.6 + 0x40d02) #2 0x00007fe06b0304b2 abort (libc.so.6 + 0x284b2) #3 0x00007fe06b26ed45 n/a (libdovecot.so.0 + 0x6ed45) #4 0x00007fe06b357597 default_fatal_handler (libdovecot.so.0 + 0x157597) #5 0x00007fe06b26ec72 i_panic (libdovecot.so.0 + 0x6ec72) #6 0x00007fe06b273b90 n/a (libdovecot.so.0 + 0x73b90) #7 0x00007fe06b3498fb n/a (libdovecot.so.0 + 0x1498fb) #8 0x00007fe06b34a784 n/a (libdovecot.so.0 + 0x14a784) #9 0x00007fe06b34a93b buffer_append (libdovecot.so.0 + 0x14a93b) #10 0x00007fe06b344cc6 dregex_code_match_groups (libdovecot.so.0 + 0x144cc6) #11 0x00007fe06b806a27 n/a (libdovecot-sieve.so.0 + 0x6da27) #12 0x00007fe06b806cd0 n/a (libdovecot-sieve.so.0 + 0x6dcd0) #13 0x00007fe06b7ee411 sieve_match_value (libdovecot-sieve.so.0 + 0x55411) #14 0x00007fe06b7ee711 sieve_match (libdovecot-sieve.so.0 + 0x55711) #15 0x00007fe06b7f5862 n/a (libdovecot-sieve.so.0 + 0x5c862) #16 0x00007fe06b7eccde sieve_interpreter_continue (libdovecot-sieve.so.0 + 0x53cde) #17 0x00007fe06b7fbfbe sieve_execute (libdovecot-sieve.so.0 + 0x62fbe) #18 0x00005638c0022fcf main (sieve-filter + 0x3fcf) #19 0x00007fe06b031f77 n/a (libc.so.6 + 0x29f77) #20 0x00007fe06b032027 __libc_start_main (libc.so.6 + 0x2a027) #21 0x00005638c0023791 _start (sieve-filter + 0x4791) ELF object binary architecture: AMD x86-64

2 1

sieve-test (Pigeonhole) on Ubuntu: PCRE support and address test restrictions on Return-Path
by michael.merz＠wtnet.de 11 Jul '26

11 Jul '26

Hello, I've been evaluating sieve-test (Pigeonhole, dovecot-sieve package, version 2.3.21, as shipped with Ubuntu 24.04 LTS) as an offline testing tool for a personal Sieve ruleset that currently runs on a Cyrus IMAPd/CMU Sieve 3.0 backend. I'd like to ask for some clarification on two points, in case I'm missing a configuration option. 1) PCRE support My ruleset makes extensive use of the `(?i)` inline case-insensitivity flag within :regex tests, which is valid under PCRE (as used by Cyrus). When running sieve-test, I get: error: invalid regular expression '(?i)...' for regex match: invalid preceding regular expression. Checking `dovecot --build-options` shows no PCRE entry, which suggests the Ubuntu-packaged binary was compiled without --with-pcre and falls back to POSIX extended regex. Question: Is PCRE support in Pigeonhole's regex extension purely a compile-time decision (i.e. something that would need to be addressed at the distribution packaging level, not via runtime configuration), or is there a runtime way to enable PCRE matching that I might have missed (e.g. via -o, sieve_extensions, or similar)? 2) Return-Path in the `address` test Several of my rules use: address :domain :is "return-path" "..." Cyrus accepts this. Pigeonhole's sieve-test rejects it with: error: specified header 'return-path' is not allowed for the address test. I understand this is likely an RFC 5228 strictness difference (Cyrus being more permissive than the RFC requires). Is there a recommended alternative within Sieve itself to achieve the same effect (testing the address portion of Return-Path) while staying within what Pigeonhole accepts - e.g. via the `header` test combined with some address-parsing approach, or via the envelope test instead? For context: I'm not running a Dovecot mail server myself. I'm using sieve-test purely as an offline syntax/behaviour verification tool before deploying ruleset changes to a separate Cyrus-based production system. I understand this is a bit of an unusual use case, so any guidance on whether Pigeonhole is well suited for this kind of cross-implementation testing at all would also be appreciated. Thank you for your time and for maintaining Pigeonhole. Best regards, Michael Merz Hello, I've been evaluating sieve-test (Pigeonhole, dovecot-sieve package, version 2.3.21, as shipped with Ubuntu 24.04 LTS) as an offline testing tool for a personal Sieve ruleset that currently runs on a Cyrus IMAPd/CMU Sieve 3.0 backend. I'd like to ask for some clarification on two points, in case I'm missing a configuration option. 1) PCRE support My ruleset makes extensive use of the `(?i)` inline case-insensitivity flag within :regex tests, which is valid under PCRE (as used by Cyrus). When running sieve-test, I get: error: invalid regular expression '(?i)...' for regex match: invalid preceding regular expression. Checking `dovecot --build-options` shows no PCRE entry, which suggests the Ubuntu-packaged binary was compiled without --with-pcre and falls back to POSIX extended regex. Question: Is PCRE support in Pigeonhole's regex extension purely a compile-time decision (i.e. something that would need to be addressed at the distribution packaging level, not via runtime configuration), or is there a runtime way to enable PCRE matching that I might have missed (e.g. via -o, sieve_extensions, or similar)? 2) Return-Path in the `address` test Several of my rules use: address :domain :is "return-path" "..." Cyrus accepts this. Pigeonhole's sieve-test rejects it with: error: specified header 'return-path' is not allowed for the address test. I understand this is likely an RFC 5228 strictness difference (Cyrus being more permissive than the RFC requires). Is there a recommended alternative within Sieve itself to achieve the same effect (testing the address portion of Return-Path) while staying within what Pigeonhole accepts -- e.g. via the `header` test combined with some address-parsing approach, or via the envelope test instead? For context: I'm not running a Dovecot mail server myself. I'm using sieve-test purely as an offline syntax/behaviour verification tool before deploying ruleset changes to a separate Cyrus-based production system. I understand this is a bit of an unusual use case, so any guidance on whether Pigeonhole is well suited for this kind of cross-implementation testing at all would also be appreciated. Thank you for your time and for maintaining Pigeonhole. Best regards, Michael Merz

4 13

Client sent invalid input for CATEGORY
by Mail Operations 09 Jul '26

09 Jul '26

Hello list We are running a dovecot 2.4 instance and the following errors start to pop up after a listener is configured to scrape metrics with Prometheus: ... dovecot[2090309]: stats: Error: conn unix:stats (pid=2090333,uid=108): Client sent invalid input for CATEGORY: Category parent 'ssl' changed to 'ssl' (input: CATEGORY ssl-client ssl) Stats listener config: service stats { inet_listener http { port = 9090 ssl = yes } } So far I failed to reliably trigger the error messages and it looks like they appear randomly. Besides the logged errors, everything seems to work correctly. Could this be a possible bug, or configuration error? Kind regards, Patrik

2 2

doveadm backup / purge problems since migration from 2.3 to 2.4
by MK 09 Jul '26

09 Jul '26

Hi everyone, I have a two-server setup where all user data from the actively used server is periodically transferred to the second server using "doveadm backup". For this, I currently use a script that caches a list of all users (containing only the usernames) via "doveadm user '*' " and then executes the command "doveadm backup -u [username] tcp:1.2.3.4" (target host) for each user. I use a similarly structured script for a daily "doveadm purge", where a purge is performed per user instead of a backup. Under version 2.3, I always ran both commands using the "-A" flag (doveadm purge -A and doveadm backup -A tcp:1.2.3.4). Since migrating from Dovecot 2.3 to 2.4, this no longer works. When running the purge, the process always aborts after a certain number of users with the following error message: doveadm(john(a)doe.de): Error: auth-master: userdb list: User listing returned failure It's not the same User allways but after nearly the same time or amount of users the script has processed. Since I couldn't identify the exact cause of the error and suspect it might be related to changes in how the SQL iterate query is integrated in the 2.4 configuration (I use MySQL as the backend) I switched to the script described above. Then since is used the new script I have a non-reproducible issue. After the initial data transfer (when the target is still empty) the amount of data is equal to the source. Then after one of the next transfers the data is twice as high. It is random as to which subsequent transfer triggers this. When comparing the quotas on the source and target server, the output is identical on both. Even running `quota recalc` makes no difference. If I check the user directories, I can see that the `storage` folder contains nearly double the number of files and used disk space per user. My second problem is that when I run the purge script on the second server (target), it sometimes deletes all emails for all users, also non-reproducible. Afterwards, all files under "storage" are gone, and the quota shows 0 emails. I did not have this issue with the exactly same purge script on the first server which is the source for the backup. The version change was during the update from Debian 12 to 13, involving an upgrade of Dovecot from version 2.3.19 to 2.4.1. (both packages from debian repo) We use MySQL as the backend and mdbox as the storage format. I hope some can help me to solve this issue. Greetings, Oliver

2 1

Bug report: doveadm kick does not kick proxy sessions
by Patrick Cernko 06 Jul '26

06 Jul '26

Hi Dovecot Devs, I'm currently setting up Dovecot 2.4 as frontend proxy to our backends. The frontends simply use an LDAP attribute of the user to connect to the assigned IMAP backend server. That works fine so far. According to the documentation <https://doc.dovecot.org/2.4.4/core/man/doveadm-kick.1.html>, I should be able to disconnect a user with doveadm kick USERNAME or all sessions to a specific backend with doveadm kick -h BACKEND_IP or simply ALL sessions with doveadm kick '*' However the command reports the expected amount of connections kicked and exitcode 0, but all sessions stay connected. I also tried 'doveadm proxy kick ...' but as the the corresponding documentation states, this is only an alias for 'doveadm kick ...' and thus shows the same behavior. In the previous Dovecot 2.3 setup 'doveadm proxy kick' is not yet an alias for 'doveadm kick' and works as expected by disconnection the specified proxy sessions. 'doveadm kick' behaves similar to Dovecot 2.4 (not kicking proxy sessions) but that is understandable given that the documentation describes the dedicated kick command for proxy sessions. If this is the intended behavior for 'doveadm kick', how can I disconnect specific *proxy* sessions? If this is a bug, could you please fix it in the next releases or at least provide a patch that I can apply on my own build? I verified the bug exists in the latest dovecot packages for Debian Trixie from repo.dovecot.org. Attached is the doveconf -n output for my test setup. Best regards, -- Patrick Cernko <pcernko(a)mpi-klsb.mpg.de> +49 681 9325 5815 Joint Scientific IT and Technical Service Max-Planck-Institute für Informatik & Softwaresysteme

2 3

The SORT extension in 2.4.4 over imapc never returns any UIDs
by inaugurator＠ya.ru 29 Jun '26

29 Jun '26

Hi guys! This is really annoying. On the command to dovecot over imapc from another one: "UID SEARCH ALL" I get my UIDs: "* SEARCH 1 2 3 4" But when sending: "UID SORT (REVERSE ARRIVAL) US-ASCII NOT DELETED" I get literally nothing: "* SORT" Sorry, but this is a bug. Thank you!

3 7

How to use two databases in two passdb sections
by Евгений 26 Jun '26

26 Jun '26

Hello everyone! In version 2.3.21, I used two databases where each passdb queried its own database. After upgrading to 2.4.4, I cannot figure out how to restore this functionality. Is it no longer possible to use two databases in a single configuration after upgrade? Thank you for the best MDA!

