January 2017 - dovecot

Dovecot auth-worker error after cram-md5 auth
by Poliman - Serwis 03 Feb '17

03 Feb '17

I set up cram-md5 using this tutorial https://wiki2.dovecot.org/HowTo/CRAM-MD5 in /etc/dovecot/dovecot.conf in passdb code block: listen = *,[::] protocols = imap pop3 #auth_mechanisms = plain login cram-md5 auth_mechanisms = cram-md5 plain login #dodana nizej linia ssl = required disable_plaintext_auth = yes log_timestamp = "%Y-%m-%d %H:%M:%S " mail_privileged_group = vmail postmaster_address = postmaster(a)vps342401.ovh.net ssl_cert = </etc/postfix/smtpd.cert ssl_key = </etc/postfix/smtpd.key ssl_protocols = !SSLv2 !SSLv3 !TLSv1 !TLSv1.1 ssl_cipher_list = ECDHE-RSA-AES128-GCM-SHA256:ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-ECDSA-AES256-GCM-SHA384[image: :D]HE-RSA-AES128-GCM-SHA256[image: :D]HE-DSS-AES$ ssl_prefer_server_ciphers = yes ssl_dh_parameters_length = 2048 mail_max_userip_connections = 100 passdb { # args = /etc/dovecot/dovecot-sql.conf # driver = sql driver = passwd-file args = scheme=cram-md5 /etc/dovecot/cram-md5.pwd } userdb { driver = prefetch } userdb { args = /etc/dovecot/dovecot-sql.conf driver = sql } Of course I created cram-md5.pwd file. All mails go out and come nicely. But after I want to do default settings by commented out these two lines: driver = passwd-file args = scheme=cram-md5 /etc/dovecot/cram-md5.pwd and uncomment # args = /etc/dovecot/dovecot-sql.conf # driver = sql I can't send emails - I use Thunderbird - get error "logging on server mail.example.com not work out". Error in logs: dovecot: auth-worker(22698): Error: Auth worker sees different passdbs/userdbs than auth server. dovecot: auth: Error: read(anvil-auth-penalty) failed: EOF Is it possible that hashed password from cram-md5.pwd file was written to database (if yes then where - I have ISPconfig)? I wasn't change any userdb {} block and this second userdb block has this same lines like default settings in passdb block. -- *Pozdrawiam / Best Regards* *Piotr Bracha* *tel. 534 555 877* *serwis(a)poliman.pl <serwis(a)poliman.pl>*

4 25

Moving to new password scheme
by ＠lbutlr 02 Feb '17

02 Feb '17

dovecot is setup on a system with MD5-CRYPT password scheme for all users, and I would like to update this to something that is secure, probably SSHA256-CRYPT, but I want to do this seamlessly without the users having to jump through any hoops. The users are in mySQL (managed via postfixadmin) and the mailbox record simply stores the hash in the password field. Users access their accounts though IMAP MUAs or Roundcube. How would I setup my system so that if a user logs in and still has a $1$ password (MD5-CRYPT) their password will be encoded to the new SHCEME and then the SQL row updated with the $5$ password instead? Something where they are redirected after authentication to a page that forces them to renter their password (or choose a new one) is acceptable. And, while I am here, is it worthwhile to set the -r flag to a large number (like something over 100,000 which sets takes about 0.25 seconds to do on my machine)? -- Apple broke AppleScripting signatures in Mail.app, so no random signatures.

4 8

quota-status returns quota_status_success when email would put user over quota
by Kristian Pedersen 02 Feb '17

