March 2024 - dovecot - dovecot.org

Debug on FreeBSD: tries to open linux style /proc
by Larry Rosenman 23 Mar '24

23 Mar '24

Aki, Can you comment on this FreeBSD bug I got? https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=274540 Either here on in the bugzilla bug. Thanks! -- Larry Rosenman http://www.lerctr.org/~ler Phone: +1 214-642-9640 (c) E-Mail: larryrtx(a)gmail.com US Mail: 5708 Sabbia Dr, Round Rock, TX 78665-2106 Aki, Can you comment on this FreeBSD bug I got? https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=274540 Either here on in the bugzilla bug. Thanks! -- Larry Rosenman http://www.lerctr.org/~ler Phone: +1 214-642-9640 (c) E-Mail: larryrtx(a)gmail.com US Mail: 5708 Sabbia Dr, Round Rock, TX 78665-2106

1 0

Thousands of dovecot/dict processes
by Alex 22 Mar '24

22 Mar '24

Hi, I have a fedora38 system with dovecot-2.3.21 for a few hundred users and just noticed there are 4000 dovecot/dict processes running: 2203938 ? S 0:15 dovecot/dict or sleeping, as the case may be. Some time ago I installed the last_login_dict plugin to record login details per user. Here is my config: plugin { last_login_dict = proxy::sql last_login_key = last-login/%{service}/%{user}/%{remote_ip} last_login_precision = ms } dict { sql = mysql:/etc/dovecot/dovecot-dict-sql.conf.ext } There are about 1100 total entries in the last_login table, most of which are localhost for webmail. Is this all to be expected? I don't know how much in terms of resources these 4000 processes are utilizing, and although the system is virtually idle, it does appear to be going into swap. # free -h total used free shared buff/cache available Mem: 62Gi 9.3Gi 2.5Gi 13Mi 51Gi 52Gi Swap: 15Gi 4.4Gi 11Gi Any ideas greatly appreciated. Thanks, Alex Hi, I have a fedora38 system with dovecot-2.3.21 for a few hundred users and just noticed there are 4000 dovecot/dict processes running: 2203938 ? S 0:15 dovecot/dict or sleeping, as the case may be. Some time ago I installed the last_login_dict plugin to record login details per user. Here is my config: plugin { last_login_dict = proxy::sql last_login_key = last-login/%{service}/%{user}/%{remote_ip} last_login_precision = ms } dict { sql = mysql:/etc/dovecot/dovecot-dict-sql.conf.ext } There are about 1100 total entries in the last_login table, most of which are localhost for webmail. Is this all to be expected? I don't know how much in terms of resources these 4000 processes are utilizing, and although the system is virtually idle, it does appear to be going into swap. # free -h total used free shared buff/cache available Mem: 62Gi 9.3Gi 2.5Gi 13Mi 51Gi 52Gi Swap: 15Gi 4.4Gi 11Gi Any ideas greatly appreciated. Thanks, Alex

1 0

Disable sieve redirect filters?
by jarland＠mxroute.com 21 Mar '24

21 Mar '24

Right now we're using Dovecot for sieve filters. Good old fairly standard config: https://mxbin.io/N6hDUi I'm hoping someone has an idea, but I think I'm probably on my own here. The problem is that this is the best solution to preventing people from spoofing external senders when creating redirects to external recipients: sieve_redirect_envelope_from = recipient But also, as great as this is, it kind of sucks because users continually insist on receiving spam and phishing emails. When a user sends a phishing email, I step in to make sure our network isn't being abused. That's how we do what we do for high quality inbox delivery, we police the hell out of our network. So my problem is I end up with all of these users that I need to review every day, only to find out they just created a redirect filter and "redirected" a phishing email. So my long winded question is, how can I disable redirect filters? I don't want users to be able to create them. But I want them to have access to the rest of the functions of the sieve filtering. Is this possible, or would this need to be an entirely new function contributed to the project's code?

2 1

Notifications to over-quota accounts
by N V 15 Mar '24

15 Mar '24

Hello! I'm trying to allow a system email address to send notifications to over-quota accounts. Is there a way to do it? Thanks in advance! Hello! I'm trying to allow a system email address to send notifications to over-quota accounts. Is there a way to do it? Thanks in advance!