3 2

Issues with syncing of shared mailboxes
by fc26wuqa＠studserv.uni-leipzig.de 26 Jun '26

26 Jun '26

Hello, I have a two users, that have a personal mailbox and a shared mailbox. They have access to the same shared mailbox. Today they notified me, that a mail send to the shared mailbox exists for the one user but not the other. The user also notified me, that some mails are not being delivered. I haven't found anything in my log files and the postmaster has also no delivery failures. I relay my mails through another server I don't have access to though. My /etc/aliases only consists of `postmaster: root`. The relay owners said that I should change postfix' main.cf to `double_bounce_sender = postmaster` a while ago. I am also not sure if I am not getting any delivery failure notifications since then, since I don't have a `postmaster` account, but I thought this is not needed since I have the above line in `/etc/aliases`. I also tried forcing a resync of their shared and non-shared inboxes with: `doveadm force-resync -u user(a)emaildomain.com INBOX` as well as `doveadm force-resync -u user(a)emaildomain.com Shared` (as well as with `Shared\*` - I don't know if that makes a difference tbh). Here is the output of `dovecot -n`: https://bin.disroot.org/?4eac9ea90a72ff08#A2dim9VGqHAjjKcggF4ScFWpbGytePZjt… I am kind of lost where to start from here, because any other person I asked told me their mail worked fine ^^. I would highly appreciate some guidance :) Thank you. Best, Lucienne

1 0

Convert sqlite configuration from 2.3 to 2.4
by Horst B Simon 24 Jun '26

24 Jun '26

Hi, I succesful converted all configurations from dovecot 2.3 to 2.4, except for the sqlite configuration. I googled and all configuration are for mysql/mariadb and not for sqlite. The sqlite database is created with postfixadmin and my current configuration for dovecot 2.3 (dovecot-sql.conf.ext) is: driver = sqlite connect = /var/mail/postfixadmin.db default_pass_scheme = DES-CRYPT password_query = \ SELECT username as user, password, '/var/mail/%d/%n' as userdb_home, \ 'maildir:/var/mail/%n' as userdb_mail, 5000 as userdb_uid, 5000 as userdb_gid \ FROM mailbox WHERE username = '%u' AND active = '1' user_query = \ SELECT '/var/mail/%d/%n' as home, 'maildir:/var/mail/%n' as mail, \ 5000 AS uid, 5000 AS gid, printf('dirsize:storage=', quota) AS quota \ FROM mailbox WHERE username = '%u' AND active = '1' I tried the only converter, but it didn’t do any conversion. Any help appreciated. Thanks and Regards, Horst Hi, I succesful converted all configurations from dovecot 2.3 to 2.4, except for the sqlite configuration. I googled and all configuration are for mysql/mariadb and not for sqlite. The sqlite database is created with postfixadmin and my current configuration for dovecot 2.3 (dovecot-sql.conf.ext) is: driver = sqlite connect = /var/mail/postfixadmin.db default_pass_scheme = DES-CRYPT password_query = \ SELECT username as user, password, '/var/mail/%d/%n' as userdb_home, \ 'maildir:/var/mail/%n' as userdb_mail, 5000 as userdb_uid, 5000 as userdb_gid \ FROM mailbox WHERE username = '%u' AND active = '1' user_query = \ SELECT '/var/mail/%d/%n' as home, 'maildir:/var/mail/%n' as mail, \ 5000 AS uid, 5000 AS gid, printf('dirsize:storage=', quota) AS quota \ FROM mailbox WHERE username = '%u' AND active = '1' I tried the only converter, but it didn't do any conversion. Any help appreciated. Thanks and Regards, Horst

2 2

Incomplete Dovecot Log Entries
by Bryan Simmons 19 Jun '26

19 Jun '26

There are hundreds of dovecot “Disconnected: Connection closed” messages in the log file that do not have a corresponding connect record. Is this normal or are there configuration options that need to be modified? Note the user accounts listed, and other similar records, do not exist on this server. 2026-06-14T02:11:48.843517-04:00 mail dovecot: imap-login: Disconnected: Connection closed (auth failed, 3 attempts in 23 secs): user=<rantanplan_admin(a)nanomail.net>, method=PLAIN, rip=190.181.120.152, lip=10.0.1.211, TLS, session=<T8xvlDBUnNC+tXiY> 2026-06-14T02:12:16.614611-04:00 mail dovecot: imap-login: Disconnected: Connection closed (auth failed, 3 attempts in 24 secs): user=<rantanplan_admin>, method=PLAIN, rip=41.139.10.25, lip=10.0.1.211, TLS, session=<cvoGljBUS+YpiwoZ> 2026-06-14T03:23:55.523146-04:00 mail dovecot: imap-login: Disconnected: Connection closed (auth failed, 3 attempts in 23 secs): user=<rantanplan.info(a)nanomail.net>, method=PLAIN, rip=103.69.9.148, lip=10.0.1.211, TLS, session=<SOBYljFUlsdnRQmU> Thanks, Bryan

1 0

Announcing nauthilus-director: a Nauthilus-backed mail protocol director
by Christian Rößner 17 Jun '26

17 Jun '26

Hi, I would like to announce `nauthilus-director` for the first time on this list. `nauthilus-director` is an open-source mail protocol director and transparent proxy for Nauthilus-backed mail deployments. It sits in front of stateful mail backends, authenticates frontend sessions through Nauthilus, resolves routing and placement inside the director, selects healthy backend services, keeps active user affinity in Redis, and then proxies the established connection to the selected backend. The current focus is on deployments where backend placement, maintenance, drains, user movement and session affinity need to be explicit, observable and controllable. The director is not a mailbox server and not a general-purpose load balancer; it owns the narrow director responsibilities between clients, Nauthilus and backend mail services. Current capabilities include: - IMAP/IMAPS, POP3/POP3S, LMTP/LMTPS and ManageSieve proxying - Nauthilus-backed authentication over HTTP or gRPC - director-owned routing and backend selection - Redis-backed session affinity and runtime coordination - backend health, maintenance, runtime drains and placement overrides - an OpenAPI-based REST control API and `nauthilus-directorctl` - metrics, structured logging and OpenTelemetry tracing The demo stack includes Dovecot backend shards and exercises real IMAP, LMTP, POP3 and ManageSieve paths, so feedback from Dovecot operators would be especially useful. Related to this, Nauthilus itself has recently been released as version 3.0.0. Nauthilus is the central authentication and policy engine in this setup: it provides centralized authentication and identity handling for mail and web workloads, LDAP-backed or Lua-driven policy decisions, OIDC/SAML IdP functionality, MFA support, and Redis-backed state handling. In the director architecture, Nauthilus remains the authentication and policy authority, while `nauthilus-director` owns concrete backend placement and proxy lifecycle decisions. Project links: - nauthilus-director: https://github.com/croessner/nauthilus-director - Nauthilus: https://github.com/croessner/nauthilus - Documentation / website: https://nauthilus.org The director is still young, so I would be interested in feedback from people running Dovecot-backed infrastructures, especially around operational workflows, backend maintenance, migration scenarios and protocol behavior. The director is still in beta. Best regards, Christian Rößner -- Rößner-Network-Solutions Zertifizierter ITSiBe / CISO Marburger Str. 70a, 36304 Alsfeld Mobil: +49 171 9905345 USt-IdNr.: DE225643613, https://roessner.website PGP fingerprint: 658D 1342 B762 F484 2DDF 1E88 38A5 4346 D727 94E5 Hi, I would like to announce `nauthilus-director` for the first time on this list. `nauthilus-director` is an open-source mail protocol director and transparent proxy for Nauthilus-backed mail deployments. It sits in front of stateful mail backends, authenticates frontend sessions through Nauthilus, resolves routing and placement inside the director, selects healthy backend services, keeps active user affinity in Redis, and then proxies the established connection to the selected backend. The current focus is on deployments where backend placement, maintenance, drains, user movement and session affinity need to be explicit, observable and controllable. The director is not a mailbox server and not a general-purpose load balancer; it owns the narrow director responsibilities between clients, Nauthilus and backend mail services. Current capabilities include: - IMAP/IMAPS, POP3/POP3S, LMTP/LMTPS and ManageSieve proxying - Nauthilus-backed authentication over HTTP or gRPC - director-owned routing and backend selection - Redis-backed session affinity and runtime coordination - backend health, maintenance, runtime drains and placement overrides - an OpenAPI-based REST control API and `nauthilus-directorctl` - metrics, structured logging and OpenTelemetry tracing The demo stack includes Dovecot backend shards and exercises real IMAP, LMTP, POP3 and ManageSieve paths, so feedback from Dovecot operators would be especially useful. Related to this, Nauthilus itself has recently been released as version 3.0.0. Nauthilus is the central authentication and policy engine in this setup: it provides centralized authentication and identity handling for mail and web workloads, LDAP-backed or Lua-driven policy decisions, OIDC/SAML IdP functionality, MFA support, and Redis-backed state handling. In the director architecture, Nauthilus remains the authentication and policy authority, while `nauthilus-director` owns concrete backend placement and proxy lifecycle decisions. Project links: - nauthilus-director: https://github.com/croessner/nauthilus-director - Nauthilus: https://github.com/croessner/nauthilus - Documentation / website: https://nauthilus.org The director is still young, so I would be interested in feedback from people running Dovecot-backed infrastructures, especially around operational workflows, backend maintenance, migration scenarios and protocol behavior. The director is still in beta. Best regards, Christian Roessner -- Roessner-Network-Solutions Zertifizierter ITSiBe / CISO Marburger Str. 70a, 36304 Alsfeld Mobil: +49 171 9905345 USt-IdNr.: DE225643613, https://roessner.website PGP fingerprint: 658D 1342 B762 F484 2DDF 1E88 38A5 4346 D727 94E5

3 3

Announcement: wormhole plugin, replication for Dovecot 2.4
by Patrick Cernko 12 Jun '26

12 Jun '26

Hello List, I am proud to announce my project "wormhole". This is an extension for Dovecot 2.4.1. It allows to continue the use of the replication feature of Dovecot that was removed with the 2.4 release. History: I started this project by forking the replication code of Dovecot 2.3.21.1 and porting it to 2.4. The first result was a patch for Dovecot 2.4.1. While working with the code, I realized that it is possible to provide the feature as a plugin / extension. In the first release of the plugin, I tried to integrate it with the existing dsync commands for doveadm. This required an API for access to the dsync commands, which I proposed to the Dovecot developers in https://github.com/dovecot/core/pull/249 . Unfortunately, my pull request was rejected, though I can understand the reasons for that. Without the required API, I had to switch to another approach: I forked the dsync code from doveadm to provide a new set of commands: "repl" and "repl-server". These commands work similar to "sync" and "dsync-server" but use additional functions required by the replication feature. This approach was released as version 0.3 of the plugin. With version 0.4 I renamed the project to wormhole. Before that it was just called dovecot-replication (see https://codeberg.org/errror/dovecot-replication) I thought that the project should have a more unique name. Currently I'm working on designing a test suite for wormhole. Doing unit tests of the components is not sufficient IMHO. The test suite should check the whole replication feature using two running dovecot instances that replicate the users' mailboxes. Having a sophisticated test suite will help in future development and bug tracking. The code can be found on my codeberg site at: https://codeberg.org/errror/wormhole There is also a web page on my personal homepage at: https://people.mpi-klsb.mpg.de/~pcernko/wormhole.shtml The project is still missing documentation. For the moment, configuration is similar to the existing Dovecot 2.3 documentation. Upgrading from 2.3: When using the included patch for Dovecot, the plugin should be 100% compatible with existing Dovecot 2.3 servers. This compatibility should allow uninterrupted upgrades of the service by replacing Dovecot on one server and then the other one. After the upgrade, the plugin can be reconfigured to work without the patch for Dovecot. Upgrading without the patch is also possible. During the upgrade (aka "while one server still runs 2.3") some aspects of the replication will not work or lack performance. Upgrading is not yet tested very well and missing documentation also. I will work on that soon. I hope to get some feedback, pull requests or bug reports via mail or at codeberg (preferred). -- Patrick Cernko <pcernko(a)mpi-klsb.mpg.de> +49 681 9325 5815 Joint Scientific IT and Technical Service Max-Planck-Institute für Informatik & Softwaresysteme