02 Feb '17

Hi list, I am attempting to get quota-status service working, so I can deny email at the initial smtp dialog instead of generating bounces with lda. I can't seem to get quota-status to return quota_status_overquota, even when an email would put an account over quota. Quota in general works fine: Jan 22 06:39:23 mail dovecot: lda(xx(a)yy.dk): msgid=<25c5bdb20d58fc4f649f716a947613dc(a)zz.org>: save failed to INBOX: Quota exceeded (mailbox for user is full) Jan 22 06:39:23 mail dovecot: lda(xx(a)yy.dk): msgid=<25c5bdb20d58fc4f649f716a947613dc(a)zz.org>: rejected: Quota exceeded (mailbox for user is full) The system is a pretty standard debian 8 box running virtually on a kvm/ovirt cluster: root@mail:~# dpkg --list | grep dovec ii dovecot-core 1:2.2.13-12~deb8u1 amd64 secure POP3/IMAP server - core files ii dovecot-imapd 1:2.2.13-12~deb8u1 amd64 secure POP3/IMAP server - IMAP daemon ii dovecot-lmtpd 1:2.2.13-12~deb8u1 amd64 secure POP3/IMAP server - LMTP server ii dovecot-mysql 1:2.2.13-12~deb8u1 amd64 secure POP3/IMAP server - MySQL support ii dovecot-pop3d 1:2.2.13-12~deb8u1 amd64 secure POP3/IMAP server - POP3 daemon Mail storage is local XFS partitions as far as the dovecot server knows. Behind the scene its actually raw disk images on NFS shares from a SAN towards the kvm/ovirt cluster. I test by using nc towards the quota-status instance: root@mail:~# printf "recipient=kptest(a)asom-net.dk\nsize=1000000000\n\n" | nc -q1 localhost 12340 action=DUNNO User has 500MB quota: mysql> SELECT CONCAT('/data/vmail/',CONCAT(SUBSTRING_INDEX(email,'@',-1),'/',SUBSTRING_INDEX(email,'@',1))) AS home, 110 AS uid, 110 AS gid, CONCAT('*:storage=',mailquota,'M') AS quota_rule FROM virtual_users WHERE email='kptest(a)asom-net.dk'; +--------------------------------+-----+-----+----------------+ | home | uid | gid | quota_rule | +--------------------------------+-----+-----+----------------+ | /data/vmail/asom-net.dk/kptest | 110 | 110 | *:storage=500M | +--------------------------------+-----+-----+----------------+ 1 row in set (0.00 sec) maildirsize seems ok: root@mail:~# cat /data/vmail/asom-net.dk/kptest/maildirsize 524288000S 685 1 690 1 Quota-status will return unknown user if that is the case: root@mail:~# printf "recipient=kptest2(a)asom-net.dk\nsize=1000000000\n\n" | nc -q1 localhost 12340 action=551 5.5.1 User not found Log file (with mail_debug = yes): Jan 12 16:52:21 mail dovecot: quota-status: Debug: auth input: kptest(a)asom-net.dk home=/data/vmail/asom-net.dk/kptest uid=110 gid=110 quota_rule=*:storage=500M Jan 12 16:52:21 mail dovecot: quota-status: Debug: Added userdb setting: plugin/quota_rule=*:storage=500M Jan 12 16:52:21 mail dovecot: quota-status(kptest(a)asom-net.dk): Debug: Effective uid=110, gid=110, home=/data/vmail/asom-net.dk/kptest Jan 12 16:52:21 mail dovecot: quota-status(kptest(a)asom-net.dk): Debug: Namespace inbox: type=private, prefix=, sep=, inbox=yes, hidden=no, list=yes, subscriptions=yes location=maildir:/data/vmail/asom-net.dk/kptest/ Jan 12 16:52:21 mail dovecot: quota-status(kptest(a)asom-net.dk): Debug: maildir++: root=/data/vmail/asom-net.dk/kptest, index=, indexpvt=, control=, inbox=/data/vmail/asom-net.dk/kptest, alt= dovecot -n: # 2.2.13: /etc/dovecot/dovecot.conf # OS: Linux 3.16.0-4-amd64 x86_64 Debian 8.6 ext4 auth_default_realm = vejen-net.dk auth_mechanisms = plain login auth_verbose = yes disable_plaintext_auth = no first_valid_uid = 110 log_timestamp = "%Y-%m-%d %H:%M:%S " mail_debug = yes mail_location = maildir:/data/vmail/%d/%n/ mail_privileged_group = mail namespace inbox { inbox = yes location = mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { special_use = \Trash } prefix = } passdb { args = /etc/dovecot/local-sql.conf driver = sql } plugin { quota = maildir:User quota quota_rule = *:storage=200M quota_status_nouser = 551 5.5.1 User not found quota_status_overquota = 552 5.2.2 Mailbox is full quota_status_success = DUNNO } protocols = imap pop3 service auth { unix_listener /var/spool/postfix/private/auth { group = postfix mode = 0660 user = postfix } unix_listener auth-master { mode = 0600 user = vmail } user = root } service imap-login { client_limit = 1024 process_limit = 256 process_min_avail = 8 service_count = 0 vsz_limit = 512 M } service imap { process_limit = 10240 } service pop3-login { client_limit = 512 process_limit = 256 process_min_avail = 8 service_count = 0 vsz_limit = 512 M } service pop3 { process_limit = 10240 } service quota-status { client_limit = 1 executable = /usr/lib/dovecot/quota-status -p postfix inet_listener { port = 12340 } } ssl_cert = </etc/letsencrypt/live/mail.asom-net.dk/fullchain.pem ssl_key = </etc/letsencrypt/live/mail.asom-net.dk/privkey.pem ssl_prefer_server_ciphers = yes userdb { args = /etc/dovecot/local-sql.conf driver = sql } protocol lda { auth_socket_path = /var/run/dovecot/auth-master mail_plugins = quota postmaster_address = postmaster(a)asom-net.dk } protocol imap { mail_plugins = quota imap_quota } protocol pop3 { mail_plugins = quota pop3_uidl_format = %08Xu%08Xv } local-sql.conf: driver = mysql connect = host=xyz dbname=xyz user=xyz password=xyz default_pass_scheme = CRYPT password_query = SELECT email as user, password FROM virtual_users WHERE email='%u'; user_query = SELECT CONCAT('/data/vmail/',CONCAT(SUBSTRING_INDEX(email,'@',-1),'/',SUBSTRING_INDEX(email,'@',1))) AS home, 110 AS uid, 110 AS gid, CONCAT('*:storage=',mailquota,'M') AS quota_rule FROM virtual_users WHERE email='%u'; If I do a strace on the quota-status PID, it seems to do a stat on the directory and then give up? Not sure Im interpreting it correct (only including the last few lines): ... lseek(14, 833, SEEK_SET) = 833 munmap(0x7f165d32a000, 833) = 0 close(14) = 0 geteuid() = 0 getegid() = 110 getgid() = 110 getegid() = 110 setgroups(1, [110]) = 0 setresuid(-1, 110, -1) = 0 prctl(PR_SET_DUMPABLE, 1) = 0 stat("/data/vmail/asom-net.dk/kptest", {st_mode=S_IFDIR|0700, st_size=4096, ...}) = 0 prctl(PR_SET_DUMPABLE, 1) = 0 setsockopt(12, SOL_TCP, TCP_CORK, [1], 4) = 0 write(12, "action=DUNNO\n\n", 14) = 14 setsockopt(12, SOL_TCP, TCP_CORK, [0], 4) = 0 epoll_wait(11, {{EPOLLIN, {u32=1593554016, u64=139734059562080}}}, 5, 59999) = 1 read(12, "", 8146) = 0 epoll_ctl(11, EPOLL_CTL_DEL, 12, 7fff0be817a0) = 0 close(12) = 0 epoll_wait(11, {}, 5, 1000) = 0 write(5, "\35q\1\0007\10\0\0\1\0\0\0", 12) = 12 epoll_wait(11, ... It seems like sort of permission issue? root@mail:~# su - vmail No directory, logging in with HOME=/ $ id -a uid=110(vmail) gid=110(vmail) groups=110(vmail) $ cat /data/vmail/asom-net.dk/kptest/maildirsize 524288000S 685 1 690 1 /data/vmail/asom-net.dk is actually a symlink, maybe that could be of importance?: root@mail:~# ls -ld /data/vmail/asom-net.dk lrwxrwxrwx 1 root root 19 Jan 9 11:18 /data/vmail/asom-net.dk -> ../mnt1/asom-net.dk root@mail:~# ls -ld /data/mnt1/asom-net.dk/ drwxrwx--- 45 vmail vmail 4096 Dec 15 10:54 /data/mnt1/asom-net.dk/ root@mail:~# ls -ld /data/mnt1/asom-net.dk/kptest/ drwx------ 9 vmail vmail 4096 Jan 23 08:55 /data/mnt1/asom-net.dk/kptest/ root@mail:~# ls -ld /data/mnt1/asom-net.dk/kptest/maildirsize -rw------- 1 vmail vmail 23 Jan 12 16:50 /data/mnt1/asom-net.dk/kptest/maildirsize Anyone have any idea what might be wrong here? Regards, -- Kristian Pedersen ASOM-Net Systemadministrator www.asom-net.dk Telefon: 44 400 970

