April 2018 - dovecot - dovecot.org

sieve trace directory: error
by André Rodier 15 May '18

15 May '18

Hello, I am using a custom directory for sieve logs and trace and it apparently generate an error, in some corner case. - I have an IMAP global sieve script, that is run when appending emails into the "Archives" folder. - I have Thunderbird configured to archive emails by year When archiving an email, for instance in "Archives/2018", I often have an error message like this: Mar 31 08:03:02 homebox dovecot[14078]: imap(andre): Error: sieve: trace: creat(/home/users/andre/mails/sieve/logs//andre.Archives/2018.3.20180331-080302.14235.5.trace) failed: No such file or directory I have this configuration (amongst) for sieve logs: sieve_user_log = ~/mails/sieve/logs/ sieve_trace_dir = ~/mails/sieve/logs/ sieve_trace_level = commands sieve_trace_debug = yes sieve_trace_addresses = yes Did I a misconfigured something? Thanks for your help, André -- https://github.com/progmaticltd/homebox

3 5

mmap failed messages in syslog
by Shawn Heisey 10 May '18

10 May '18

I'm getting mmap failed messages in syslog on version 1.2.15 for dovecot.index.cache files. This old dovecot version is installed with debian packages. The OS and all software is 64-bit. I've found info on this log message, with a suggested fix of raising vsz_limit ... but this seems to be a config option for dovecot 2.x, not 1.x. The file it complains about being unable to mmap is only 90MB, and the biggest cache file I have found on the system is only 180MB. The system has 4GB of total memory with about 1GB of that allocated to OS disk cache right now. The mmap functionality provided by the OS doesn't use *actual* memory, so the amount of memory I have should be irrelevant. Checking 'ulimit -a' with several users shows that vmemory is unlimited for all of them. How do I go about allowing docevot 1.x to use more virtual memory? I do have a migration planned to Ubuntu 16, where I can run a 2.x version, but that's not going to happen for a while. Thanks, Shawn

2 3

v2.2.36 release candidate released
by Timo Sirainen 04 May '18

04 May '18

https://dovecot.org/releases/2.2/rc/dovecot-2.2.36.rc1.tar.gz https://dovecot.org/releases/2.2/rc/dovecot-2.2.36.rc1.tar.gz.sig v2.2.36 is hopefully going to be the last v2.2.x release. Please test this RC well, so we'll have a good final release! v2.3.2 is still going to take a couple of months before it's ready. * login-proxy: If ssl_require_crl=no, allow revoked certificates. Also don't do CRL checks for incoming client certificates. * stats plugin: Don't temporarily enable PR_SET_DUMPABLE while opening /proc/self/io. This may still cause security problems if the process is ptrace()d at the same time. Instead, open it while still running as root. + doveadm: Added mailbox cache decision&remove commands. See doveadm-mailbox(1) man page for details. + doveadm: Added rebuild attachments command for rebuilding $HasAttachment or $HasNoAttachment flags for matching mails. See doveadm-rebuild(1) man page for details. - dsync: Lock file generation failed if home directory didn't exist - Snippet generation for HTML mails didn't ignore &entities inside blockquotes, producing strange looking snippets. - imapc: Fix assert-crash if getting disconnected and after reconnection all mails in the selected mailbox are gone. - pop3c: Handle unexpected server disconnections without assert-crash - fts: If mails contained NUL characters, the text around it wasn't indexed. - Obsolete dovecot.index.cache offsets were sometimes used. Trying to fetch a field that was just added to cache file may not have always found it.

3 4

Corrupted dbox file around offset= EOF reading msg header
by DurgaPrasad - DatasoftComnet 02 May '18

02 May '18

Hello Dovecoters, I am using dovecot-2.2.10-8.el7.x86_64 on RHEL 7.4 When I backup mdboxes (using SIS) using this command doveadm -Dv -o plugin/quota= sync -1f -u user1(a)user.com mdbox:/mailbackup/user1 mails get backedup but all m files are not copied. Last file copied is m.4 If I run the command again I get below error I tried with doveadm backup command as well (many times). Error: Corrupted dbox file / mailbackup/user1/storage/m.4 (around offset=3732536): EOF reading msg header (got 0/30 bytes) Result is all mails are not backedup. Destination is a folder on a NAS with NFS mapped to the server. Thanks/DP --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

