- dovecot - dovecot.org

doveadm force-resync crashes at list/mailbox-list-fs-iter.c:507
by Ákos Szőts 29 Jul '25

29 Jul '25

Hi All, In Dovecot 2.4.1 if I start doveadm force-resync with a non-existent mailbox that ends in a slash, the command crashes: gdb -ex run --args doveadm force-resync -A test/ Thread 1 (Thread 0x7f9e5cc4a980 (LWP 442) "doveadm"): #0 fs_list_next_root(ctx=ctx@entry=0x55d248df6898) at list/mailbox-list-fs-iter.c:507 roots= <error reading variable roots (Cannot access memory at address 0x0)> count= <error reading variable count (Cannot access memory at address 0x8)> __func__= "fs_list_next_root" #1 0x00007f9e5dc417c2in fs_list_iter_reset(ctx=0x55d248df6898) at list/mailbox-list-fs-iter.c:830 __func__= <optimized out> #2 fs_list_iter_next(_ctx=0x55d248df6898) at list/mailbox-list-fs-iter.c:912 ctx= <optimized out> ret= 0 __func__= "fs_list_iter_next" #3 0x00007f9e5dc4f8ddin mailbox_list_iter_next_call(ctx=0x55d248df6898) at list/mailbox-list-iter.c:1009 info= <optimized out> set= 0x148d86ee0 ret= <optimized out> #4 mailbox_list_iter_next(ctx=0x55d248df6898) at list/mailbox-list-iter.c:1098 _data_stack_cur_id= 5 info= <optimized out> #5 0x00007f9e5dc50899in mailbox_list_ns_iter_try_next(_ctx=0x55d248df6488, info_r=<synthetic pointer>) at list/mailbox-list-iter.c:616 ctx= 0x55d248df6488 ns= <optimized out> info= <optimized out> suffix= <optimized out> has_children= <optimized out> __func__= <optimized out> #6 mailbox_list_ns_iter_next(_ctx=0x55d248df6488) at list/mailbox-list-iter.c:678 info= 0x0 #7 0x00007f9e5dc4f8ddin mailbox_list_iter_next_call(ctx=0x55d248df6488) at list/mailbox-list-iter.c:1009 info= <optimized out> set= 0x2fffffffd0 ret= <optimized out> #8 mailbox_list_iter_next(ctx=0x55d248df6488) at list/mailbox-list-iter.c:1098 _data_stack_cur_id= 4 info= <optimized out> #9 0x000055d20bc81a20in cmd_force_resync_run(_ctx=0x55d248db57a8, user=0x55d248dc6318) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:369 ctx= 0x55d248db57a8 iter_flags= (MAILBOX_LIST_ITER_RAW_LIST| MAILBOX_LIST_ITER_NO_AUTO_BOXES| MAILBOX_LIST_ITER_STAR_WITHIN_NS| MAILBOX_LIST_ITER_RETURN_NO_FLAGS| MAILBOX_LIST_ITER_FORCE_RESYNC) ns_mask= (MAIL_NAMESPACE_TYPE_PRIVATE| MAIL_NAMESPACE_TYPE_SHARED| MAIL_NAMESPACE_TYPE_PUBLIC) iter= 0x55d248df6488 info= <optimized out> ret= <optimized out> patterns= {0x55d248db5300"test/", 0x0} #10 0x000055d20bc856c2in doveadm_mail_next_user(ctx=ctx@entry=0x55d248db57a8, error_r=error_r@entry=0x7ffdbc3f0778) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:506 _data_stack_cur_id= 3 cctx= 0x55d248db39b8 input= {event_parent= 0x0, service= 0x55d20bcbd810"doveadm", protocol= 0x0, username= 0x55d248db81a0"akos(a)szots.dev", session_id= 0x0, session_id_prefix= 0x0, session_create_ time= 0, local_ip= {family= 0, u= {ip6= {__in6_u= {__u6_addr8= '\000' <repeats 15 times>, __u6_addr16= {0, 0, 0, 0, 0, 0, 0, 