3 5

dovecot 2.2.27 welcome plugins script
by Carlos R Laguna 01 Feb '17

01 Feb '17

Hello everyone, I am trying to send a mail to newly autocreated users, using the welcome plugins, however, i am get unable to pipe user information correctly to the script, acording to the wiki the scripts are called just as the quota warnig scripts. Also, every time that a try something new i have created a new user...... is a pain. what can i do to test this without to create a new user after any little change? # 2.2.27 (c0f36b0): /etc/dovecot/dovecot.conf # Pigeonhole version 0.4.16 (fed8554) # OS: Linux 4.4.0-59-generic x86_64 Ubuntu 14.04.5 LTS nfs auth_mechanisms = plain login mail_debug = yes mail_gid = vmail mail_location = maildir:/var/nfsvmail/%d/%Ln/ mail_plugins = " quota welcome" mail_uid = vmail namespace inbox { inbox = yes location = mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { special_use = \Trash } prefix = } passdb { args = /etc/dovecot/dovecot-ldap.conf.ext driver = ldap } plugin { quota = maildir:User quota quota_grace = 10%% quota_rule = *:storage=2M quota_rule2 = Trash:storage=+1M quota_status_nouser = DUNNO quota_status_overquota = 552 5.2.2 El buzón esta lleno quota_status_success = DUNNO quota_warning = storage=95%% quota-warning 95 %u quota_warning2 = storage=80%% quota-warning 80 %u welcome_script = welcome %u welcome_wait = yes } postmaster_address = postmaster(a)cubava.cu protocols = " imap lmtp pop3" inet_listener pop3s { port = 0 } } service quota-status { client_limit = 1 executable = quota-status -p postfix inet_listener { port = 12340 } } service welcome { executable = script /usr/local/bin/welcome.sh unix_listener welcome { user = vmail } user = dovecot } ssl_cert = </etc/ssl/certs/mailcubava.pem ssl_key = # hidden, use -P to show it userdb { args = /etc/dovecot/dovecot-ldap.conf.ext driver = ldap } protocol imap { mail_plugins = " quota welcome imap_quota" } -rwxr-xr-x 1 dovecot vmail 397 Jan 31 16:11 welcome.sh #!/bin/sh USER=$1 cat << EOF | /usr/lib/dovecot/dovecot-lda -d $USER From: postmaster(a)example.com Subject: Welcome Message EOF -- Carlos R Laguna Mendoza Esp. Principal NDC TINORED Joven Club de Computación y Electrónica Tel: 52793189, 7863-3174, 7868-4230

1 0

Vacation Sending To Root
by Shaun Forsyth 31 Jan '17

31 Jan '17

Dovecot Version 2.0.9 I am having an issue where all vacation messages are being sent to root. Is any one able to advise. Dovecot Log Shows :- Sieve: msgid=< CAMD5TViw0sSigGm8JY5WOv0ROs1UdCJFnvb9vo-RzcNN74E92Q(a)mail.xxxxxx.com>: sent vacation response to <root> The emails are handed off to dovecot from sendmail using procmail :0 w | $DELIVER The variable $DELIVER is set as DELIVER="/usr/libexec/dovecot/deliver" I do have DROPPRIVS=YES in procmail, however I don't think deliver works without this. Any advice would greatly help. Shaun

2 2

mail_location in userdb inconsistent with imap and pigeonhole
by Ladislav Laska 31 Jan '17