1 2

Panic when using doveadm fts lookup
by André Rodier 29 Apr '18

29 Apr '18

Hello, I am actually testing dovecot fts search, using doveadm fts lookup. I run this command: doveadm fts lookup -u andre subject "Test" And this generate this error: doveadm(andre): Panic: file mail-storage.c: line 1694 (mailbox_get_open_status): assertion failed: (box->opened) doveadm(andre): Error: Raw backtrace: /usr/lib/dovecot/libdovecot.so.0(+0x95e92) [0x7fa7ac9e8e92] -> /usr/lib/dovecot/libdovecot.so.0(+0x95f19) [0x7fa7ac9e8f19] -> /usr/lib/dovecot/libdovecot.so.0(i_fatal+0) [0x7fa7ac97ea91] -> /usr/lib/dovecot/libdovecot-storage.so.0(mailbox_get_open_status+0x68) [0x7fa7accb5f68] -> /usr/lib/dovecot/modules/lib21_fts_solr_plugin.so(+0x494d) [0x7fa7ab90394d] -> /usr/lib/dovecot/modules/lib20_fts_plugin.so(fts_backend_lookup+0x4d) [0x7fa7abf6861d] -> /usr/lib/dovecot/modules/doveadm/lib20_doveadm_fts_plugin.so(+0x2821) [0x7fa7aabff821] -> doveadm(+0x28fcd) [0x5617545e8fcd] -> doveadm(+0x29ba8) [0x5617545e9ba8] -> doveadm(doveadm_cmd_ver2_to_mail_cmd_wrapper+0x21c) [0x5617545ea9ec] -> doveadm(doveadm_cmd_run_ver2+0x555) [0x5617545f9d25] -> doveadm(doveadm_cmd_try_run_ver2+0x37) [0x5617545f9d77] -> doveadm(main+0x1da) [0x5617545d9afa] -> /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf1) [0x7fa7ac5d42e1] -> doveadm(_start+0x2a) [0x5617545d9eda] Do I have to do anything special to test the dovecot fts index with doveadm lookup? I have not found the documentation on the dovecot web site. Thanks for your advices. -- André Rodier. https://github.com/progmaticltd/homebox

1 0

Listening on multiple ports
by Derek Harding 28 Apr '18

28 Apr '18

Hi, I want dovecot to listen for imaps (ssl) on both 993 and a non-standard port 8839. I've tried using variations on the inet_listener entries in 10-master.conf as shown in a couple of threads on here but none of the attempts have worked. Can someone provide me with a cut-and-paste solution and confirm into which file it should be placed (i.e. is the dovecot.conf format using the inet_listener = {address} appropriate)? -- Best wishes, Derek

3 6

Virtual mailbox doesn't refresh without deleting indexes
by Bret Martin 26 Apr '18

26 Apr '18

Hello, I have a virtual mailbox with the following dovecot-virtual file: mdbox/* -mdbox/Archive/* -mdbox/Junk -mdbox/Notes all younger 604800 not deleted This works at first, but after messages are older than 7 days, they continue to appear in this mailbox forever. Even if I close all IMAP clients and reconnect, I can't get the mailbox to omit the older-than-7-days messages without deleting its indexes. I have other virtual mailboxes that update as expected, but I don't think any of those are date-based. For example, this one "refreshes" as I'd expect: mdbox/* -mdbox/Archive/* -mdbox/Junk flagged Am I doing something wrong with the first virtual mailbox I mentioned above? I'm using dovecot 2.2.13 on Debian GNU/Linux. Thanks! --Bret