4 4

New Backup Issue
by Doug Hardie 11 Jun '26

11 Jun '26

I have dovecot using maildir storage on the mail machine. I have an external drive directly connected to use a backup for the mailboxes. Doveadm backup works perfectly for this setup. I also have two external backup servers. Backup to those tends to fail with a variety of errors. I deleted the Maildir directory on the backup servers and created the original backups with doveadm backup. No issues were reported. However, the next backup generates errors: dsync-remote(jeanne)<HV6JE+i/KWo0NwAAq6l46g>: Warning: Deleting mailbox 'INBOX': UID=11166 GUID=1712570780.M604032P10319.mail,S=99689,W=101000 is missing locally dsync-remote(jeanne)<HV6JE+i/KWo0NwAAq6l46g>: Error: Couldn't delete mailbox INBOX: INBOX can't be deleted. dsync-local(jeanne)<HqU9Aei/KWoBMAAAZU03Dg>: Error: Remote command returned error 65: ssh sermons doveadm dsync-server -ujeanne -U dsync-remote(beneke)<eARRAem/KWpdNwAAq6l46g>: Warning: Deleting mailbox 'INBOX': UID=232 GUID=1641564460.M859673P8687.mail,S=5917,W=6035 is missing locally dsync-remote(beneke)<eARRAem/KWpdNwAAq6l46g>: Error: Couldn't delete mailbox INBOX: INBOX can't be deleted. dsync-local(beneke)<mw5bM+i/KWoJMAAAZU03Dg>: Error: Remote command returned error 65: ssh sermons doveadm dsync-server -ubeneke -U dsync-remote(doug)<6gmSJ+m/KWp7NwAAq6l46g>: Warning: Deleting mailbox 'INBOX': UID=92440 GUID=1465118975.V4eI7cfa32M232845.mail is missing locally dsync-remote(doug)<6gmSJ+m/KWp7NwAAq6l46g>: Error: Couldn't delete mailbox INBOX: INBOX can't be deleted. dsync-local(doug)<hwsCH+m/KWoNMAAAZU03Dg>: Error: Remote command returned error 65: ssh sermons doveadm dsync-server -udoug -U There is always one error per user, but they are different between the two backup machine. With the first error, I found the message 1712570780.M604032P10319.mail,S=99689,W=101000 in the mail server cur directory and deleted it. That messge did not exist in the backup server. I ran backup again and no issues. However, the next backup produced the following error: dsync-remote(jeanne)<dNWXGZbUKWoaOgAAq6l46g>: Warning: Deleting mailbox 'INBOX': UID=11167 GUID=1712574432.M425061P11306.mail,S=45149,W=45775 is missing locally dsync-remote(jeanne)<dNWXGZbUKWoaOgAAq6l46g>: Error: Couldn't delete mailbox INBOX: INBOX can't be deleted. dsync-local(jeanne)<Jg3tCZbUKWp5NQAAZU03Dg>: Error: Remote command returned error 65: ssh sermons doveadm dsync-server -ujeanne -U That message is in the cur directory on the mail server. It is not in the backup server. I don't understand the message that those files are missing locally (primary mail server I presume) when they are actually there and missing on the remote server. I don't know why they didn't get copied originally. The first email was recieved on Mon, 8 Apr 2024 so it can't be a timming issue. -- Doug

2 1

Replication performance question and environment variables sanitization
by Jesus Cea 10 Jun '26

10 Jun '26

Hi, there. I am using dovecot 2.3.21.1. I am experimenting with replication, currently doing it manually, using this: doveadm -Dv sync -u jcea ssh email doveadm dsync-server -u prueba_backup; echo $? It works fine so far, but I have a few hundreds of mailboxes in my imap4 account and an "empty" (no changes) sync takes 30 seconds, since apparently it compare every mailbox (according to the debug info printed). I am using "mdbox" storage in both sides. I would expect an event replication, where time is proportional to the volume of changes, not to the total data volume or number of mailboxes. Am I doing anything wrong?. It would be faster to use "replicator" + "mail_replica". Can I improve this in some way? Replication must be both ways, because in one server new emails are arriving and in the other imapserver I am doing my mail work. I want new email appearing in my work imapserver (living my own laptop) and my email processing to be replicated back to the server for backup/disaster recovery (lost laptop, broken HD) and access from a different laptop, phone, etc. Thanks. PS: By default dovecot sanitizes the environment and doesn't allow "SSH_AUTH_SOCK" for ssh-agent. Why?. That is very inconvenient. I am using this environment variable to overcome that: DOVECOT_PRESERVE_ENVS="HOME USER SSH_AUTH_SOCK SSH_AGENT_PID LOG_STDERR_TIMESTAMP TZ CORE_OUTOFMEM CORE_ERROR" Is this the right way to do this? -- Jesús Cea Avión _/_/ _/_/_/ _/_/_/ jcea(a)jcea.es - https://www.jcea.es/ _/_/ _/_/ _/_/ _/_/ _/_/ Twitter: @jcea _/_/ _/_/ _/_/_/_/_/ jabber / xmpp:jcea@jabber.org _/_/ _/_/ _/_/ _/_/ _/_/ "Things are not so easy" _/_/ _/_/ _/_/ _/_/ _/_/ _/_/ "My name is Dump, Core Dump" _/_/_/ _/_/_/ _/_/ _/_/ "El amor es poner tu felicidad en la felicidad de otro" - Leibniz

2 3

dovecot 2.4 and Outlook
by Jochen Bern 08 Jun '26

08 Jun '26

Since I don't remember seeing *this* mentioned here so far, but *do* remember people asking about Outlook-dovecot compatibility issues every now and then, here's a potential cause for seeing more of those in the immediate future: https://fedoramagazine.org/fedora-43-upgrade-revealed-20-years-old-outlook-… Kind regards, -- Jochen Bern Systemingenieur Binect GmbH

2 3

Using 2 LDAP sources for passdb
by Tom 03 Jun '26

03 Jun '26

I can't seem to find documentation that shows how to set up 2 LDAP passdb blocks. I can get each of them working properly, but only the second of the two works at any given time. I can't figure out the syntax needed to get both to work, even though I've been all over the Dovecot 2.4x official documentation. The docs seem to suggest settings that the server rejects. Anyone have any experience doing this, or know the correct syntax? Thanks in advance! passdb ldap { # works when standalone; fails when combined with 2nd block below #passdb_name = ldap1 #driver = ldap ldap_version = 3 bind = yes bind_userdn = %{user|username} ldap_auth_dn = cn=dovecot,ou=apps,dc=example,dc=com ldap_auth_dn_password = D0vec0t ldap_base = ou=apps,dc=foscore,dc=com ldap_uris = ldapi://%2Frun%2Fldapi filter = (&(objectClass=applicationProcess)(cn=%{user})) passdb_ldap_bind = yes passdb_ldap_bind_userdn = cn=%{user},ou=apps,dc=example,dc=com result_success = return-ok } passdb ldap { #driver = ldap #name = ldap_users ldap_version = 3 bind = yes bind_userdn = %{user|username} ldap_auth_dn = cn=dovecot,ou=apps,dc=example,dc=com ldap_auth_dn_password = D0vec0t ldap_base = ou=people,dc=example,dc=com ldap_uris = ldapi://%2Frun%2Fldapi filter = (&(objectClass=posixAccount)(uid=%{user|username})(memberOf=cn=mail,ou=%{user|domain},ou=groups,dc=example,dc=com)) passdb_ldap_bind = yes passdb_ldap_bind_userdn = uid=%{user|username},ou=people,dc=example,dc=com result_success = return-ok }

2 7

Building on FreeBSD 14.4
by Frank Volf 03 Jun '26

03 Jun '26

Hi, I'm trying to build dovecot-2.4.4 on FreeBSD 14.4 and I get a strange error message. What I did is download and extract dovecot-2.4.4.tar.gz and then run ./configure without any arguments. After typing make the build started without any problems until this point: .... Making all in auth CC test_auth_cache-auth-cache.o CC test_auth_cache-test-auth-cache.o CC auth-main.o CC auth-auth.o CC auth-auth-cache.o CC auth-auth-client-connection.o CC auth-auth-master-connection.o CC auth-auth-policy.o CC auth-auth-penalty.o CC auth-auth-request.o CC auth-auth-request-fields.o CC auth-auth-request-handler.o CC auth-auth-request-var-expand.o CC auth-auth-sasl-mech-apop.o CC auth-auth-sasl-mech-dovecot-token.o CC auth-auth-sasl-mech-oauth2.o CC auth-auth-sasl.o CC auth-auth-settings.o CC auth-auth-fields.o CC auth-auth-token.o CC auth-auth-worker-connection.o CC auth-auth-worker-server.o CC auth-db-oauth2.o CC auth-db-sql.o CC auth-db-passwd-file.o CC auth-passdb.o CC auth-passdb-blocking.o CC auth-passdb-bsdauth.o CC auth-passdb-cache.o CC auth-passdb-oauth2.o CC auth-passdb-passwd.o CC auth-passdb-passwd-file.o CC auth-passdb-pam.o CC auth-passdb-sql.o CC auth-passdb-static.o CC auth-userdb.o CC auth-userdb-blocking.o CC auth-userdb-passwd.o CC auth-userdb-passwd-file.o CC auth-userdb-prefetch.o CC auth-userdb-static.o CC auth-userdb-sql.o CC auth-db-ldap.o CC auth-db-ldap-sasl.o CC auth-db-ldap-settings.o CC auth-passdb-ldap.o CC auth-userdb-ldap.o CC auth-db-lua.o CC auth-passdb-lua.o CC auth-userdb-lua.o make[3]: don't know how to make -L/usr/local/lib. Stop make[3]: stopped making "all" in /home/volf/dovecot-2.4.4/src/auth *** Error code 1 So, apparently make is confused and thinks that the compiler/linker commandline option is an item to build. I'm at loss how to solve this. I would appreciate it, if somebody could give me a hint how to solve this. Kind regards, Frank Hi, I'm trying to build dovecot-2.4.4 on FreeBSD 14.4 and I get a strange error message. What I did is download and extract dovecot-2.4.4.tar.gz and then run ./configure without any arguments. After typing make the build started without any problems until this point: .... Making all in auth CC test_auth_cache-auth-cache.o CC test_auth_cache-test-auth-cache.o CC auth-main.o CC auth-auth.o CC auth-auth-cache.o CC auth-auth-client-connection.o CC auth-auth-master-connection.o CC auth-auth-policy.o CC auth-auth-penalty.o CC auth-auth-request.o CC auth-auth-request-fields.o CC auth-auth-request-handler.o CC auth-auth-request-var-expand.o CC auth-auth-sasl-mech-apop.o CC auth-auth-sasl-mech-dovecot-token.o CC auth-auth-sasl-mech-oauth2.o CC auth-auth-sasl.o CC auth-auth-settings.o CC auth-auth-fields.o CC auth-auth-token.o CC auth-auth-worker-connection.o CC auth-auth-worker-server.o CC auth-db-oauth2.o CC auth-db-sql.o CC auth-db-passwd-file.o CC auth-passdb.o CC auth-passdb-blocking.o CC auth-passdb-bsdauth.o CC auth-passdb-cache.o CC auth-passdb-oauth2.o CC auth-passdb-passwd.o CC auth-passdb-passwd-file.o CC auth-passdb-pam.o CC auth-passdb-sql.o CC auth-passdb-static.o CC auth-userdb.o CC auth-userdb-blocking.o CC auth-userdb-passwd.o CC auth-userdb-passwd-file.o CC auth-userdb-prefetch.o CC auth-userdb-static.o CC auth-userdb-sql.o CC auth-db-ldap.o CC auth-db-ldap-sasl.o CC auth-db-ldap-settings.o CC auth-passdb-ldap.o CC auth-userdb-ldap.o CC auth-db-lua.o CC auth-passdb-lua.o CC auth-userdb-lua.o make[3]: don't know how to make -L/usr/local/lib. Stop make[3]: stopped making "all" in /home/volf/dovecot-2.4.4/src/auth *** Error code 1 So, apparently make is confused and thinks that the compiler/linker commandline option is an item to build. I'm at loss how to solve this. I would appreciate it, if somebody could give me a hint how to solve this. Kind regards, Frank