31 Jan '17

Hi! I'm still fighting with lost mails and what I think is a locking problem. As I've lost some pretty important mails already, I decided to mitigate my troubles by moving to maildir format, which should not be susceptible to these problems. I read the docs and wanted to test on my account, before I mess with all the others and possibly break something. The way to do it seems to be to do an override via userdb. I already have a passwd-file authdb, so I uncommented those few lines in configuration, and added this to my passwd-file: krakonos:{SHA256}xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx=:1000:100::/home/krakonos:/bin/bash:userdb_mail=maildir:/home/krakonos/Mail Now, I'm not really sure why it has to be prefixed with userdb_, but the example in docs has it, and it seems to work. I've attached the relevant part of a logfile, with some debugging enabled, but in short: Dec 10 18:39:22 ibex dovecot: imap(krakonos): Debug: Added userdb setting: mail=maildir:/home/krakonos/Mail [...] Dec 10 18:39:22 ibex dovecot: imap(krakonos): Debug: maildir++: root=/home/krakonos/Mail, index=, indexpvt=, control=, inbox=/home/krakonos/Mail, alt= [...] Dec 10 18:39:52 ibex dovecot: lda(krakonos): Debug: mbox: INBOX defaulted to /home/krakonos/.mbox/inbox I'm confused why the last line chooses my old mbox over the new location. I'm even more confused by the fact the IMAP works fine, and as I connect from mutt over imap, I can see empty mailboxes (the ~/Mail has only the basic cur/new/tmp structure). However, pigeonhole delivers into the .mbox, and not only if it's for INBOX, but also for spam and others. I read somewhere in the docs that the order is first try userdb, then mail_location and namespaces override the rest. This might be the problem, as I have: mail_location = mbox:~/.mbox namespace inbox { inbox = yes location = mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { special_use = \Trash } prefix = } Due to inbox=yes this would explain why inbox is using the default path in mail_location. But why even the spam is delivered into the old mbox? And how do I solve the problem anyway, should I just remove the namespace? Anyway, it doesn't explain why the IMAP part and pigeonhole apply different rules and show me different inboxes, which is probably a bug. Thanks! ----------------------------------------------------------------------- Here is the rest of my configuration for reference, the log is attached: ----------------------------------------------------------------------- # 2.2.26.0 (23d1de6): /etc/dovecot/dovecot.conf # Pigeonhole version 0.4.16 (fed8554) # OS: Linux 4.4.26-gentoo x86_64 Gentoo Base System release 2.2 auth_username_format = %n hostname = ibex.krakonos.org login_greeting = Dovecot(a)krakonos.org ready. mail_debug = yes mail_location = mbox:~/.mbox mbox_read_locks = dotlock fcntl namespace inbox { inbox = yes location = mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { special_use = \Trash } prefix = } passdb { args = * driver = pam } passdb { args = scheme=CRYPT username_format=%u /etc/dovecot/users driver = passwd-file } plugin { sieve = file:~/sieve;active=~/.dovecot.sieve sieve_execute_bin_dir = /usr/lib/dovecot/sieve-execute sieve_execute_socket_dir = sieve-execute sieve_extensions = +vnd.dovecot.filter +editheader sieve_filter_bin_dir = /usr/lib/dovecot/sieve-filter sieve_filter_socket_dir = sieve-filter sieve_pipe_bin_dir = /usr/lib/dovecot/sieve-pipe sieve_pipe_socket_dir = sieve-pipe sieve_plugins = sieve_extprograms } postmaster_address = postmaster(a)krakonos.org protocols = imap service auth { unix_listener /var/spool/postfix/private/auth { mode = 0666 } } ssl_cert = </etc/ssl/dovecot/server.pem ssl_key = # hidden, use -P to show it userdb { driver = passwd } userdb { args = username_format=%u /etc/dovecot/users driver = passwd-file } protocol lda { mail_plugins = sieve } -- S pozdravem Ladislav "Krakonoš" Láska http://www.krakonos.org/

2 2

Sieve & Public folders issue (2nd episode)
by dovecot＠avv.solutions 31 Jan '17

31 Jan '17

Hello Community, Getting back to my previous post, here a a complement: - Having enabled sieve, I can store mails and create folder under the recipient - Having enabled public folders, I can move mail into them using a client (eg Thunderbird); security is therefore ok (actually I gave full control to the user) eg Public/Newsletters/SomeCompany - *BUT* running a sieve script to store to Public/Newsletters/SomeCompany returns the following in the logs: "failed to store into mailbox 'Public/Newsletters/SomeCompany': Permission denied." Any tip? Thank you for your help. Greetz