1 0

stripping dovecot...
by Rupert Gallagher 26 Apr '18

26 Apr '18

Hello, I do not need the following, and have configured both source and runtime accordingly: pam, gssapi, zlib, bzlib, lzma, pop3, quota, fts, welcome, lda. Therefore, is it *safe* to manually remove the following from /usr/lib/dovecot/? lib05_pop3_migration_plugin.a lib05_pop3_migration_plugin.la lib05_pop3_migration_plugin.so lib20_fts_plugin.a lib20_fts_plugin.la lib20_fts_plugin.so lib21_fts_squat_plugin.a lib21_fts_squat_plugin.la lib21_fts_squat_plugin.so libdovecot-fts.0.dylib libdovecot-fts.a libdovecot-fts.dylib libdovecot-fts.la lib20_doveadm_fts_plugin.a lib20_doveadm_fts_plugin.la lib20_doveadm_fts_plugin.so lib10_quota_plugin.a lib10_quota_plugin.la lib10_quota_plugin.so lib11_imap_quota_plugin.a lib11_imap_quota_plugin.la lib11_imap_quota_plugin.so lib20_quota_clone_plugin.a lib20_quota_clone_plugin.la lib20_quota_clone_plugin.so lib10_doveadm_quota_plugin.a lib10_doveadm_quota_plugin.la lib10_doveadm_quota_plugin.so lib99_welcome_plugin.a lib99_welcome_plugin.la lib99_welcome_plugin.so libdovecot-lda.0.dylib libdovecot-lda.a libdovecot-lda.dylib libdovecot-lda.la

