- dovecot - dovecot.org

Dovecot UIDs and POP.
by Plutocrat 26 Sep '19

26 Sep '19

Hi, I recently performed a mail server migration for a client under fairly serious time constraints (ahem, yes one of those jobs). I would normally use imapsync to get all the mail copied to the new server, but under the circumstances, I had to write a script to rsync each of the mailboxes over ssh, from one server to another. This all worked fine, except for the fact that if a client was using a POP email client, when they connected to the new server, they re-downloaded all their mail, creating duplicates of everything in their Outlook client. Of course they weren't happy about this. IMAP email clients were OK. No duplicates, and everything was fine. So while I was migrating the mail, I did try for a while to understand the format of the UID files, but failed to do so in the available time, so the client just had to deal with duplicate emails. But now the smoke has cleared, I'd like to understand the problem a little better, and I was hoping someone on this forum could explain it to me, and the changes I'd need to make to the files so that the POP client DIDN'T download the duplicate emails. Here's what I understand. The list of dovecot files in an account looks like this: -rw-r----- 1 admin mail 0 Jul 11 2018 dovecot-acl-list -rw-r----- 1 admin mail 14K Sep 20 08:36 dovecot.index -rw-r----- 1 admin mail 216K Sep 20 17:24 dovecot.index.cache -rw-r----- 1 admin mail 29K Sep 20 17:24 dovecot.index.log -rw-r----- 1 admin mail 388 Dec 17 2017 dovecot.list.index -rw-r----- 1 admin mail 1.6K Jul 11 2018 dovecot.list.index.log -rw-r----- 1 admin mail 24 Jul 11 2018 dovecot.mailbox.log -rw-r----- 1 admin mail 53 Sep 12 09:42 dovecot-quota -rw-r----- 1 admin mail 52K Sep 20 17:24 dovecot-uidlist -rw-r----- 1 admin mail 8 Jul 11 2018 dovecot-uidvalidity -rw-r----- 1 admin mail 0 Jul 4 2016 dovecot-uidvalidity.5779bbeb It seems that the dovecot-uidlist is an index of all the mail files in the account. As far as I can gather, this should be the same on both servers after I've rsynced them over ssh: The files and the list of them in dovecot-uidlist should correspond. So what is it that the POP client looks at and decides to re-download all the emails? And how could I have stopped it from doing that, and get it to realize that the list of emails it holds internally is the same as the list on the server? What would I need to change in the dovecot files? Thanks for any insight into the process. As a secondary question -- and perhaps I should put this in a separate message -- I did notice that "doveadm sync" would apparently have helped me with this, but I wasn't able to get that to work either. I believe it was something to do with the fact that all the mailboxes were under the same linux user account on the target server, and I couldn't figure out all the paths and permissions in time. Would 'doveadm sync' have fixed all the UID and duplicate POP email issues? P.