2 1

panic when doveadm sieve put between multiple hosts
by Tyler Jachetta 31 Jan '17

31 Jan '17

Dovecot version: 2.2.25 (7be1766) I’m looking into an issue we’re having with a new setup. We have one fronted host with two backend hosts, and we’re attempting to use doveadm to allow us to administer sieve on the frontend and have it replicated to the backend hosts. When I attempt to use doveadm sieve put on the frontend host, i get the following on the frontend host: frontend-machine $ cat /tmp/default.sieve | doveadm sieve put default doveadm(root): Error: doveadm server disconnected before handshake: Connection reset by peer doveadm(root): Error: 10.1.11.82:10993: Command sieve put failed for root: Connection reset by peer when this happens, on the backend, I see this in /var/log/maillog: Jan 9 17:00:35 was-imap1 dovecot: doveadm: Panic: epoll_ctl(add, 0) failed: Operation not permitted (fd doesn't support epoll - instead of '<file', try 'cat file|') Jan 9 17:00:35 was-imap1 dovecot: doveadm: Error: Raw backtrace: /usr/lib64/dovecot/libdovecot.so.0(+0x8d52e) [0x7f7a86ba952e] -> /usr/lib64/dovecot/libdovecot.so.0(+0x8d60e) [0x7f7a86ba960e] -> /usr/lib64/dovecot/libdovecot.so.0(i_fatal+0) [0x7f7a86b481b6] -> /usr/lib64/dovecot/libdovecot.so.0(io_loop_handle_add+0x147) [0x7f7a86bbe067] -> /usr/lib64/dovecot/libdovecot.so.0(+0xa0311) [0x7f7a86bbc311] -> /usr/lib64/dovecot/libdovecot.so.0(io_add+0xd) [0x7f7a86bbc3bd] -> dovecot/doveadm-server(doveadm_mail_get_input+0xde) [0x7f7a875c53de] -> dovecot/doveadm-server(doveadm_mail_single_user+0x73) [0x7f7a875c5633] -> dovecot/doveadm-server(+0x2874d) [0x7f7a875c574d] -> dovecot/doveadm-server(doveadm_cmd_ver2_to_mail_cmd_wrapper+0x27d) [0x7f7a875c65fd] -> dovecot/doveadm-server(doveadm_cmd_run_ver2+0x50c) [0x7f7a875d492c] -> dovecot/doveadm-server(+0x3b223) [0x7f7a875d8223] -> /usr/lib64/dovecot/libdovecot.so.0(io_loop_call_io+0x4c) [0x7f7a86bbce8c] -> /usr/lib64/dovecot/libdovecot.so.0(io_loop_handler_run_internal+0xff) [0x7f7a86bbe2ef] -> /usr/lib64/dovecot/libdovecot.so.0(io_loop_handler_run+0x25) [0x7f7a86bbcf15] -> /usr/lib64/dovecot/libdovecot.so.0(io_loop_run+0x38) [0x7f7a86bbd0c8] -> /usr/lib64/dovecot/libdovecot.so.0(master_service_run+0x13) [0x7f7a86b4e613] -> dovecot/doveadm-server(main+0x186) [0x7f7a875b6276] -> /lib64/libc.so.6(__libc_start_main+0xf5) [0x7f7a8677bb15] -> dovecot/doveadm-server(+0x19321) [0x7f7a875b6321] Jan 9 17:00:36 was-imap1 dovecot: doveadm: Fatal: master: service(doveadm): child 28542 killed with signal 6 (core dumped) Now, if I log in to either of the backend hosts, I can run the sieve put command and it all works out fine, so it’s something in the communication between the frontend and backend. dovecot -n output for both a frontend and backend below frontend-machine $ dovecot -n # 2.2.25 (7be1766): /etc/dovecot/dovecot.conf # Pigeonhole version 0.4.15 (97b3da0) # OS: Linux 3.10.0-327.18.2.el7.x86_64 x86_64 CentOS Linux release 7.2.1511 (Core) auth_username_format = %Ln director_mail_servers = 10.1.11.82 10.1.11.83 director_servers = 10.1.11.81 director_username_hash = %Ln doveadm_port = 10993 managesieve_notify_capability = mailto managesieve_sieve_capability = fileinto reject envelope encoded-character vacation subaddress comparator-i;ascii-numeric relational regex imap4flags copy include variables body enotify environment mailbox date index ihave duplicate mime foreverypart extracttext mbox_write_locks = fcntl namespace inbox { inbox = yes location = mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { special_use = \Trash } prefix = } passdb { args = proxy=y nopassword=y ssl=any-cert driver = static } plugin { sieve = file:~/sieve;active=~/.dovecot.sieve } protocols = imap pop3 lmtp sieve service director { fifo_listener login/proxy-notify { mode = 0666 } inet_listener { port = 1109 } unix_listener director-userdb { mode = 0600 } unix_listener login/director { mode = 0666 } } service doveadm { inet_listener { port = 10993 } } service imap-login { executable = imap-login director } service imap { process_limit = 250 } service ipc { unix_listener ipc { user = dovecot } } service managesieve-login { executable = managesieve-login director } service pop3-login { executable = pop3-login director } ssl = required ssl_ca = </etc/pki/CA/ca-cert-***.pem ssl_cert = </etc/pki/tls/certs/dovecot.cert ssl_client_ca_file = </etc/pki/CA/ca-cert-***.pem ssl_key = </etc/pki/tls/private/dovecot.key ssl_protocols = !SSLv2 !SSLv3 ssl_require_crl = no userdb { driver = passwd } protocol doveadm { auth_socket_path = director-userdb } protocol sieve { passdb { args = proxy=y nopassword=y starttls=any-cert driver = static name = } } local 10.1.11.0/24 { doveadm_password = # hidden, use -P to show it } backend-machine $ dovecot -n # 2.2.25 (7be1766): /etc/dovecot/dovecot.conf # Pigeonhole version 0.4.15 (97b3da0) # OS: Linux 3.10.0-327.18.2.el7.x86_64 x86_64 CentOS Linux release 7.2.1511 (Core) auth_username_format = %Ln default_vsz_limit = 2 G lda_mailbox_autocreate = yes lda_mailbox_autosubscribe = yes mail_location = maildir:~/Maildir mail_plugins = " notify replication" managesieve_notify_capability = mailto managesieve_sieve_capability = fileinto reject envelope encoded-character vacation subaddress comparator-i;ascii-numeric relational regex imap4flags copy include variables body enotify environment mailbox date index ihave duplicate mime foreverypart extracttext mbox_write_locks = fcntl namespace inbox { inbox = yes location = mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { special_use = \Trash } prefix = separator = / } passdb { driver = pam } plugin { mail_replica = tcp:10.1.11.82:10993 sieve = file:~/sieve;active=~/.dovecot.sieve } protocols = imap pop3 lmtp sieve service aggregator { fifo_listener replication-notify-fifo { mode = 0666 } unix_listener replication-notify { mode = 0666 } } service doveadm { inet_listener { port = 10993 } } service lmtp { inet_listener lmtp { port = 24 } } service replicator { process_min_avail = 1 unix_listener replicator-doveadm { mode = 0666 } } ssl = required ssl_ca = </etc/pki/CA/ca-cert-***.pem ssl_cert = </etc/pki/tls/certs/dovecot.cert ssl_client_ca_file = </etc/pki/CA/ca-cert-***.pem ssl_key = </etc/pki/tls/private/dovecot.key ssl_protocols = !SSLv2 !SSLv3 ssl_require_crl = no userdb { driver = passwd } protocol lmtp { auth_username_format = %Ln info_log_path = /var/log/dovecot/lmtp.log mail_plugins = " notify replication sieve" postmaster_address = postmaster(a)***global.com } protocol sieve { passdb { args = proxy=y nopassword=y starttls=any-cert driver = static name = } } local 10.1.11.0/24 { doveadm_password = # hidden, use -P to show it }