6 14

Re: User authentication for shared mailbox
by Stéphane Veyret 31 May '26

31 May '26

I managed to make it work. In case some other people want to do the same, the trick was to use the master user concept, as suggested by Aki Tuomi, with a lua script to check if user can be master. My configuration looks like this: passdb pam_master { driver = pam master = yes result_success = continue } passdb pam { } passdb lua { lua_file = /etc/dovecot/master-auth.lua result_failure = return-fail result_internalfail = return-fail result_success = continue-ok } passdb ldap { ldap_base = ou=Group,dc=example,dc=com fields { password = {SHA1}IQ== user = %{ldap:cn} } filter = (cn=%{user | username}) } and the lua script: function script_init() return 0 end function auth_passdb_lookup(req) if req.master_user then req:log_info("Looking for groups for " .. tostring(req.master_user)) handle = io.popen("/bin/groups " .. tostring(req.master_user)) output = handle:read("*a") req:log_debug("Searching " .. tostring(req.user) .. " in found groups " .. output) for group in output:gmatch("%w+") do if group == req.user then return dovecot.auth.PASSDB_RESULT_OK, { password="!" } end end end return dovecot.auth.PASSDB_RESULT_USER_UNKNOWN, "No such user" end -- Bien cordialement, / Plej kore, Stéphane Veyret I managed to make it work. In case some other people want to do the same, the trick was to use the master user concept, as suggested by Aki Tuomi, with a lua script to check if user can be master. My configuration looks like this: passdb pam_master { driver = pam master = yes result_success = continue } passdb pam { } passdb lua { lua_file = /etc/dovecot/master-auth.lua result_failure = return-fail result_internalfail = return-fail result_success = continue-ok } passdb ldap { ldap_base = ou=Group,dc=example,dc=com fields { password = {SHA1}IQ== user = %{ldap:cn} } filter = (cn=%{user | username}) } and the lua script: function script_init() return 0 end function auth_passdb_lookup(req) if req.master_user then req:log_info("Looking for groups for " .. tostring(req.master_user)) handle = io.popen("/bin/groups " .. tostring(req.master_user)) output = handle:read("*a") req:log_debug("Searching " .. tostring(req.user) .. " in found groups " .. output) for group in output:gmatch("%w+") do if group == req.user then return dovecot.auth.PASSDB_RESULT_OK, { password="!" } end end end return dovecot.auth.PASSDB_RESULT_USER_UNKNOWN, "No such user" end -- Bien cordialement, / Plej kore, Stephane Veyret

1 0

dsync between 2.3.16 and 2.4.2
by Jean-Daniel Beaubien 28 May '26

28 May '26

Hi, I currently have a doveadm backup running my prod and backup server. Both are currently running 2.3.16, and the following command runs fine: sudo doveadm -v backup -u XXXX -P remote:XXXX@192.168.111.XX I am in the process of updating my servers to ubuntu 26.04 which runs dovecot 2.4.2 and I am running into problems trying to run this command. For some reason, it tries to delete the INBOX on the destination and I get the following error: dsync-remote(XXXX)<7hkbKDl6F2rlxQAArPp36Q>: Error: brain S: Mailbox INBOX sync: mailbox_delete failed: The /private/specialuse attribute cannot be changed doveadm(228686): Error: Remote command returned error 65: ssh -lXXXX 192.168.111.XX doveadm dsync-server -uXXXX I am not sure what/why this is happening. I read the docs and man page, but could not find a solution. Anyone knows what is happening, and how to solve this? Thanks, -JD ----- Addendum ----- Another thing that might help someone else: I had to add a line to the config file to avoid the -U problem: dsync_remote_cmd = "ssh -l%{login} %{host} doveadm dsync-server -u%{user}" This seems to already be solved ( https://dovecot.org/mailman3/archives/list/dovecot@dovecot.org/thread/B3TQF…) but not yet in the ubuntu 26.04 release? Hi, I currently have a doveadm backup running my prod and backup server. Both are currently running 2.3.16, and the following command runs fine: sudo doveadm -v backup -u XXXX -P remote:XXXX@192.168.111.XX I am in the process of updating my servers to ubuntu 26.04 which runs dovecot 2.4.2 and I am running into problems trying to run this command. For some reason, it tries to delete the INBOX on the destination and I get the following error: dsync-remote(XXXX)<7hkbKDl6F2rlxQAArPp36Q>: Error: brain S: Mailbox INBOX sync: mailbox_delete failed: The /private/specialuse attribute cannot be changed doveadm(228686): Error: Remote command returned error 65: ssh -lXXXX 192.168.111.XX doveadm dsync-server -uXXXX I am not sure what/why this is happening. I read the docs and man page, but could not find a solution. Anyone knows what is happening, and how to solve this? Thanks, -JD ----- Addendum ----- Another thing that might help someone else: I had to add a line to the config file to avoid the -U problem: dsync_remote_cmd = "ssh -l%{login} %{host} doveadm dsync-server -u%{user}" This seems to already be solved ([1]https://dovecot.org/mailman3/archives/list/dovecot@dovecot.org/thread/B3T…) but not yet in the ubuntu 26.04 release? References Visible links 1. https://dovecot.org/mailman3/archives/list/dovecot@dovecot.org/thread/B3TQF…

2 1

I can't find up to date documentation about dovecot authentication process design
by kristoffermollerhoj＠gmail.com 27 May '26

27 May '26

Hi, I am currently about to port a legacy / custom authentication mechanism from 2.3 to 2.4. Many things have changed in this area, and I have unfortunately not been able to find any valid documentation about this topic. I have looked here: https://doc.dovecot.org/2.4.4/developers/design/auth_process.html but as far as I can see all the content is about the code design for dovecot 2.3, not 2.4. Just to mention a few examples: There is not anymore a struct named mech_module, there is no such file as mech.h, and also the API for the central callback functions such as auth_new() and auth_initial() seems to have changed a lot (both arguments and return types). Maybe a somewhat unfinished version of the documentation can be found on a specific branch? Thank you, Kristoffer Møllerhøj.

4 5

Sent, Junk and Drafts are not created
by Thomas Jensen 26 May '26

26 May '26

I have a new dovecot 2.4.2 setup. All seems to work fine, except Sent, Junk and Drafts are not created. In 15-mailboxes.conf I have: namespace inbox { # These mailboxes are widely used and could perhaps be created automatically: mailbox Drafts { auto = create special_use = \Drafts } mailbox Junk { auto = create special_use = \Junk } mailbox Trash { auto = create special_use = \Trash } # For \Sent mailboxes there are two widely used names. We'll mark both of # them as \Sent. User typically deletes one of them if duplicates are created. mailbox Sent { auto = create special_use = \Sent } mailbox "Sent Messages" { auto = create special_use = \Sent } } But nothing is created. And dovecot -n return this: (note the "\\") root@rpix:~# dovecot -n # 2.4.2 (0962ed2104): /etc/dovecot/dovecot.conf # Pigeonhole version 2.4.2 (767418c3) # OS: Linux 7.0.0-15-generic x86_64 Ubuntu 26.04 LTS # Hostname: rpix.dk # 9 default setting changes since version 2.4.0 dovecot_config_version = 2.4.0 auth_default_domain = rpix.dk auth_mechanisms = plain login auth_username_format = %{user|lower} dovecot_storage_version = 2.4.0 fts_autoindex = yes fts_autoindex_max_recent_msgs = 999 fts_search_add_missing = yes mail_driver = maildir mail_home = /var/vmail/%{user | domain}/%{user | username} mail_path = ~/Maildir mail_privileged_group = mail protocols = imap lmtp sql_driver = mysql ssl = required passdb sql { default_password_scheme = ARGON2I query = SELECT username AS user,password FROM mailbox WHERE username = '%{user}' AND active='1' } userdb sql { iterate_query = SELECT username AS user FROM mailbox query = SELECT 2000 AS uid, 2000 AS gid FROM mailbox WHERE username = '%{user}' AND active='1' } mysql localhost { dbname = postfixadmin password = # hidden, use -P to show it user = postfixadmin } namespace inbox { inbox = yes mailbox Drafts { auto = create special_use = "\\Drafts" } mailbox Junk { auto = create special_use = "\\Junk" } mailbox Trash { auto = create special_use = "\\Trash" } mailbox Sent { auto = create special_use = "\\Sent" } mailbox "Sent Messages" { auto = create special_use = "\\Sent" } } service imap-login { inet_listener imap { } inet_listener imaps { } } service pop3-login { inet_listener pop3 { } inet_listener pop3s { } } service submission-login { inet_listener submission { } inet_listener submissions { } } service lmtp { unix_listener /var/spool/postfix/private/dovecot-lmtp { group = postfix mode = 0600 user = postfix } } service imap { } service pop3 { } service submission { } service auth { unix_listener /var/spool/postfix/private/auth { group = postfix mode = 0666 user = postfix } } service auth-worker { } service dict { unix_listener dict { } } service stats { unix_listener stats-reader { group = www-data mode = 0660 user = www-data } unix_listener stats-writer { group = www-data mode = 0660 user = www-data } } ssl_server { cert_file = /etc/letsencrypt/live/mail/fullchain.pem key_file = /etc/letsencrypt/live/mail/privkey.pem prefer_ciphers = server } protocol lmtp { auth_username_format = %{user | lower} } What am I doing wrong? -- /Thomas Jensen, Denmark

3 3

Possible FTS language detection packaging issue in Dovecot 2.4.x Debian packages
by Jens Urban 26 May '26

26 May '26

Hello, Is anyone running Dovecot 2.4.4 with multilingual FTS flatcurve on Debian 12/13, using the packages from Dovecot.org? I suspect a packaging/build issue related to language detection. It looks like libexttextcat is not included in the Debian build. Simple reproduction with doveadm: doveadm fts tokenize -u localpart@domain 'This is the house. Das ist das Haus‘ —> Warning: Can't detect its language - assuming de …… My suspicion: If the "libexttextcat-dev" package was not installed during the "configure" step of the package creation process, then "HAVE_LANG_EXTTEXTCAT" will not be set. And if the define "HAVE_LANG_EXTTEXTCAT" is not set during the compilation of "src/lib-language/language.c", then "language_textcat_init" always returns "LANGUAGE_DETECT_RESULT_UNKNOWN;". Consequently, doveadm invariably translates this return value into the message: "Warning: Can't detect its language - assuming $DEFAULT-LANG". If I download the source code from Dovecot.org and build the dovecot vendor package myself, "language detection" works—provided that "libexttextcat-dev" has been installed beforehand. My simple test to check whether the installed version contains „libexttextcat“: (My build finds references to libexttextcat; the build from Dovecot.org finds none.) nm --print-file-name --dynamic --undefined-only /usr/lib/x86_64-linux-gnu/dovecot/* | grep textcat —> /usr/lib/x86_64-linux-gnu/dovecot/libdovecot-language.so.0.0.0: U special_textcat_Init —> /usr/lib/x86_64-linux-gnu/dovecot/libdovecot-storage.so.0.0.0: U textcat_ReleaseClassifyFullOutput Background: In my self-built package, exttextcat-related symbols appear in both libdovecot-language and libdovecot-storage; I assume this is expected due to the build/link layout. Regards, Jens Hello, Is anyone running Dovecot 2.4.4 with multilingual FTS flatcurve on Debian 12/13, using the packages from Dovecot.org? I suspect a packaging/build issue related to language detection. It looks like libexttextcat is not included in the Debian build. Simple reproduction with doveadm: doveadm fts tokenize -u localpart@domain 'This is the house. Das ist das HausaEUR~ aEUR"> Warning: Can't detect its language - assuming de aEUR|aEUR| My suspicion: If the "libexttextcat-dev" package was not installed during the "configure" step of the package creation process, then "HAVE_LANG_EXTTEXTCAT" will not be set. And if the define "HAVE_LANG_EXTTEXTCAT" is not set during the compilation of "src/lib-language/language.c", then "language_textcat_init" always returns "LANGUAGE_DETECT_RESULT_UNKNOWN;". Consequently, doveadm invariably translates this return value into the message: "Warning: Can't detect its language - assuming $DEFAULT-LANG". If I download the source code from Dovecot.org and build the dovecot vendor package myself, "language detection" worksaEUR"provided that "libexttextcat-dev" has been installed beforehand. My simple test to check whether the installed version contains aEURzlibexttextcataEURoe: (My build finds references to libexttextcat; the build from Dovecot.org finds none.) nm --print-file-name --dynamic --undefined-only /usr/lib/x86_64-linux-gnu/dovecot/* | grep textcat aEUR"> /usr/lib/x86_64-linux-gnu/dovecot/libdovecot-language.so.0.0.0: U special_textcat_Init aEUR"> /usr/lib/x86_64-linux-gnu/dovecot/libdovecot-storage.so.0.0.0: U textcat_ReleaseClassifyFullOutput Background: In my self-built package, exttextcat-related symbols appear in both libdovecot-language and libdovecot-storage; I assume this is expected due to the build/link layout. Regards, Jens