0}, __u6_addr32= {0, 0, 0, 0}}}, ip4= {s_addr= 0}}, sco pe_id= 0}, remote_ip= {family= 0, u= {ip6= {__in6_u= {__u6_addr8= '\000' <repeats 15 times>, __u6_addr16= {0, 0, 0, 0, 0, 0, 0, 0}, __u6_addr32= {0, 0, 0, 0}}}, ip4= {s_addr= 0}}, scope_id= 0}, local_port= 0, remote_port= 0, local_name= 0x0, userdb_fields= 0x0, code_override_fields= 0x0, forward_fields= 0x0, set_instance= 0x0, flags_override_add= 0, flags_over ride_remove= 0, no_userdb_lookup= false, debug= false, end_client_tls_secured= false, autocreated= false, no_free_init_failure= false} error= 0x0 ip= <optimized out> ret= <optimized out> dropping_privs= true cur_uid= 0 cur_cwd= 0x55d248d86df8"/" cur_ctx= 0x55d248dc2c80 reason= 0x55d248df57f0 __func__= "doveadm_mail_next_user" #11 0x000055d20bc87525in doveadm_mail_all_users(ctx=0x55d248db57a8, wildcard_user=0x55d20bcc015d "*") at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:584 _data_stack_cur_id= 2 cctx= 0x55d248db39b8 user_idx= 0 ip= <optimized out> user= 0x55d248db81a0"akos(a)szots.dev" error= 0x55d248d8e7d8"\t\350\330H\322U" ret= 1 #12 doveadm_mail_cmd_exec(ctx=0x55d248db57a8, wildcard_user=0x55d20bcc015d "*") at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:721 cctx= <optimized out> cli= <optimized out> ret= <optimized out> error= 0x55d248d8e7d8"\t\350\330H\322U" #13 doveadm_cmd_ver2_to_mail_cmd_wrapper(cctx=<optimized out>) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:1039 mctx= 0x55d248db57a8 mail_cmd= {alloc= 0x55d20bc803f0<cmd_force_resync_alloc>, name= 0x55d20bcbe92d"force-resync", usage_args= 0x55d20bcc5008"[-u <user>|-A] [-S <socket_path>] [-f] <mailbox mask>"} wildcard_user= 0x55d20bcc015d"*" #14 0x000055d20bc9ba77in doveadm_cmdline_run(argc=<optimized out>, argv=0x55d248d8e7c8, cctx=cctx@entry=0x55d248db39b8) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-cmd-parse.c:451 prefix= <optimized out> pargv= {arr= {buffer= 0x55d248db4388, element_size= 112}, v= 0x55d248db4388, v_modifiable= 0x55d248db4388} pargc= 8 pool= 0x55d248db4350 ptr_count= <optimized out> ptr= <optimized out> ptr_end= <optimized out> ptr_last_non_kv= <optimized out> #15 0x000055d20bc9bc2cin doveadm_cmdline_try_run(cmd_name=cmd_name@entry=0x55d248d8e7f9 "force-resync", argc=<optimized out>, argv=<optimized out>, cctx=cctx@entry=0x55d248db39b8) at /usr/s rc/debug/dovecot-2.4.1-4/src/doveadm/doveadm-cmd.c:160 cmd= <optimized out> #16 0x000055d20bc74fdein main(argc=<optimized out>, argv=<optimized out>) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm.c:386 service_flags= <optimized out> cmd_name= 0x55d248d8e7f9"force-resync" i= <optimized out> quick_init= false c= <optimized out> longopts= {{name= 0x7f9e5dadb7a8"help", has_arg= 0, flag= 0x0, val= 0}, {name= 0x0, has_arg= 0, flag= 0x0, val= 0}} help_requested= <optimized out> longopt= 0x0 cctx= 0x55d248db39b8 Ákos Hi All, In Dovecot 2.4.1 if I start doveadm force-resync with a non-existent