1 1

director stuck in inifite loop on 2.2.35
by Jan-Pieter Cornet 26 Apr '18

26 Apr '18

Occasionally, all of our directors get stuck in an endless loop. Likely triggered by some network event, but I haven't been able to isolate exactly what triggers it, and am unable to reproduce at the moment. When this happens, it makes the dovecot/director process on all directors unresponsive, making login impossible. Restarting dovecot "fixes" it. Here's the info collected from the stuck process. I hope that a dovecot dev could use this to either fix a bug in the async IO code or point to a possible external source that could have triggered this. Would it help to compile dovecot with -g and run a version with debugging info, to be able to inspect more when the crash happens again? "strace" showed this endless loop: 09:27:52.837960 epoll_ctl(14, EPOLL_CTL_DEL, 213, 0x7ffe8e642cdc) = 0 09:27:52.837993 epoll_ctl(14, EPOLL_CTL_ADD, 213, {EPOLLIN|EPOLLPRI|EPOLLERR|EPOLLHUP, {u32=3317538000, u64=94320799358160}}) = 0 09:27:52.838035 epoll_ctl(14, EPOLL_CTL_DEL, 213, 0x7ffe8e642cdc) = 0 09:27:52.838070 epoll_ctl(14, EPOLL_CTL_ADD, 213, {EPOLLIN|EPOLLPRI|EPOLLERR|EPOLLHUP, {u32=3317538000, u64=94320799358160}}) = 0 "ltrace" showed the same loop as: 10:38:01.773867 timeout_add(0x7530, 0x55c8c351df75, 965, 0x55c8c3515a00) = 0x55c8c5c658b0 10:38:01.774060 timeout_remove(0x55c8c5acf7a8, 0, 0xffffffff, 0) = 0 10:38:01.774239 io_add(213, 1, 0x55c8c351df75, 1070) = 0x55c8c5c658b0 10:38:01.774429 io_set_pending(0x55c8c5c658b0, 1, 0x55c8c54b8aa0, 0x7fb4ab11b06a) = 0x55c8c54b8aa0 10:38:01.774598 p_strarray_dup(0x7fb4ab703660, 0x55c8c5786e00, 5, 0x55c8c3723740) = 0x55c8c56d7530 10:38:01.774771 io_remove(0x55c8c5acf788, 0x55c8c3516d60, 5, 48) = 0 10:38:01.774957 timeout_add(0x7530, 0x55c8c351df75, 965, 0x55c8c3515a00) = 0x55c8c5c658b0 10:38:01.775129 timeout_remove(0x55c8c5acf7a8, 0, 0xffffffff, 0) = 0 10:38:01.775301 io_add(213, 1, 0x55c8c351df75, 1070) = 0x55c8c5c658b0 10:38:01.775492 io_set_pending(0x55c8c5c658b0, 1, 0x55c8c54b8aa0, 0x7fb4ab11b06a) = 0x55c8c54b8aa0 10:38:01.775662 p_strarray_dup(0x7fb4ab703660, 0x55c8c56d7530, 5, 0x55c8c3723740) = 0x55c8c5786e00 10:38:01.775833 io_remove(0x55c8c5acf788, 0x55c8c3516d60, 5, 48) = 0 That fd 213 is the next-highest FD opened, and is bound to /var/run/dovecot/director-admin. In fact the last few FDs are all connections to director-admin. Last few lines of "lsof" say: director 16109 dovecot 207u unix 0xffff940d13c93800 0t0 1106341075 type=STREAM director 16109 dovecot 208u unix 0xffff9409f5280000 0t0 1106337217 /var/run/dovecot/director-userdb type=STREAM director 16109 dovecot 209u unix 0xffff940d580fa000 0t0 1106337218 type=STREAM director 16109 dovecot 210u unix 0xffff940ac0f61000 0t0 1106343439 /var/run/dovecot/director-userdb type=STREAM director 16109 dovecot 211u unix 0xffff940bed105400 0t0 1106343440 type=STREAM director 16109 dovecot 212u unix 0xffff940b76d31c00 0t0 1106342408 /var/run/dovecot/director-admin type=STREAM director 16109 dovecot 213u unix 0xffff940ab241d000 0t0 1106340510 /var/run/dovecot/director-admin type=STREAM director 16109 dovecot 214u unix 0xffff9409c317ec00 0t0 1106340511 /var/run/dovecot/director-admin type=STREAM I attached with "gdb" and did "bt" at two different moments, resulting in: (gdb) bt #0 0x00007fb4ab11b06a in epoll_ctl () from target:/lib/x86_64-linux-gnu/libc.so.6 #1 0x00007fb4ab48964d in io_loop_handle_add () from target:/usr/lib/dovecot/libdovecot.so.0 #2 0x00007fb4ab4876b9 in ?? () from target:/usr/lib/dovecot/libdovecot.so.0 #3 0x00007fb4ab48776d in io_add () from target:/usr/lib/dovecot/libdovecot.so.0 #4 0x000055c8c35179d3 in doveadm_connections_ring_synced () #5 0x00007fb4ab488166 in io_loop_handle_timeouts () from target:/usr/lib/dovecot/libdovecot.so.0 #6 0x00007fb4ab489963 in io_loop_handler_run_internal () from target:/usr/lib/dovecot/libdovecot.so.0 #7 0x00007fb4ab48839c in io_loop_handler_run () from target:/usr/lib/dovecot/libdovecot.so.0 #8 0x00007fb4ab488548 in io_loop_run () from target:/usr/lib/dovecot/libdovecot.so.0 #9 0x00007fb4ab40c0a3 in master_service_run () from target:/usr/lib/dovecot/libdovecot.so.0 #10 0x000055c8c350b46d in main () (gdb) c Continuing. ^C Program received signal SIGINT, Interrupt. 0x00007fb4ab11b06a in epoll_ctl () from target:/lib/x86_64-linux-gnu/libc.so.6 (gdb) bt #0 0x00007fb4ab11b06a in epoll_ctl () from target:/lib/x86_64-linux-gnu/libc.so.6 #1 0x00007fb4ab489873 in io_loop_handle_remove () from target:/usr/lib/dovecot/libdovecot.so.0 #2 0x00007fb4ab487adf in ?? () from target:/usr/lib/dovecot/libdovecot.so.0 #3 0x000055c8c35158b7 in ?? () #4 0x000055c8c3516b4a in ?? () #5 0x000055c8c3516d78 in ?? () #6 0x000055c8c35179e4 in doveadm_connections_ring_synced () #7 0x00007fb4ab488166 in io_loop_handle_timeouts () from target:/usr/lib/dovecot/libdovecot.so.0 #8 0x00007fb4ab489963 in io_loop_handler_run_internal () from target:/usr/lib/dovecot/libdovecot.so.0 #9 0x00007fb4ab48839c in io_loop_handler_run () from target:/usr/lib/dovecot/libdovecot.so.0 #10 0x00007fb4ab488548 in io_loop_run () from target:/usr/lib/dovecot/libdovecot.so.0 #11 0x00007fb4ab40c0a3 in master_service_run () from target:/usr/lib/dovecot/libdovecot.so.0 #12 0x000055c8c350b46d in main () For completeness, doveconf -n output attached. -- Jan-Pieter Cornet <johnpc(a)xs4all.net> Systeembeheer XS4ALL Internet bv www.xs4all.nl

2 1

Alt Storage and dsync
by Matt Bryant 26 Apr '18

26 Apr '18

Stupid question but if dovecot mdbox is configured using alt path storage on a bi directional dysnc replicated pair of servers ... will the doveadm altmove need to be run on both halves of the cluster ??? rgds Matt -- <https://www.postbox-inc.com/?utm_source=email&utm_medium=siglink&utm_campai…>

1 0