1 0

dovecot 2.4.4 community edition, bug in variables
by Terry Braun 22 May '26

22 May '26

I'm using this passdb sql to a postgres db passdb sql { query = SELECT NULL AS password, 'Y' as nopassword, userid AS user \ FROM users \ WHERE addr = '%{user}' AND password = '%{password}' } This succeed and the addr passed in as user is foo(a)example.com as expected I use this userdb sql userdb sql { query = SELECT home, uid, gid \ FROM users \ WHERE addr = '%{user}' # For using doveadm -A: # iterate_query = SELECT userid AS username, domain FROM users } This fails because postgres receives just "foo" not "foo(a)example.com" as user. maybe there is a bug in my stuff but I cannot see it. I can see the query in the postgres log where it clearly says: WHERE addr = 'foo' It also logs the passdb query: WHERE addr = 'foo(a)example.com' AND password='secret' thanks Terry

2 2

Problem with authentication in dovecot
by Freek de Kruijf 22 May '26

22 May '26

I have the following dovecot.conf, striped from comment: dovecot_config_version = 2.4.2 dovecot_storage_version = 2.4.0 protocols = imap auth_allow_cleartext = yes auth_mechanisms = plain auth_verbose = yes mail_plugin_dir = /usr/lib64/dovecot/modules namespace inbox { inbox = yes mail_driver = maildir mail_path = ~/Maildir prefix = INBOX. mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Trash { auto = subscribe special_use = \Trash } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } } passdb pam { service_name = dovecot } userdb passwd { use_worker = yes } service auth { unix_listener /var/spool/postfix/private/auth { group = postfix mode = 0666 user = postfix } } ssl = yes ssl_server { cert_file = /etc/letsencrypt/live/<mydomain>/fullchain.pem dh_file = /etc/dovecot/dh.pem key_file = /etc/letsencrypt/live/<mydomain>/privkey.pem } sieve_script personal { active_path = ~/.dovecot.sieve driver = file path = ~/sieve } When connecting my email client to dovecot, the connection succeeds, but after awhile it disconnects. In /var/log/dovecot I get: May 18 17:36:15 imap-login: Warning: Auth process not responding, delayed sending initial response (greeting): user=<>, rip=192.168.178.6, lip=192.168.178.15, session=<2IMbUhlS8K3AqLIG> May 18 17:36:45 imap-login: Info: Login aborted: Connection closed (disconnected before auth was ready, waited 40 secs) (auth_process_not_ready): user=<>, rip=192.168.178.6, lip=192.168.178.15, session=<2IMbUhlS8K3AqLIG> May 18 17:36:45 imap-login: Error: auth-client: conn unix:login: connect(login) in directory / failed: Permission denied (euid=473(<getpwuid() error>) egid=473(<getgrgid() error>) missing +x perm: /, we're not in group 474, dir owned by 0:474 mode=0750) May 18 17:36:45 imap-login: Error: auth-client: conn unix:login: connect(login) in directory / failed: Permission denied (euid=473(<getpwuid() error>) egid=473(<getgrgid() error>) missing +x perm: /, we're not in group 474, dir owned by 0:474 mode=0750) 473 is connected with account/group dovenull an 474 is connected with account/group dovecot I can't find a solution when searching the internet. -- fr.gr. member openSUSE Freek de Kruijf I have the following dovecot.conf, striped from comment: dovecot_config_version = 2.4.2 dovecot_storage_version = 2.4.0 protocols = imap auth_allow_cleartext = yes auth_mechanisms = plain auth_verbose = yes mail_plugin_dir = /usr/lib64/dovecot/modules namespace inbox { inbox = yes mail_driver = maildir mail_path = ~/Maildir prefix = INBOX. mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Trash { auto = subscribe special_use = \Trash } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } } passdb pam { service_name = dovecot } userdb passwd { use_worker = yes } service auth { unix_listener /var/spool/postfix/private/auth { group = postfix mode = 0666 user = postfix } } ssl = yes ssl_server { cert_file = /etc/letsencrypt/live/<mydomain>/fullchain.pem dh_file = /etc/dovecot/dh.pem key_file = /etc/letsencrypt/live/<mydomain>/privkey.pem } sieve_script personal { active_path = ~/.dovecot.sieve driver = file path = ~/sieve } When connecting my email client to dovecot, the connection succeeds, but after awhile it disconnects. In /var/log/dovecot I get: May 18 17:36:15 imap-login: Warning: Auth process not responding, delayed sending initial response (greeting): user=<>, rip=192.168.178.6, lip=192.168.178.15, session=<2IMbUhlS8K3AqLIG> May 18 17:36:45 imap-login: Info: Login aborted: Connection closed (disconnected before auth was ready, waited 40 secs) (auth_process_not_ready): user=<>, rip=192.168.178.6, lip=192.168.178.15, session=<2IMbUhlS8K3AqLIG> May 18 17:36:45 imap-login: Error: auth-client: conn unix:login: connect(login) in directory / failed: Permission denied (euid=473(<getpwuid() error>) egid=473(<getgrgid() error>) missing +x perm: /, we're not in group 474, dir owned by 0:474 mode=0750) May 18 17:36:45 imap-login: Error: auth-client: conn unix:login: connect(login) in directory / failed: Permission denied (euid=473(<getpwuid() error>) egid=473(<getgrgid() error>) missing +x perm: /, we're not in group 474, dir owned by 0:474 mode=0750) 473 is connected with account/group dovenull an 474 is connected with account/group dovecot I can't find a solution when searching the internet. -- fr.gr. member openSUSE Freek de Kruijf

3 4

Regression/Bug report: crypt_private_key_file (for crypt_global_private_key) no longer supports inline base64 (from userdb) in dev branch
by Ömer Güven 22 May '26

22 May '26

Hi, I’ve run into what appears to be a regression (or at least a behavioral change) in Dovecot 3.3.0+ / current HEAD regarding crypt_private_key_file handling for the crypt_global_private_key. I don‘t use per-folder user keys but a global encryption key per user. Previously, it was possible to provide the private key via userdb using an inline base64:-encoded PEM directly to the crypt_global_private_key_file and Dovecot would correctly parse and use it. With the latest config scheme change, this no longer works, as you cannot really set named config variables from userdb. I did workaround by setting a custom userdb variable and setting the needed named/filter variables manually in the config file. To make it work, I had to enclose the config in a „protocol imap“ scope. The crypt_global_public_key_file is accessible from userdb (as it is not within a named filter scope), so inline-base64 works as with Dovecot 2.3+. When using the following workaround to reach the scoped variables, Dovecot tries to load the file „inline:BASE64ENCODEDPEM“ as a literal filename (error open(inline:AB….C==) failed): protocol imap { crypt_global_private_key user { crypt_private_key_file = %{userdb:private_key_file} crypt_private_key_password = %{userdb:private_key_password} } } As a workaround, I now have to: • store the private key as a file (e.g. %{home}/privkey.pem) • pass only the password via userdb • inject the path indirectly via static config + userdb variables This works, but is less elegant: protocol imap { crypt_global_private_key user { crypt_private_key_file = %{home}/privkey.pem crypt_private_key_password = %{userdb:private_key_password} } } Again, with the new named filter mechanism, it’s no longer possible to directly inject such config values via userdb, which makes fully dynamic per-user key provisioning harder. I now have to export all encrypted private keys from the database to the users home directory. Ömer Hi, I've run into what appears to be a regression (or at least a behavioral change) in Dovecot 3.3.0+ / current HEAD regarding crypt_private_key_file handling for the crypt_global_private_key. I don`t use per-folder user keys but a global encryption key per user. Previously, it was possible to provide the private key via userdb using an inline base64:-encoded PEM directly to the crypt_global_private_key_file and Dovecot would correctly parse and use it. With the latest config scheme change, this no longer works, as you cannot really set named config variables from userdb. I did workaround by setting a custom userdb variable and setting the needed named/filter variables manually in the config file. To make it work, I had to enclose the config in a "protocol imap" scope. The crypt_global_public_key_file is accessible from userdb (as it is not within a named filter scope), so inline-base64 works as with Dovecot 2.3+. When using the following workaround to reach the scoped variables, Dovecot tries to load the file "inline:BASE64ENCODEDPEM" as a literal filename (error open(inline:AB....C==) failed): protocol imap { crypt_global_private_key user { crypt_private_key_file = %{userdb:private_key_file} crypt_private_key_password = %{userdb:private_key_password} } } As a workaround, I now have to: o store the private key as a file (e.g. %{home}/privkey.pem) o pass only the password via userdb o inject the path indirectly via static config + userdb variables This works, but is less elegant: protocol imap { crypt_global_private_key user { crypt_private_key_file = %{home}/privkey.pem crypt_private_key_password = %{userdb:private_key_password} } } Again, with the new named filter mechanism, it's no longer possible to directly inject such config values via userdb, which makes fully dynamic per-user key provisioning harder. I now have to export all encrypted private keys from the database to the users home directory. Oemer

4 11

warning: /usr/lib/postfix/sbin/virtual: bad command startup -- throttling
by David Koski 22 May '26