mailbox that ends in a slash, the command crashes: gdb -ex run --args doveadm force-resync -A test/ Thread 1 (Thread 0x7f9e5cc4a980 (LWP 442) "doveadm"): #0 fs_list_next_root (ctx=ctx@entry=0x55d248df6898) at list/mailbox-list-fs-iter.c:507 roots = <error reading variable roots (Cannot access memory at address 0x0)> count = <error reading variable count (Cannot access memory at address 0x8)> __func__ = "fs_list_next_root" #1 0x00007f9e5dc417c2 in fs_list_iter_reset (ctx=0x55d248df6898) at list/mailbox-list-fs-iter.c:830 __func__ = <optimized out> #2 fs_list_iter_next (_ctx=0x55d248df6898) at list/mailbox-list-fs-iter.c:912 ctx = <optimized out> ret = 0 __func__ = "fs_list_iter_next" #3 0x00007f9e5dc4f8dd in mailbox_list_iter_next_call (ctx=0x55d248df6898) at list/mailbox-list-iter.c:1009 info = <optimized out> set = 0x148d86ee0 ret = <optimized out> #4 mailbox_list_iter_next (ctx=0x55d248df6898) at list/mailbox-list-iter.c:1098 _data_stack_cur_id = 5 info = <optimized out> #5 0x00007f9e5dc50899 in mailbox_list_ns_iter_try_next (_ctx=0x55d248df6488, info_r=<synthetic pointer>) at list/mailbox-list-iter.c:616 ctx = 0x55d248df6488 ns = <optimized out> info = <optimized out> suffix = <optimized out> has_children = <optimized out> __func__ = <optimized out> #6 mailbox_list_ns_iter_next (_ctx=0x55d248df6488) at list/mailbox-list-iter.c:678 info = 0x0 #7 0x00007f9e5dc4f8dd in mailbox_list_iter_next_call (ctx=0x55d248df6488) at list/mailbox-list-iter.c:1009 info = <optimized out> set = 0x2fffffffd0 ret = <optimized out> #8 mailbox_list_iter_next (ctx=0x55d248df6488) at list/mailbox-list-iter.c:1098 _data_stack_cur_id = 4 info = <optimized out> #9 0x000055d20bc81a20 in cmd_force_resync_run (_ctx=0x55d248db57a8, user=0x55d248dc6318) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:369 ctx = 0x55d248db57a8 iter_flags = (MAILBOX_LIST_ITER_RAW_LIST | MAILBOX_LIST_ITER_NO_AUTO_BOXES | MAILBOX_LIST_ITER_STAR_WITHIN_NS | MAILBOX_LIST_ITER_RETURN_NO_FLAGS | MAILBOX_LIST_ITER_FORCE_RESYNC) ns_mask = (MAIL_NAMESPACE_TYPE_PRIVATE | MAIL_NAMESPACE_TYPE_SHARED | MAIL_NAMESPACE_TYPE_PUBLIC) iter = 0x55d248df6488 info = <optimized out> ret = <optimized out> patterns = {0x55d248db5300 "test/", 0x0} #10 0x000055d20bc856c2 in doveadm_mail_next_user (ctx=ctx@entry=0x55d248db57a8, error_r=error_r@entry=0x7ffdbc3f0778) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:506 _data_stack_cur_id = 3 cctx = 0x55d248db39b8 input = {event_parent = 0x0, service = 0x55d20bcbd810 "doveadm", protocol = 0x0, username = 0x55d248db81a0 [1]"akos(a)szots.dev", session_id = 0x0, session_id_prefix = 0x0, session_create_ time = 0, local_ip = {family = 0, u = {ip6 = {__in6_u = {__u6_addr8 = '\000' <repeats 15 times>, __u6_addr16 = {0, 0, 0, 0, 0, 0, 0, 0}, __u6_addr32 = {0, 0, 0, 0}}}, ip4 = {s_addr = 0}}, sco pe_id = 0}, remote_ip = {family = 0, u = {ip6 = {__in6_u = {__u6_addr8 = '\000' <repeats 15 times>, __u6_addr16 = {0, 0, 0, 0, 0, 0, 0, 0}, __u6_addr32 = {0, 0, 0, 0}}}, ip4 = {s_addr = 0}}, scope_id = 0}, local_port = 0, remote_port = 0, local_name = 0x0, userdb_fields = 0x0, code_override_fields = 0x0, forward_fields = 0x0, set_instance = 0x0, flags_override_add = 0, flags_over ride_remove = 0, no_userdb_lookup = false, debug = false, end_client_tls_secured = false, autocreated = false, no_free_init_failure = false} error = 0x0 ip = <optimized out> ret = <optimized out> dropping_privs = true cur_uid = 0 cur_cwd = 0x55d248d86df8 "/" cur_ctx = 0x55d248dc2c80 reason = 0x55d248df57f0 __func__ = "doveadm_mail_next_user" #11 0x000055d20bc87525 in doveadm_mail_all_users (ctx=0x55d248db57a8, wildcard_user=0x55d20bcc015d "*") at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:584 _data_stack_cur_id = 2 cctx = 0x55d248db39b8 user_idx = 0 ip = <optimized out> user = 0x55d248db81a0 [2]"akos(a)szots.dev" error = 0x55d248d8e7d8 "\t\350\330H\322U" ret = 1 #12 doveadm_mail_cmd_exec (ctx=0x55d248db57a8, wildcard_user=0x55d20bcc015d "*") at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:721 cctx = <optimized out> cli = <optimized out> ret = <optimized out> error = 0x55d248d8e7d8 "\t\350\330H\322U" #13 doveadm_cmd_ver2_to_mail_cmd_wrapper (cctx=<optimized out>) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-mail.c:1039 mctx = 0x55d248db57a8 mail_cmd = {alloc = 0x55d20bc803f0 <cmd_force_resync_alloc>, name = 0x55d20bcbe92d "force-resync", usage_args = 0x55d20bcc5008 "[-u <user>|-A] [-S <socket_path>] [-f] <mailbox mask>"} wildcard_user = 0x55d20bcc015d "*" #14 0x000055d20bc9ba77 in doveadm_cmdline_run (argc=<optimized out>, argv=0x55d248d8e7c8, cctx=cctx@entry=0x55d248db39b8) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm-cmd-parse.c:451 prefix = <optimized out> pargv = {arr = {buffer = 0x55d248db4388, element_size = 112}, v = 0x55d248db4388, v_modifiable = 0x55d248db4388} pargc = 8 pool = 0x55d248db4350 ptr_count = <optimized out> ptr = <optimized out> ptr_end = <optimized out> ptr_last_non_kv = <optimized out> #15 0x000055d20bc9bc2c in doveadm_cmdline_try_run (cmd_name=cmd_name@entry=0x55d248d8e7f9 "force-resync", argc=<optimized out>, argv=<optimized out>, cctx=cctx@entry=0x55d248db39b8) at /usr/s rc/debug/dovecot-2.4.1-4/src/doveadm/doveadm-cmd.c:160 cmd = <optimized out> #16 0x000055d20bc74fde in main (argc=<optimized out>, argv=<optimized out>) at /usr/src/debug/dovecot-2.4.1-4/src/doveadm/doveadm.c:386 service_flags = <optimized out> cmd_name = 0x55d248d8e7f9 "force-resync" i = <optimized out> quick_init = false c = <optimized out> longopts = {{name = 0x7f9e5dadb7a8 "help", has_arg = 0, flag = 0x0, val = 0}, {name = 0x0, has_arg = 0, flag = 0x0, val = 0}} help_requested = <optimized out> longopt = 0x0 cctx = 0x55d248db39b8 Akos References Visible links 1. mailto:akos@szots.dev 2. mailto:akos@szots.dev