3 2

Question about sieve language
by Lev Serebryakov 15 Mar '24

15 Mar '24

Hello! I need to match all messages sent from some specific domain and all its sub-domains. I can achieve this with: if address :match :domain "From" [ "domain.com", "*.domain.com" ] { ... } But it looks ugly, especially if repeated for 5-6 domains. Is here way better? `:match ["*domain.com"]` will match "notdomain.com" and `:is [ "domain.com" ] doesn't match sub-domains. I understand, that it is (slightly?) off-topic, but I can not find any sieve-specific resources (all google find ins questions about PronotMail and FastMail filters, and these questions are very basic). Thank you. -- // Black Lion AKA Lev Serebryakov

3 4

antispoofing with dovecot submission
by Adrian M 15 Mar '24

15 Mar '24

Hi, it is possible to do antispoofing, like reject_authenticated_sender_login_mismatch from postfix ? Hi, it is possible to do antispoofing, like reject_authenticated_sender_login_mismatch from postfix ?

4 3

How to get a memory pointer in the core process
by Joan Moreau 15 Mar '24

15 Mar '24

Hi Is it possible, from a plugin perspective, to create and recover a pointer in the core process (i.e. memory not lost between 2 calls to the plugin, even after the "deinit" of the plugin" ) ? Thanks Hi Is it possible, from a plugin perspective, to create and recover a pointer in the core process (i.e. memory not lost between 2 calls to the plugin, even after the "deinit" of the plugin" ) ? Thanks