22 May '26

Hello, I have set up a new dovecot mail server using MySQL backend. All users, passwords and related data are in SQL. I'm seeing the following warning in the mail.log file: warning: /usr/lib/postfix/sbin/virtual: bad command startup -- throttling It appears to have no impact on mail delivery. Since paths to user directories are yielded by the SQL database, I'm left wondering why this is even a warning. Regards, David Koski dkoski(a)sutinen.com # 2.4.1-4 (7d8c0e5759): /etc/dovecot/dovecot.conf # Pigeonhole version 2.4.1-4 (0a86619f) # OS: Linux 6.12.69+deb13-amd64 x86_64 Debian 13.5 # Hostname: B914839.online-server.cloud # 4 default setting changes since version 2.4.0 dovecot_config_version = 2.4.0 auth_debug = yes auth_debug_passwords = yes auth_mechanisms = plain login gssapi cram-md5 digest-md5 debug_log_path = /var/log/dovecot/debug.log default_vsz_limit = 512M dovecot_storage_version = 2.4.0 fts_autoindex = yes fts_autoindex_max_recent_msgs = 999 fts_search_add_missing = yes info_log_path = /var/log/dovecot/info.log lda_mailbox_autocreate = yes lda_mailbox_autosubscribe = yes log_debug = category=auth log_path = /var/log/dovecot/dovecot.log mail_driver = maildir mail_plugins { quota = yes } mailbox_directory_name_legacy = no protocols { imap = yes lmtp = yes sieve = yes } quota_exceeded_message = User %{user} has exceeded the storage volume. / User %{user} has exhausted allowed storage space. sieve_plugins { sieve_imapsieve = yes sieve_extprograms = yes } sql_driver = mysql ssl = required namespace inbox { inbox = yes prefix = separator = . type = private mailbox Drafts { auto = subscribe special_use = "\\Drafts" } mailbox Outbox { special_use = "\\Drafts" } mailbox Junk { auto = subscribe special_use = "\\Junk" } mailbox "Junk Email" { auto = no special_use = "\\Junk" } mailbox "Junk E-mail" { auto = no special_use = "\\Junk" } mailbox Trash { auto = subscribe special_use = "\\Trash" } mailbox "Deleted Items" { auto = no special_use = "\\Trash" } mailbox Spam { auto = no special_use = "\\Junk" } mailbox Archives { auto = no special_use = "\\Archive" } mailbox Sent { auto = subscribe special_use = "\\Sent" } mailbox "Sent Items" { auto = no special_use = "\\Sent" } mailbox "Sent Messages" { auto = no special_use = "\\Sent" } } service imap-login { inet_listener imap { } inet_listener imaps { } } service pop3-login { inet_listener pop3 { } inet_listener pop3s { } } service submission-login { inet_listener submission { } inet_listener submissions { } } service lmtp { unix_listener /var/spool/postfix/private/dovecot-lmtp { group = postfix mode = 0600 user = postfix } } service imap { } service pop3 { } service submission { } service auth { unix_listener auth-userdb { group = postfix } unix_listener /var/spool/postfix/private/auth { group = postfix mode = 0660 user = postfix } } service auth-worker { } service dict { unix_listener dict { } } ssl_server { cert_file = /etc/letsencrypt/live/mail.sutinen.com/fullchain.pem key_file = /etc/letsencrypt/live/mail.sutinen.com/privkey.pem } protocol lda { mail_plugins { sieve = yes quota = yes } } protocol imap { mail_plugins { imap_sieve = yes } } protocol lmtp { mail_plugins { quota = yes sieve = yes notify = yes push_notification = yes } postmaster_address = postmaster(a)sutinen.com } service managesieve-login { inet_listener sieve { port = 4190 } inet_listener sieve_deprecated { port = 2000 } } service managesieve { } sieve_script personal { active_path = /var/vmail/sieve/%{user | domain }/%{user | username }/active-script.sieve driver = file path = /var/vmail/sieve/%{user | domain }/%{user | username }/scripts type = personal } sieve_script spam-global { path = /var/vmail/sieve/global/spam-global.sieve type = before } mailbox Spam { sieve_script spam { cause = copy path = /var/vmail/sieve/global/learn-spam.sieve type = before } } imapsieve_from Spam { sieve_script ham { cause = copy path = /var/vmail/sieve/global/learn-ham.sieve type = before } } mysql /var/run/mysqld/mysqld.sock { dbname = vmail password = # hidden, use -P to show it user = vmail } passdb sql { driver = sql query = CALL password_query('%{user}', '%{password}') } userdb sql { driver = sql query = CALL user_query_dovecot('%{user | username}', '%{user | domain}') }

3 2

Dovecot 2.4 authentication
by David Koski 22 May '26

22 May '26

Hello, I just put a new Postfix with Dovecot 2.4 server on line migrating the email from Dovecot 2.3. I have discovered a number of clients (all copiers so far) that no longer authenticate SMTP. Mechanisms configured: auth_mechanisms = plain login gssapi cram-md5 digest-md5 I found that if I disable all encrypted password methods (some copiers list them separately with a checkbox) and leave only plain/login on the client side, then it succeeds. But sometimes it is not necessary. It's almost as if it only tries once and gives up if it fails. I'm not sure how it is actually supposed to work. But it worked with Dovecot 2.3. I did test one account with an encrypted password like "{CRAM-MD5}6e3..." successfully. But other accounts failed. It almost looks like the encryption method must be specified on the client side to match the server before it will work consistently. I know that password encryption is unnecessary with STARTTLS but I have to live with it now. I have noticed that Thunderbird configured with encryption silently falls back to plain/login so I didn't notice a problem. Regards, David Koski dkoski(a)sutinen.com

2 4

dovecot 2.4 Restricting Access - journal/log differences for failed login?
by David C Rankin 20 May '26

20 May '26

All, I have restricted access for several users using the Restring Access documentation https://doc.dovecot.org/2.4.0/howto/restrict_access.html This was to cut down on the brute-force attempts against dovecot for certain users whose usernames/e-mails are routinely used. I have a very simply passdb config: passdb passwd-file { passwd_file_path = /etc/dovecot/deny.%{protocol} deny = yes } passdb pam { driver = pam } userdb pam { driver = passwd } I created the /etc/dovecot/deny.imap file with one username per-line. But what confuses me is the server logs the exact same failure for unsuccessful login after restricting access as it did before restricting access. E.g. 2026-05-20T07:22:43.039585-05:00 valkyrie auth[28618]: pam_unix(dovecot:auth): authentication failure; logname= uid=0 euid=0 tty=dovecot ruser=janice(a)mydomain.com rhost=70.186.179.37 2026-05-20T07:22:51.578711-05:00 valkyrie auth[28618]: pam_unix(dovecot:auth): authentication failure; logname= uid=0 euid=0 tty=dovecot ruser=janice(a)mydomain.com rhost=70.186.179.37 May 20 07:25:36 valkyrie dovecot[1015]: imap-login: Login aborted: Inactivity (auth failed, 2 attempts in 177 secs) (auth_failed): user=<janice(a)mydomain.com>, method=PLAIN, rip=70.186.179.37, lip=192.168.6.14, TLS, session=<ouXP2T5SJpZGurMl> Is this normal, or should dovecot respond to restricted users in some way that differs from a normal login failure? All could be working fine, I'm just not certain about the log entries. -- David C. Rankin, J.D.,P.E.

1 0

"imap" command not usable in dovecot 2.4
by Mark Hills 20 May '26

20 May '26

Since upgrading to 2.4 (from 2.3) the "imap" command no longer works: $ /usr/libexec/dovecot/imap * PREAUTH [CAPABILITY IMAP4rev1 SASL-IR LOGIN-REFERRALS ID ENABLE IDLE SORT SORT=DISPLAY THREAD=REFERENCES THREAD=REFS THREAD=ORDEREDSUBJECT MULTIAPPEND URL-PARTIAL CATENATE UNSELECT CHILDREN NAMESPACE UIDPLUS LIST-EXTENDED I18NLEVEL=1 CONDSTORE QRESYNC ESEARCH ESORT SEARCHRES WITHIN CONTEXT=SEARCH LIST-STATUS BINARY MOVE REPLACE SNIPPET=FUZZY PREVIEW=FUZZY PREVIEW SPECIAL-USE STATUS=SIZE SAVEDATE COMPRESS=DEFLATE INPROGRESS NOTIFY LITERAL+] Logged in as mark imap(mark)<6109><5GlNEEHaDWrdFwAABxE9Dg>: Fatal: Namespace inbox: mail_storage settings: Failed to parse configuration: Invalid setting mail_path=~/Maildir: mail_path setting used home directory (~/) but there is no mail_home and userdb didn't return it Whereas I would expect an authenticated IMAP session. Removing the offending /etc/dovecot.conf line temporarily: dovecot_config_version = 2.4.1 dovecot_storage_version = 2.4.0 protocols = imap lmtp sieve mail_driver = maildir #mail_path = ~/Maildir passdb passwd-file { passwd_file_path = /etc/dovecot/passwd } userdb passwd { } [...] And re-testing suggests the cause runs deeper: $ /usr/lib/exec/dovecot/imap [...] imap(mark)<6364><tetIC6/aDWrcGAAABxE9Dg>: Fatal: Namespace inbox: mail_storage settings: Failed to parse configuration: Failed to expand mail_path setting variables: Setting used home directory (%h) but there is no mail_home and userdb didn't return it Attempts to hard-wire it seem to suggest that usedb is not involved. I imagine this may be not consulted, because the process is pre-authenticated and running as a local UID: userdb passwd { fields_import_all = no fields { mail_home = /home/mark mail_path = /home/mark/Maildir } } $ /usr/libexec/dovecot/imap [...] imap(mark)<6656><qPqUB17cDWoAGgAABxE9Dg>: Fatal: Namespace inbox: mail_storage settings: Failed to parse configuration: Failed to expand mail_path setting variables: Setting used home directory (%h) but there is no mail_home and userdb didn't return it It seems like 2.3 looked up these fields from the user invoking it, but no longer is this the case? Running "imap" like this is very useful, especially when a client like Pine/Alpine can access over SSH. To get to here I had to workaround the missing "shadow" passdb module, which is also useful as Alpine Linux 3.23 neither uses PAM extensively nor provides dovecot built with PAM. -- Mark

1 0

undocumented changed behavior of 2.4.3 (possible bug report)
by Patrick Cernko 20 May '26

20 May '26

Hello Dovecot-Devs, hello list, while upgrading my Dovecot 2.4.2 setup to 2.4.3 I observed two instances of changed behavior. I could not find notes about those issues in the corresponding documentation (https://doc.dovecot.org/2.4.3/installation/upgrade/2.4-to-2.4.x.html) or release notes: 1. I had to change userdb_ldap_iterate_filter = (&(objectClass=posixAccount)%{if(protocol,"eq","pop3","(istIsMemberOf=POP3-users)","")}) to userdb_ldap_iterate_filter = (&(objectClass=posixAccount)%{if(protocol,"eq","pop3","(istIsMemberOf=POP3-users)","")|safe) otherwise the characters '()=' in the "true"-case of the condition got somehow "quoted" with escaped ASCII codes, which led to a broken LDAP filter string. Maybe you could add some documentation (in "Upgrading" as well as "Settings Variables") about that? 2. The behavior of sieve matches using ":matches" has changed: Example email (only the relevant header): List-Id: Dovecot Mailing List <dovecot.dovecot.org> Example Sieve script: if header :matches "List-Id" "*<dovecot*.dovecot.org>" { fileinto "dovecot"; } In 2.4.2 (and earlier) the example email matched the condition and got stored in the dovecot folder, while in 2.4.3 the condition does NOT match. As a workaround I added a '*' at the end of the match ("*<dovecot*.dovecot.org>*") to get my mails stored correctly again. Is this changed behavior for Sieve intentional or a bug? In case of a bug, could you please provide a fix? In case of intentional behavior: * Could you please explain the intention? Why was it necessary? * What would be the correct fix for the sieve example to match a List-Id header that ENDS WITH ".dovecot.org>"? * I assume, that other users on my servers will run into similar problems once I upgrade their servers. Is it possible to detect such problematic matches in Sieve scripts? E.g. "all matches without a trailing '*' are affected", ... This would give me a chance, to inform my users or even fix their scripts in advance. Best regards, -- Patrick Cernko <pcernko(a)mpi-klsb.mpg.de> +49 681 9325 5815 Joint Scientific IT and Technical Service Max-Planck-Institute für Informatik & Softwaresysteme