2 2

"lock/disable" user's mailbox
by roughnecks 29 Jul '25

29 Jul '25

Hello, not sure if relevant here, but I'm in need of disabling or locking down an user's mailbox, without completely removing the related system account. It would be nice if said account was not able to receive emails anymore, or "better", until it was re-enabled. Would also be cool if the sender received a warning back, upon the failed dispatch. Does that even make sense? :D I'm running postfix and dovecot on a debian bookworm VPS. Thanks -- roughnecks

2 2

When does dovecot use sendmail_path?
by Joseph Tam 29 Jul '25

29 Jul '25

I'm trying to plug up all the holes where local mail leaves our mail system without a DKIM signature. One avenue is bounce messages. I wrote a small sendmail wrapper to add DKIM signatures, and defined it as the value for dovecot's sendmail_path. I presume Dovecot invokes this executable when it has to send a send message. Under what circumstances does this happen? I tried to provoke a mailout by sending messages to recipients with full quotas and permission locked mailbox files, but could not generate any mail back. Joseph Tam <jtam.home(a)gmail.com>

2 2

"imap-login" crashes at lib-ssl-iostream/iostream-ssl.c:55
by Ákos Szőts 29 Jul '25

29 Jul '25

Hi all, When I start "imap-login" without arguments, it crashes: gdb -ex run --args /usr/lib/dovecot/imap-login (gdb) thread apply all bt full Thread 1 (Thread 0x7f5bd50c6740 (LWP 625) "imap-login"): #0 0x00007f5bd545fe97in io_stream_ssl_global_init(set=0x56205ca355c8, error_r=0x7ffdf0e669a8) at ../lib-ssl-iostream/iostream-ssl.c:55 No locals. #1 0x00007f5bd558d15bin login_ssl_init() at /usr/src/debug/dovecot-2.4.1-4/src/login-common/main.c:343 ssl_set= 0x56205ca355c8 error= 0xc00000<error: Cannot access memory at address 0xc00000> #2 main_preinit() at /usr/src/debug/dovecot-2.4.1-4/src/login-common/main.c:355 max_fds= <optimized out> __func__= <optimized out> #3 login_binary_run(binary=<optimized out>, argc=<optimized out>, argv=<optimized out>) at /usr/src/debug/dovecot-2.4.1-4/src/login-common/main.c:528 service_flags= <optimized out> login_socket= 0x7f5bd5590b66"login" error= 0x1000000<error: Cannot access memory at address 0x1000000> c= <optimized out> input= {config_path= 0x0, config_fd= 0, preserve_environment= false, preserve_user= false, preserve_home= false, no_service_filter= false, check_full_config= false, hide_obsol ete_warnings= false, no_protocol_filter= false, no_key_validation= false, reload_config= false, never_exec= false, always_exec= false, return_config_fd= false, use_sysexits= false, pr otocol= 0x562033d837ac"imap"} output= {specific_protocols= 0x56205ca15398, config_fd= -1, permission_denied= false} #4 0x00007f5bd51662aein __libc_start_call_main() from /lib64/libc.so.6 No symbol table info available. #5 0x00007f5bd5166379in __libc_start_main_impl() from /lib64/libc.so.6 No symbol table info available. #6 0x0000562033d79885in _start() at ../sysdeps/x86_64/start.S:115 No locals. Probably "ssl_vfuncs->[...]" is a NULL dereference? I'm debugging a crash within Dovecot and this was one of my findings. Dovecot version: 2.4.1. Ákos

1 0

'doveadm' error
by Andreas Kossack 28 Jul '25

28 Jul '25

Hi, are there anywhere recommendations or 'best practices' for compiling dovecot-2.4.x at IBMs AIX7.3? With some own patches i can compile it, but one unexpected result is # /usr/local/bin/doveadm mailbox list -u <user> /usr/local/bin/doveconf: invalid option -- u What's wrong? -- Mit freundlichen Grüßen / Best regards / Salutations / Cordialmente ___________________________________________________________________ Brandenburgische Technische Universität Cottbus – Senftenberg E Andreas.Kossack(a)B-TU.De

1 0

one-off doveadm commands broken in 2.4.1 when using docker
by jlh 28 Jul '25

28 Jul '25

Hi all In 2.3, I could use doveadm from the docker image as a standalone tool, but this no longer works in 2.4: $ docker run --rm -ti dovecot/dovecot:2.3-latest doveadm usage: doveadm [-Dv] [-f <formatter>] <command> [<args>] altmove [-u <user>|-A] [-S <socket_path>] [-r] <search query> [...snip...] $ docker run --rm -ti dovecot/dovecot:latest-2.4 doveadm doveconf: Fatal: Cannot expand variable %{env:USER_PASSWORD This was useful for hashing passwords with "doveadm pw". I assume this wasn't an intentional change in how doveadm works. For now I'll just use the old image when hashing passwords. Thank you! Jean-Luc