2 1

dovecot 2.2.18 replication I/O Timeout
by Jan Büren 31 Jan '17

31 Jan '17

Dear list, I am having an issue like this http://www.dovecot.org/list/dovecot-cvs/2014-September/024898.html with my current dovecot 2.2.18 Concret: doveadm -Dv replicator replicate -p high -f info throws an I/O Timeout error and the only thing I can see are some lock issues: Error: Couldn't lock /var/vmail/info/.dovecot-sync.lock: Timed out after 30 seconds Can I debug the communication one level deeper? I replicate over a somewhat slow DSL connection. I had the some problem with dovecot 2.2.9 and after upgrading to 2.2.15 everything went smoothly, just a wild guess, maybe the timeout_reset was removed again? Well, doesn't look like it: https://github.com/dovecot/core/blob/master/src/doveadm/dsync/dsync-ibc-str… or https://github.com/dovecot/core/blame/master/src/doveadm/dsync/dsync-ibc-st… Some debug hints would be nice. Thanks! Jan -- Blog http://blog.kivitendo.de/ kivitendo GmbH Jan Büren Kölnstr. 311 53117 Bonn USt-IdNr. DE292363254 Telefon: 0228 92 98 2012 persönliche Durchwahl: 0228 92 97 8965

1 0

Auth cache does not take %real_rip into account
by Tom Sommer 31 Jan '17

31 Jan '17

I run a Director setup with a webmail in front, the webmail is in login_trusted_networks and sends IMAP-ID x-original-ip to log the client IP. If I enable auth_debug on the director, I see that the cache key contains the client IP, and not the %real_rip. This is causing problems because in my passdb SQL query, I use the %real_rip to determine if login is allowed. Should %real_rip not be added to the auth cache key? Or should it be the cache key instead of the %rip? Thanks -- Tom

1 0