2 3

flatcurve index size (10x larger than lucene)
by Jean-Daniel Beaubien 18 May '26

18 May '26

Hi everyone, I am currently trying to update from an old dovecot 2.3 setup with fts-lucene to 2.4.2 with fts-flatcurve. However, I am seeing a giant jump in the size of the indexes. For an email account that takes 12GB of storage, I get the following index sizes: - lucene: 270 MB - flatcurve: 2400 MB That is almost a 10x increase in the index size (with flatcurve the indexes represent 20% of the size of the mailbox). Considering I have a mailbox that takes 270GB, I can't imagine the size of the indexes. *Am I doing something wrong with flatcurve? How can the indexes be so much larger than lucene?* I took the default flatcurve configuration from the Ubuntu 26.04 install: mail_plugins { fts = yes fts_flatcurve = yes } fts_autoindex = yes language_filters = normalizer-icu snowball stopwords language_tokenizers = generic email-address language_tokenizer_generic_algorithm = simple language en { default = yes filters = lowercase snowball english-possessive stopwords } fts flatcurve { # All of these are optional, and indicate the default values. # They are listed here for documentation purposes; most people should not # need to define/override in their config. # commit_limit = 500 # max_term_size = 30 # min_term_size = 2 # optimize_limit = 10 # rotate_count = 5000 # rotate_time = 5000 substring_search = yes } *Any help/insight would be greatly appreciated.* Best regards, -J Hi everyone, I am currently trying to update from an old dovecot 2.3 setup with fts-lucene to 2.4.2 with fts-flatcurve. However, I am seeing a giant jump in the size of the indexes. For an email account that takes 12GB of storage, I get the following index sizes: o lucene: 270 MB o flatcurve: 2400 MB That is almost a 10x increase in the index size (with flatcurve the indexes represent 20% of the size of the mailbox). Considering I have a mailbox that takes 270GB, I can't imagine the size of the indexes. Am I doing something wrong with flatcurve? How can the indexes be so much larger than lucene? I took the default flatcurve configuration from the Ubuntu 26.04 install: mail_plugins { fts = yes fts_flatcurve = yes } fts_autoindex = yes language_filters = normalizer-icu snowball stopwords language_tokenizers = generic email-address language_tokenizer_generic_algorithm = simple language en { default = yes filters = lowercase snowball english-possessive stopwords } fts flatcurve { # All of these are optional, and indicate the default values. # They are listed here for documentation purposes; most people should not # need to define/override in their config. # commit_limit = 500 # max_term_size = 30 # min_term_size = 2 # optimize_limit = 10 # rotate_count = 5000 # rotate_time = 5000 substring_search = yes } Any help/insight would be greatly appreciated. Best regards, -J

3 3

[Dovecot] batch convert mbox to maildir
by Netlink Tech 18 May '26

18 May '26

Hello, I have used mb2md for an individual user like this: su ~username cd ~ ./mb2md -m This converted /var/spool/mail/username mbox to /home/users/username/Maildir just as I expected. To convert 100s of users the same way would be a chore (su as each user for permissions, etc.). Is there a way to batch process all /var/spool/mail/$USER to /home/users/$USER/Maildir with the correct permissions/file owners for each user?

4 4

Lua authentication
by John Fawcett 18 May '26

18 May '26

Hi I'm finally getting round to taking up Aki's suggestion here to look into lua. The original requirement was to check the ip reputation giving an OK or KO before proceeding to the login. I'd be interested in any feedback on the dovecot.conf settings and lua script. Also I've encountered a strange caching issue. https://dovecot.org/mailman3/archives/list/dovecot@dovecot.org/thread/E5FLX… To do this I've added a new passdb before the existing sql passdb. I am assuming that I will need to reverse the defaults (in 2.3.21.1) for result_success and result_failure. passdb { driver = lua args = file=/etc/dovecot/passdb.lua blocking=yes result_success = continue result_failure = return-fail } This is my lua test script. function auth_passdb_lookup(req) dovecot.i_info("login request from " .. req.remote_ip .. " user=" .. req.user) if (req.user == "user_to_deny(a)example.com") then dovecot.i_info("deny login") return dovecot.auth.PASSDB_RESULT_USER_UNKNOWN, "" end return dovecot.auth.PASSDB_RESULT_NEXT, "" end I am trying to get this working with this very simple logic before developing the real logic. But basically I am considering two responses from the lua script, one which is OK (returning dovecot.auth.PASSDB_RESULT_NEXT) in which case the next passdb should be consulted for the real password lookup. The other case is a KO (returning dovecot.auth.PASSDB_RESULT_USER_UNKNOWN) where this should be taken as an immediate failure without proceeding to password lookup in the next passd.I have passdb caching disabled. While testing the KO case which seems to be working, if I modify the script changing the user_to_deny(a)example.com so it no longer matches, the script still outputs "deny login". It's not a passdb caching issue since it is going through the "deny login" code despite the user not matching. Could it be that lua scripts are being cached by dovecot? John

1 0

Re: [EXT] Re: postfix/dovecot migration
by bcs＠nanomail.net 16 May '26

16 May '26

Was able to execute the process as root with "doveadm -D backup -Ru tst tcp:10.0.1.212:12354" after updating some destination side configuration options that do not appear to be documented in the link provided, but there are a few errors that I am unable to understand or correct. 1. Debug: open(/proc/self/io) failed: Permission denied - This file is read only for root and permissions cannot be modified. 2. Error: Mailbox INBOX sync: mailbox_delete failed: INBOX can't be deleted. - Not sure why a mailbox would need to be deleted other than the backup is using a delete and recreate to retain only current emails, assuming this is on the destination server. 3. All emails for tst account are deleted from the source server - This is a significant issue when only copies should be made for a backup. Full log is attached. Any additional insights would be greatly appreciated, Bryan May 15, 2026 10:01 AM, "Aki Tuomi" <aki.tuomi(a)open-xchange.com> wrote: > Yeah, you can do migration with dsync or imap between 2.3 and 2.4 > > Aki > >> On 15/05/2026 16:10 EEST Bryan Simmons <bcs(a)nanomail.net> wrote: >> >> Aki, >> >> Thank you for the guidance. Just to confirm, doveadm/imap based >> migration is the section under "Migrating Mailboxes from Another Dovecot >> Server" in the link referenced below since this migration will be >> between servers? >> >> Seems an upgrade from dovecot 2.3.21 (current version on source server) >> to 2.4.2 (installed version on the destination server) would be >> appropriate before migrating any user accounts based on information from >> some of the other posts regarding migration. >> >> Thanks , Bryan >> >> On 5/13/26 13:41, Aki Tuomi via dovecot wrote: >> On 13/05/2026 20:31 EEST Bryan Simmons via dovecot <dovecot(a)dovecot.org> wrote: >> >> Moving postfix and dovecot to a new server including all current users >> and emails. There exists several dovecot files in each user mail directory: >> >> dovecot.index >> dovecot.index.cache >> dovecot.index.log >> dovecot-keywords >> dovecot.list.index.log >> dovecot.mailbox.log >> dovecot-uidlist >> dovecot-uidvalidity >> dovecot-uidvalidity.612d3d34 >> >> Should these files be migrated also or will dovecot recreate them as needed? >> >> Thank you, Bryan >> >> https://doc.dovecot.org/2.4.4/core/admin/migration.html#migrating-mailboxes >> >> I would recommend seeing this. Depending on your mailbox format, it can be important that those >> files are copied, but I would suggest doveadm/imap based migration. >> >> Aki >> >> _______________________________________________ >> dovecot mailing list -- dovecot(a)dovecot.org >> To unsubscribe send an email to dovecot-leave(a)dovecot.org

2 2

Panic: file charset-iconv.c: line 83 (charset_to_utf8_try): assertion failed: (srcleft <= CHARSET_MAX_PENDING_BUF_SIZE)
by Alex Rosenberg 15 May '26

15 May '26

Attached is an LLM-reduced reproduction of the crash in the title. My particular setup is dovecot 2.3.21.1 (d492236fa0) in a FreeBSD jail (13.5). I realize that this is an older release but there is no FreeBSD port/pkg for dovecot 2.4.x yet. The message in the attachment is reduced from an old (2017!) email to one of the LLVM compiler mailing lists. The original malformed email had this header: X-Mailer: Evolution 3.22.5 (3.22.5-1.fc25) The bug occurs when dovecot's FTS indexer processes a MIME part that: 1. Declares charset="UTF-7" 2. Contains base64-encoded content that, when decoded, has bare '+' characters 3. Causes UTF-7 decoder buffer overflow in charset-iconv.c:83 The base64 content decodes to C source code with expressions like: - argc + 4 - state++ - state-- These '+' characters in UTF-7 context cause the decoder's pending buffer to exceed CHARSET_MAX_PENDING_BUF_SIZE, triggering the assertion failure. Alex

4 5

Can't find unicode/ustring.h in test
by mr.thesaw＠gmail.com 14 May '26

14 May '26

I get this error during compilation: make[3]: Entering directory '/home/hacksaw/src/dovecot-2.4.4/src/lib-language' CC test-lang-icu.o In file included from test-lang-icu.c:8: lang-icu.h:4:10: fatal error: unicode/ustring.h: No such file or directory 4 | #include <unicode/ustring.h> | ^~~~~~~~~~~~~~~~~~~ compilation terminated. I have unicode 66, at least as far as Ubuntu is concerned.

2 8

Dovecot 2.4.4 - mail_crypt with per-user Global Keys from LDAP (not Folder Keys)
by luis.neapolis＠gmail.com 14 May '26

14 May '26

Hi everyone, I'm trying to configure Dovecot 2.4.4 with per-user encryption keys stored in LDAP, using Global Keys mode (NOT Folder Keys). My goal: Each user has their own public key and private key (encrypted with the user's login password) stored in LDAP attributes. 1) LMTP should use the public key to encrypt incoming emails. 2) IMAP should use the user's login password to decrypt the private key from LDAP, then decrypt the emails. I want to use Global Keys mode (where keys are provided via userdb), NOT Folder Keys (where Dovecot stores keys in dovecot-attributes files). What works: LMTP receives emails and encrypts them correctly using the user's public key from LDAP and the encrypted files are saved to disk. What doesn't work: IMAP can log in and see the email list (subject, from, date), but the email body never loads. Error: Decryption error: no private key available May 14 18:58:45 host dovecot[295725]: imap(test(a)example.com)<295760><SESSIONID>: Debug: Mailbox INBOX: UID 11: Opened mail because: prefetch May 14 18:58:45 host dovecot[295725]: imap(test(a)example.com)<295760><SESSIONID>: Debug: dict(file): Looking up 'priv/a82ea60599ba056ad47904003a5aeee9/vendor/vendor.dovecot/pvt/crypt/privkeys/aacbcb614fb59f81e9519186b12ac8862858d89fa3bcc1be5f25b886d3e95ee3' May 14 18:58:45 host dovecot[295725]: imap(test(a)example.com)<295760><SESSIONID>: Debug: dict(file): Lookup finished for 'priv/a82ea60599ba056ad47904003a5aeee9/vendor/vendor.dovecot/pvt/crypt/privkeys/aacbcb614fb59f81e9519186b12ac8862858d89fa3bcc1be5f25b886d3e95ee3': not found May 14 18:58:45 host dovecot[295725]: imap(test(a)example.com)<295760><SESSIONID>: Error: Mailbox INBOX: UID 11: read() failed: read(/mail/example.com/test/cur/1778773528.M279604P294174.host.example.com,… failed: Decryption error: no private key available (FETCH BODY[1]) The confusing part: IMAP has the user's password (the user authenticated successfully). The private key in LDAP is encrypted with that same password. So IMAP should be able to decrypt the private key, then decrypt the email. But Dovecot is not using the password it already has to unlock the private key. My configuration: dovecot_config_version = 2.4.4 mail_plugins { quota = yes } # Empty global key block (workaround for LMTP not to fail) crypt_global_private_key main { } passdb ldap { driver = ldap filter = (&(objectClass=Account) (mail=%{user})) fields { mail = %{ldap:mail} password = %{ldap:userPassword} userdb_home = %{ldap:homeDirectory} } } userdb ldap { driver = ldap filter = (&(objectClass=Account) (mail=%{user})) fields { home = %{ldap:homeDirectory} uid = %{ldap:uidNumber|default('500')} gid = %{ldap:gidNumber|default('500')} # Per-user keys from LDAP (Global Keys mode) crypt_global_public_key_file = inline:%{ldap:userPublicKey} crypt_global_private_key/main/crypt_private_key_file = inline:%{ldap:userPrivateKey} crypt_private_key_password = %{password} } } protocol imap { mail_plugins { imap_quota = yes mail_crypt = yes } protocol lmtp { mail_plugins { sieve = yes mail_crypt = yes } } Note: I do NOT have mail_attribute_dict configured. I do NOT want Dovecot to store keys in local files. I want everything to come from LDAP. What I've verified: 1) The private key is valid and decrypts correctly with the user's password using openssl pkey -passin pass:xxx. 2) doveadm user shows the keys are retrieved from LDAP. The error persists. The question: Since IMAP already has the user's password (from the successful login), how can I make Dovecot use that password to decrypt the private key from LDAP and then decrypt the emails? Is there a specific setting I'm missing? Does crypt_private_key_password = %{password} need to be placed somewhere else (e.g., inside the protocol imap block) to work correctly? Also, I see in the logs that Dovecot is still looking for dict(file) keys (Folder Keys mode). How can I completely disable Folder Keys behavior and force Dovecot to only use Global Keys from LDAP? Thank you! Dovecot version: 2.4.4-5 OS: AlmaLinux 9.7 Mail storage: Maildir