2 1

siebe settings in 2.4.0
by Helmut K. C. Tessarek 23 Jul '25

23 Jul '25

I've been reading the docs and the migration docs for hours, but the sieve config is a mess. In 2.3.x I am using: plugin { sieve = ~/.sieve sieve_dir = ~/.sieve_storage } and this is how it looks in my home dir: $ ls -la .sie* lrwxrwxrwx 32 a b c d .sieve -> .sieve_storage/MailFilters.sieve .rw------- 0 a b c d .sieve.log .rw------- 17,414 a b c d 2025-07-08 03:00 .sieve.svbin So How do I do the same in 2.4.x? I've come so far as: sieve_script personal { driver = file } But then there are path, bin_path, active_path, name, and a bunch of other settings. Sorry to be a little bit frustrated, but my old settings are not out of the ordinary (2 freaking parameters), but there is no documentation on how to migrate these 2 very simple settings to the new format. The upgrade docs explain the new var expansion and a matrix with old and new parameters. But this info is completely missing for sieve. The doc only points to "script storage". Really? You couldn't have added these 2 parameters to the old-new matrix? I am not asking you to do this for every single sieve parameter, but at least for the 2 or 3 most used ones. Sorry for my rant, but I had to get it out there to not go insane. Cheers, K. C. -- regards Helmut K. C. Tessarek KeyID 0x172380A011EF4944 Key fingerprint = 8A55 70C1 BD85 D34E ADBC 386C 1723 80A0 11EF 4944 /* Thou shalt not follow the NULL pointer for chaos and madness await thee at its end. */

4 6

discarded vacation reply to own address <xxxx> (as specified using :addresses argument)
by Marc 23 Jul '25

23 Jul '25

Did I configure this? Or is this default behaviour of dovecot? How can I turn this off if this is a standard config?

2 1

Re: Strange mailq errors
by Xavier Humbert 23 Jul '25

23 Jul '25

Le 22/07/2025 16:28, Odhiambo Washington a écrit : > > > On Tue, Jul 22, 2025 at 4:42 PM Xavier Humbert via dovecot > <dovecot(a)dovecot.org> wrote: > > Hi, > > I recently experienced strange errors in mailq : > > [root@numenor ~]# mailq > -Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient------- > 56BF31C9391 2320 Fri Jul 18 03:42:33 info(a)quora.com > (temporary failure. Command output: lda(postfix): Error: > net_connect_unix(/var/run/dovecot/stats-writer) failed: Permission > denied) > postfix(a)groumpf.org > > All other messages are getting in without error. > Furthermore these are probably spam which shouldn't get in, since > postfix@ is not in the authorized recipients > > When it happens, it is always spam for postfix@ > > Any idea ? > > > Check your dovecot config for the permissions: Postfix user needs > access to /var/run/dovecot/stats-writer > > V2.4: > service stats { > unix_listener stats-writer { > mode = 0660 > user = ?? > } > } > > v2.3 > unix_listener stats-writer { > user = ?? > group = ?? > mode = 0660 > } > } > > > -- > Best regards, > Odhiambo WASHINGTON, > Nairobi,KE > +254 7 3200 0004/+254 7 2274 3223 > In an Internet failure case, the #1 suspect is a constant: DNS. > "Oh, the cruft.", egrep -v '^$|^.*#' ¯\_(ツ)_/¯ :-) > [How to ask smart questions: > http://www.catb.org/~esr/faqs/smart-questions.html] Here is the relevant part of dovecot config : service stats { unix_listener stats-reader { user = $default_internal_user group = $default_internal_user mode = 0660 } unix_listener stats-writer { user = $default_internal_user group = $default_internal_user mode = 0660 } } All other mails get in without problem. I suspect it gets in, then confuses the mail system because of proxy:unix:passwd.byname in postfix config local_recipient_maps I don't need this btw. Xavier -- Xavier HUMBERT - Unix/Win/MacOSX Sysadmin/Network Engineer https://www.amdh.fr Le 22/07/2025 16:28, Odhiambo Washington a ecrit : On Tue, Jul 22, 2025 at 4:42PM Xavier Humbert via dovecot <[1]dovecot(a)dovecot.org> wrote: Hi, I recently experienced strange errors in mailq : [root@numenor ~]# mailq -Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient------- 56BF31C9391 2320 Fri Jul 18 03:42:33 [2]info(a)quora.com (temporary failure. Command output: lda(postfix): Error: net_connect_unix(/var/run/dovecot/stats-writer) failed: Permission denied) [3]postfix(a)groumpf.org All other messages are getting in without error. Furthermore these are probably spam which shouldn't get in, since postfix@ is not in the authorized recipients When it happens, it is always spam for postfix@ Any idea ? Check your dovecot config for the permissions: Postfix user needs access to /var/run/dovecot/stats-writer V2.4: service stats { unix_listener stats-writer { mode = 0660 user = ?? } } v2.3 unix_listener stats-writer { user = ?? group = ?? mode = 0660 } } -- Best regards, Odhiambo WASHINGTON, Nairobi,KE +254 7 3200 0004/+254 7 2274 3223 In an Internet failure case, the #1 suspect is a constant: DNS. "Oh, the cruft.", egrep -v '^$|^.*#' -\_(Tu)_/- :-) [How to ask smart questions: [4]http://www.catb.org/~esr/faqs/smart-questions.html] Here is the relevant part of dovecot config : service stats { unix_listener stats-reader { user = $default_internal_user group = $default_internal_user mode = 0660 } unix_listener stats-writer { user = $default_internal_user group = $default_internal_user mode = 0660 } } All other mails get in without problem. I suspect it gets in, then confuses the mail system because of proxy:unix:passwd.byname in postfix config local_recipient_maps I don't need this btw. Xavier -- Xavier HUMBERT - Unix/Win/MacOSX Sysadmin/Network Engineer [5]https://www.amdh.fr References Visible links 1. mailto:dovecot@dovecot.org 2. mailto:info@quora.com 3. mailto:postfix@groumpf.org 4. http://www.catb.org/~esr/faqs/smart-questions.html 5. https://www.amdh.fr/