4 11

dovecot panic
by Michael Grant 15 Mar '24

15 Mar '24

I'm running dovecot 2.3.19.1 on debian stable. I just started seeing a slew of these in my log: Mar 14 10:02:38 strange.networkguild.org dovecot[865654]: imap(mgrant)<1939553><DcMIWJ8T6s7ZIx04>: Panic: file istream-header-filter.c: line 663 (i_stream_header_filter_snapshot_free): assertion failed: (snapshot->mstream->snapshot_pending) Mar 14 10:02:38 strange.networkguild.org dovecot[865654]: imap(mgrant)<1939553><DcMIWJ8T6s7ZIx04>: Error: Raw backtrace: #0 test_subprocess_fork[0x7f1a33d7ca20] -> #1 backtrace_append[0x7f1a33d7cc90] -> #2 backtrace_get[0x7f1a33d7ce20] -> #3 execvp_const[0x7f1a33d89f90] -> #4 i_syslog_fatal_handler[0x7f1a33d8a900] -> #5 i_panic[0x7f1a33ce00a4] -> #6 fs_wrapper_unlock[0x7f1a33cdb71d] -> #7 i_stream_snapshot_free[0x7f1a33d96110] -> #8 i_stream_unref[0x7f1a33d96180] -> #9 index_mail_get_virtual_size[0x7f1a33f2d190] -> #10 index_mail_save_finish[0x7f1a33f2d470] -> #11 index_mail_get_special[0x7f1a33f2d690] -> #12 mail_get_special[0x7f1a33eaf950] -> #13 cmd_select[0x560714f983e0] -> #14 imap_fetch_begin[0x560714f9d170] -> #15 imap_fetch_more[0x560714f9d550] -> #16 cmd_fetch[0x560714f91c00] -> #17 command_exec[0x560714f9a880] -> #18 cmd_x_cancel[0x560714fa0510] -> #19 cmd_x_cancel[0x560714fa0510] -> #20 cmd_x_cancel[0x560714fa0510] -> #21 client_handle_input[0x560714fa0880] -> #22 client_input[0x560714fa0d80] -> #23 io_loop_call_io[0x7f1a33da1c70] -> #24 io_loop_handler_run_internal[0x7f1a33da3970] -> #25 io_loop_handler_run[0x7f1a33da3b00] -> #26 io_loop_run[0x7f1a33da3cd0] -> #27 master_service_run[0x7f1a33d14180] -> #28 main[0x560714f8ca00] -> #29 __libc_init_first[0x7f1a33ac51d0] -> #30 __libc_start_main[0x7f1a33ac5280] -> #31 _start[0x560714f8d000] I realize this is a slightly old version of dovecot but this is what's in debian's stable package repository. I see 2.3.21+dfsg1-2 in testing and 2.3.21+dfsg1-3 in unstable. I really try to avoid installing this on production. I didn't recently change anything in my configs on the server. I did start recently using 'eM Client' on Windows a couple days ago, could that have caused this? Mail seems to continue to come in to my imap clients, but these errors in the log are worrying! Michael Grant I'm running dovecot 2.3.19.1 on debian stable. I just started seeing a slew of these in my log: Mar 14 10:02:38 strange.networkguild.org dovecot[865654]: imap (mgrant)<1939553><DcMIWJ8T6s7ZIx04>: Panic: file istream-header-filter.c: line 663 (i_stream_header_filter_snapshot_free): assertion failed: (snapshot- >mstream->snapshot_pending) Mar 14 10:02:38 strange.networkguild.org dovecot[865654]: imap (mgrant)<1939553><DcMIWJ8T6s7ZIx04>: Error: Raw backtrace: #0 test_subprocess_fork[0x7f1a33d7ca20] -> #1 backtrace_append[0x7f1a33d7cc90] - > #2 backtrace_get[0x7f1a33d7ce20] -> #3 execvp_const[0x7f1a33d89f90] -> #4 i_syslog_fatal_handler[0x7f1a33d8a900] -> #5 i_panic[0x7f1a33ce00a4] -> #6 fs_wrapper_unlock[0x7f1a33cdb71d] -> #7 i_stream_snapshot_free[0x7f1a33d96110] -> #8 i_stream_unref[0x7f1a33d96180] -> #9 index_mail_get_virtual_size [0x7f1a33f2d190] -> #10 index_mail_save_finish[0x7f1a33f2d470] -> #11 index_mail_get_special[0x7f1a33f2d690] -> #12 mail_get_special[0x7f1a33eaf950] -> #13 cmd_select[0x560714f983e0] -> #14 imap_fetch_begin[0x560714f9d170] - > #15 imap_fetch_more[0x560714f9d550] -> #16 cmd_fetch[0x560714f91c00] -> #17 command_exec[0x560714f9a880] -> #18 cmd_x_cancel[0x560714fa0510] -> #19 cmd_x_cancel[0x560714fa0510] -> #20 cmd_x_cancel[0x560714fa0510] -> #21 client_handle_input[0x560714fa0880] -> #22 client_input[0x560714fa0d80] -> #23 io_loop_call_io[0x7f1a33da1c70] -> #24 io_loop_handler_run_internal [0x7f1a33da3970] -> #25 io_loop_handler_run[0x7f1a33da3b00] -> #26 io_loop_run [0x7f1a33da3cd0] -> #27 master_service_run[0x7f1a33d14180] -> #28 main [0x560714f8ca00] -> #29 __libc_init_first[0x7f1a33ac51d0] -> #30 __libc_start_main[0x7f1a33ac5280] -> #31 _start[0x560714f8d000] I realize this is a slightly old version of dovecot but this is what's in debian's stable package repository. I see 2.3.21+dfsg1-2 in testing and 2.3.21+dfsg1-3 in unstable. I really try to avoid installing this on production. I didn't recently change anything in my configs on the server. I did start recently using 'eM Client' on Windows a couple days ago, could that have caused this? Mail seems to continue to come in to my imap clients, but these errors in the log are worrying! Michael Grant

2 1

Re: [EXT] Re: How to get a memory pointer in the core process
by Joseph Tam 15 Mar '24

15 Mar '24

From: Joan Moreau > I am trying to avoid closing/ reopening a file pointer to the exact same file > between each call to the plugin Can't you do an end-run around all this by having a persistent standalone process that holds all your persistent data, and the plugin collects session data and passes it to the persistent process via sockets? You'll have to have some initial handshake protocol to establish session context, but this seems the easiest way to accomplish what you want. Joseph Tam <jtam.home(a)gmail.com>

1 0

dovecot and solr clustering
by Maciej Milaszewski 13 Mar '24

13 Mar '24

Hi I have 1 director and ~8 dovecot nodes. I thinking about Solr cluster beacuse one server solr probably is not enough. I thinking about SolrCloud Mode but I don't know if it will work with Dovecot And if not SolrCloud Mode, then what? I don't have much experience in Solr clustering and I want to choose the most optimal solution with scalable aand redundancy (bare metal servers is not problem) Hi I have 1 director and ~8 dovecot nodes. I thinking about Solr cluster beacuse one server solr probably is not enough. I thinking about SolrCloud Mode but I don't know if it will work with Dovecot And if not SolrCloud Mode, then what? I don't have much experience in Solr clustering and I want to choose the most optimal solution with scalable aand redundancy (bare metal servers is not problem)

1 0