1 0

Backup problem
by Doug Hardie 13 May '26

13 May '26

I just migrated my mail server to a new machine. When the new machine was back up and running, I found one mailbox that used to have many emails, now only had 3. I went back to the old machine and found there were only those three messages in the mailbox. I suspect somehow I managed to delete them prior to the backup to the new machine. I have a copy of that mailbox on a backup machine. I just checked and all are there. They have file names like: 1658299911.M966551P8821.mail,S=5001,W=5108:2,S 1664572476.M701757P94428.mail,S=4051,W=4135:2,Sab 1665119988.M609674P37848.mail,S=8455,W=8629:2,Sab 1665757235.M137163P32739.mail,S=5838,W=5950:2,S 1665774793.M299412P37846.mail,S=4669,W=4755:2,S 1665782571.M706364P40033.mail,S=12191,W=12379:2,S The backup machine has all the dovecot files for that user. How can I restore those email from the backup machine to the new server? -- Doug

2 1

Can I use a static userdb or no?
by social+17dw1scw＠lombra.net 13 May '26

13 May '26

Hi, I've been researching this for a while now, and can't find a definitive answer. I have a virtual setup, with globally static UID, GID and mail locations set. The documentation is clear that this lets me use a static userdb instead of SQL. But then, doveadm mailbox list -A requires that a valid iterate_query is set, and that requires an SQL userdb. So, which is it? If querying does not work with a static userdb, that seems like a significant caveat. I'm also curious how this works with prefetch. I understand I can make the passdb query return static values, but I figure since I already have those configured, I shouldn't need to do that. How can I determine whether prefetching is working as intended? Thank you.

3 2

postfix/dovecot migration
by Bryan Simmons 13 May '26

13 May '26

Moving postfix and dovecot to a new server including all current users and emails. There exists several dovecot files in each user mail directory: dovecot.index dovecot.index.cache dovecot.index.log dovecot-keywords dovecot.list.index.log dovecot.mailbox.log dovecot-uidlist dovecot-uidvalidity dovecot-uidvalidity.612d3d34 Should these files be migrated also or will dovecot recreate them as needed? Thank you, Bryan

2 1

Expand dovecot recipients from ldap group
by Maksim Rodin 13 May '26

13 May '26

Hello, I have set up dovecot (2.3.21) to get its recipients from Active Directory. I can successfully get recipients from Active Directory user objects using their sAMAccountName: # cat /etc/dovecot/local.conf ... passdb users { driver = ldap args = /etc/dovecot/dovecot-ldap_domain.conf.ext } userdb users { driver = static args = uid=vmail gid=vmail } ... Here is the relevant part of the ldap config: # cat /etc/dovecot/dovecot-ldap_domain.conf.ext ... user_attrs = \ =home=/vmail/%{ldap:sAMAccountName}, \ =uid=vmail, \ =gid=vmail pass_filter = (&(objectCategory=person)(mail=%u)) user_filter = (&(objectCategory=person)(mail=%u)) ... I can also successfully get user aliases from Active Directory user objects using their otherMailbox attribute: # cat /etc/dovecot/local.conf ... userdb aliases { driver = ldap args = /etc/dovecot/dovecot-ldap_domain_aliases.conf.ext } ... Here is the relevant part of the ldap config: # cat /etc/dovecot/dovecot-ldap_domain_aliases.conf.ext: ... user_attrs = \ =home=/vmail/%{ldap:sAMAccountName}, \ =uid=vmail, \ =gid=vmail pass_filter = (&(objectCategory=person)(otherMailbox=%u)) user_filter = (&(objectCategory=person)(otherMailbox=%u)) Now I am struggling with the task of getting recipients from Active Directory group objects to handle them as distribution groups. Is there a way to expand an Active Directory group object to multiple recipients? I know all of this can be done with postfix ldap module, but I would like to handle all the delivery tasks to dovecot LMTP. -- Best regards Maksim Roin

1 0

Sync'ing two dovecot servers
by Marc Singer 12 May '26

12 May '26

Greetings, I was pleased to see that dovecot has a sync function in doveadm. So far, other tools for moving mailbox contents from one server to another has not been successful. However. Perhaps I don't understand how this should work. I have a new containerized dovecot v2.4.2 # 2.4.2 (0962ed2104): /etc/dovecot/dovecot.conf # OS: Linux 6.12.63+deb13-amd64 x86_64 and I have an existing dovecot running in a VPS: # 2.3.19.1 (9b53102964): /etc/dovecot/dovecot.conf # Pigeonhole version 0.5.19 (4eae2f79) # OS: Linux 6.1.0-35-amd64 x86_64 Debian 12.11 I want to move one user, as a test. I execute this in the container. # doveadm sync -R1 -u USER tcps:OLD_SERVER:993 doveadm(USER): Error: conn OLD_SERVER:993 (IP:993): doveadm server sent invalid handshake: * OK [CAPABILITY IMAP4rev1 SASL-IR LOGIN-REFERRALS ID ENABLE IDLE LITERAL+ AUTH=PLAIN AUTH=LOGIN] Dovecot (Debian) ready. doveadm(USER): Error: cmd sync: Disconnected from remote: Connection closed I'm pretty sure that doveadm sync was able to negotiate the TLS connection. I don't see what it could object to in the sign-on. What am I missing? Cheers -- /Marc Oscar Singer/ *Woollysoft* Dial 817.1823.602.1+ from right to left. Greetings, I was pleased to see that dovecot has a sync function in doveadm. So far, other tools for moving mailbox contents from one server to another has not been successful. However. Perhaps I don't understand how this should work. I have a new containerized dovecot v2.4.2 # 2.4.2 (0962ed2104): /etc/dovecot/dovecot.conf # OS: Linux 6.12.63+deb13-amd64 x86_64 and I have an existing dovecot running in a VPS: # 2.3.19.1 (9b53102964): /etc/dovecot/dovecot.conf # Pigeonhole version 0.5.19 (4eae2f79) # OS: Linux 6.1.0-35-amd64 x86_64 Debian 12.11 I want to move one user, as a test. I execute this in the container. # doveadm sync -R1 -u USER tcps:OLD_SERVER:993 doveadm(USER): Error: conn OLD_SERVER:993 (IP:993): doveadm server sent invalid handshake: * OK [CAPABILITY IMAP4rev1 SASL-IR LOGIN-REFERRALS ID ENABLE IDLE LITERAL+ AUTH=PLAIN AUTH=LOGIN] Dovecot (Debian) ready. doveadm(USER): Error: cmd sync: Disconnected from remote: Connection closed I'm pretty sure that doveadm sync was able to negotiate the TLS connection. I don't see what it could object to in the sign-on. What am I missing? Cheers -- Marc Oscar Singer Woollysoft Dial 817.1823.602.1+ from right to left.

5 7

Dovecot and Pigeonhole v2.4.4 Released
by Aki Tuomi 12 May '26

12 May '26

Hi! We are happy to publish version 2.4.4 of Dovecot and Pigeonhole. These contain CVEs, discovered by external researches. The majority of these have been discovered with help of automated code analysis tools like claude code security, which is why some of these are rather old, missed bugs. No new supported distros have been added or old removed, no new dependencies have been added. Note that there are experimental features in 2.4, one is enabled with `--enable-experimental-mail-utf8`, and another with `--enable-experimental-imap4rev2`, and you also need to set mail_utf8_extensions=yes and imap4rev2_enabled=yes to enable them in config. https://dovecot.org/releases/2.4/dovecot-2.4.4.tar.gz https://dovecot.org/releases/2.4/dovecot-2.4.4.tar.gz.sig https://pigeonhole.dovecot.org/releases/2.4/dovecot-pigeonhole-2.4.4.tar.gz https://pigeonhole.dovecot.org/releases/2.4/dovecot-pigeonhole-2.4.4.tar.gz… Binary packages in https://repo.dovecot.org/ Docker images in https://hub.docker.com/r/dovecot/dovecot --- * CVE-2026-27851: lib-var-expand: Safe filter marks all following pipelines safe. * CVE-2026-33603: auth: CRAM-SHA-*-PLUS channel binding could be faked. MITM attacker with a certificate trusted by the client could have bypassed the requirement for channel binding. * CVE-2026-40020: IMAP folders can be shared-spammed to everyone. * CVE-2026-42006: An attacker can cause uncontrolled memory usage with excessive bracing over IMAP. The fix in CVE-2026-27857 was incomplete. * indexer-worker, quota-status, script-login, program-client-local: Root privileges are now dropped permanently before serving requests. * indexer-worker: Default restart_request_count changed to 1 to work correctly after permanent root privilege drop. * lmtp: Add back service_extra_groups=$SET:default_internal_group that was incorrectly removed in v2.4.3. * master: inet_listener_reuse_port has been replaced by service_reuse_port. The new setting properly pre-creates all listener sockets at startup and assigns one unique socket per process. Using this allows evenly distributing incoming connections to login processes. See https://doc.dovecot.org/latest/core/config/service.html#service_reuse_port for details. - auth: Fix LDAP escaping of 0x13 control character. - auth: Use timing-safe comparison for certificate and public key fingerprints. - fts: Correctly handle internal http-client response errors. - fts: Don't send request to Tika if there is no body text. - fts: Fix address header indexing for RFC 2047 encoded-words. - fts: tika, fts-solr: Fix use-after-free crash during DNS lookup. - imap: Fix assertion panic on invalid REPLACE 0 command. - lib-auth-client: Avoid "unknown id" errors for aborted auth requests. - lib-dcrypt: Fix potential crash if trying to access untrusted/corrupted keys. - lib-dcrypt: Improve error message if keys aren't in hex format as expected. - lib-index: Fix potential crash if fsck fails. - lib-ldap: Fix using OpenLDAP default CA when ssl_client_ca_dir/file is unset. v2.4.3 regression. - lib-master, master: Fix behavior for services with client_limit>1 and restart_request_count so that processes reaching restart_request_count are no longer counted towards process_limit. - lib-master: Fix crash when reaching client_limit with restart_request_count>1. - lib-master: haproxy - Don't trust client certificate common name when HAProxy reports verification failure. - lib-sasl: cram-md5 - Fix out of bounds memory read. - lib-sasl: oauth2 - Fix one byte out of bounds read. - lib-sql: cassandra - Fix reusing Cassandra SSL connections. - lib-sql: sqlite - Fix sqlite_journal_mode=wal to actually work. - lib-storage: Auto-rename non-NFC subscription file entries to NFC on read. - lib-storage: Prevent non-atom SEARCH keywords from causing IMAP command injection. - lib-var-expand-crypt: Return error if hex decoding fails. - lib-var-expand: Fix crash (SIGFPE) with non-positive divisor for / and %. - log: Fix memory leak at deinit. - login-common: When process is full, don't destroy clients waiting on master auth. - login-proxy: Fix crash with rawlog and multiplexing during reconnection. - mail-compress: Fix panic when save method unavailable. - mail-crypt: Fix crash when HMAC-based algorithm is used. - mail-crypt: Use AEAD instead of HMAC with ChaCha20-Poly1305. - mdbox: Create files with O_NOFOLLOW. - push-notification: ox - Fix use-after-free crash during DNS lookup. - quota: quota-status - Limit input buffer size to 1 kB. ––- * CVE-2026-40016: sieve :contains and :matches operators could have been using excessive amount of CPU. Limit the CPU to sieve_max_cpu_time. - Fix potential crashes parsing corrupted Sieve binaries. - lib-sieve: matches - Fix trailing literal match when it fills value exactly. v2.4.3 regression.

2 2