2 1

Anyone have working examples of how one might go about *debugging* a userdb/passdb LUA script for authentication?
by Andy Baugh 22 Jul '25

22 Jul '25

Howdy, I've been working on upgrading the version cPanel ships to Dovecot 2.4. It appears that the "dict" driver for passdb/authdb has been removed per your documentation, with a recommendation to use a lua script instead. As such I've followed the online documentation to do so and wound up with a userdb/passdb section that looks like so: passdb cpauthd { driver = lua lua_file = /usr/local/cpanel/etc/dovecot/cpauthd.lua lua_settings { socket = /usr/local/cpanel/var/cpdoveauthd.sock } result_internalfail = continue result_failure = return-fail } userdb cpauthd { driver = lua lua_file = /usr/local/cpanel/etc/dovecot/cpauthd.lua } My prospective lua script only takes from the socket arg passed in and reads from the socket, delivering back the same kind of answer we would have previously got from the auth dict driver (converting password_key/user_key to auth_passdb/userdb_get_cache_key in script as appears to be suggested). Unfortunately, I'm thinking that this entire setup is being ignored, as when I do: doveadm -D -o auth_debug=yes auth lookup for a user I know to be valid on the system with the previous 2.3 setup, I don't even see any indication that the lua authentication database is even consulted. Jul 18 12:32:30 Debug: auth-master: passdb lookup(zippy(a)cptest.local): Started passdb lookup Jul 18 12:32:30 Debug: auth-master: conn unix:/var/run/dovecot/auth-userdb: Connecting Jul 18 12:32:30 Debug: auth-master: conn unix:/var/run/dovecot/auth-userdb (pid=800187,uid=0): Client connected (fd=9) Jul 18 12:32:30 Debug: auth-master: conn unix:/var/run/dovecot/auth-userdb (pid=800187,uid=0): auth input: Jul 18 12:32:30 Error: auth-master: passdb lookup(zippy(a)cptest.local): Auth PASS lookup failed Jul 18 12:32:30 Debug: auth-master: passdb lookup(zippy(a)cptest.local): auth PASS input: Jul 18 12:32:30 Debug: auth-master: passdb lookup(zippy(a)cptest.local): Passdb lookup failed Jul 18 12:32:30 Error: cmd auth lookup: passdb lookup failed for zippy(a)cptest.local Jul 18 12:32:30 Debug: auth-master: conn unix:/var/run/dovecot/auth-userdb (pid=800187,uid=0): Disconnected: Connection closed (fd=9) Similarly, in my lua script, I try to use the "dovecot.auth.log_debug/log_info" etc. functions in documentation within the scope of auth_passdb_lookup or auth_userdb_lookup, yet I never see any debug prints even if I add a debug line to the start of the function to indicate entry. What I would like to know is whether there is some debug flag I'm missing to see the routing from config -> lua script. Does this actually exist? I didn't see anything enlightening in the man pages or online documentation. Hopefully I'm just missing something simple. Cheers, -- Thomas "Andy" Baugh | Software Development Engineer IV andy.baugh(a)webpros.com www.webpros.com

1 1