6 11

NFS Locking and Submission Service Authentication
by Asai 25 Sep '19

25 Sep '19

Greetings, We're in the process of upgrading our Dovecot server to new hardware and new expanded storage. We planned on using an NFS share for the mail storage, as we're running Postfix / Dovecot on a VM and wanted to separate out the mail storage from the VM for backup reasons. I read as much as I could find on line regarding configuring Dovecot to use NFS, and set it up as best I could, but I'm still running into lock errors e.g.: Sep 25 10:30:35 triata4 dovecot: imap(user(a)triata.globalchange.media)<75580><enQ/AWSTkQkKCgDu>: Error: fcntl(/vmail/triata.globalchange.media/user/dovecot.index.log, write-lock, F_SETLKW) locking failed: No locks available Sep 25 10:30:35 triata4 dovecot: imap(user(a)triata.globalchange.media)<75580><enQ/AWSTkQkKCgDu>: Error: mail_index_wait_lock_fd() failed with file /vmail/triata.globalchange.media/user/dovecot.index.log: No locks available Also, I have been trying to configure the Dovecot submission service to authenticate with Postfix, but I keep running into errors which are related to how I'm configuring the authentication. How do we set the user@domain and password string for submission in the configs? Thank you. Please see doveconf -n below, if you see any other errors in this config, I'd be grateful if you pointed them out: # 2.3.7.2 (3c910f64b): /etc/dovecot/dovecot.conf # Pigeonhole version 0.5.7.2 (7372921a) # OS: Linux 3.10.0-957.27.2.el7.x86_64 x86_64 CentOS Linux release 7.6.1810 (Core) nfs auth_debug = yes auth_debug_passwords = yes auth_mechanisms = plain login auth_verbose = yes auth_verbose_passwords = plain debug_log_path = /var/log/dovecot-debug.log deliver_log_format = msgid=%m: from=%f: %$ hostname = triata.globalchange.media mail_debug = yes mail_fsync = always mail_home = /vmail/%d/%n/home mail_location = maildir:/vmail/%d/%n 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 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 = / type = private } passdb { args = /etc/dovecot/dovecot-mysql.conf driver = sql } plugin { mail_log_fields = uid box msgid from flags sieve = file:/vmail/%d/%n/sieve;active=/vmail/%d/%n/.dovecot.sieve } protocols = imap lmtp submission sieve service auth { unix_listener /var/spool/postfix/private/auth { mode = 0666 } unix_listener auth-userdb { group = vmail mode = 0666 user = vmail } } service imap-login { inet_listener imap { port = 143 } inet_listener imaps { port = 993 ssl = yes } } service lmtp { unix_listener lmtp { mode = 0666 } } service managesieve-login { inet_listener sieve { port = 4190 } process_min_avail = 0 service_count = 1 vsz_limit = 64 M } service managesieve { process_limit = 1024 } service submission-login { inet_listener submission { port = 587 } } ssl_cert = </etc/letsencrypt/live/triata.globalchange.media/fullchain.pem ssl_key = # hidden, use -P to show it submission_relay_host = triata.globalchange.media submission_relay_password = # hidden, use -P to show it submission_relay_port = 587 submission_relay_ssl = starttls submission_relay_trusted = yes submission_relay_user = %u userdb { args = /etc/dovecot/dovecot-mysql.conf driver = sql } protocol lda { mail_plugins = }

3 4

Spam Blocking by filtering on username / id
by Plutocrat 25 Sep '19

25 Sep '19

This is probably quite an easy question, but I haven't been able to find the answer. I'm running a server where all the email addresses are in the format "user(a)domain.com". I've noticed that a large number of fake login attempts use the format "user" eg. reception, service, root, admin. Is it possible to prevent any such logins to these email users without an @domain.com? Or maybe ignore them. Or drop them from the logging. P.

5 5

'director_tag' field returned from passdb lookup results in 'unknown passdb extra field'
by Dan 25 Sep '19

25 Sep '19

Hello all, I'm working with director in tandem with 'director_tags' to group some backend hosts. I've assigned the groups to the hosts by appending a '@' character after the IP address of a backend host under director_mail_servers: ... director_mail_servers = 192.168.0.1@foo 192.168.0.2@bar 192.168.0.3@foobar ... I can confirm that the backend hosts possess the correct tags by issuing: # doveadm director status mail server ip tag vhosts state state changed users 192.168.0.1 foo 100 up - 0 192.168.0.2 bar 100 up - 0 192.168.0.3 foobar 100 up - 0 And examining that the tags have been assigned correctly. My passdb lookup is using mysql, and it appears as such: passdb { args = /etc/dovecot/dovecot-sql.conf driver = sql } This file looks like: driver = mysql connect = host=localhost dbname=xxx user=xxx password=xxx password_query = select director_tag, username, domain, password, 'Y' as `proxy` from users where username = '%n' and domain = '%d' iterate_query = select director_tag, username, domain from users I have a database that contains the following columns: username, domain, password, home, UID, GID, active, and director_tag To assign users to a backend, I went off of the documentation that states: > " A passdb lookup can return "director_tag" field containing the wanted tag name. " To fulfill this, I return the 'director_tag' in the passdb lookup, through password_query. The mapping seems to work. When I change the row in the database, the user gets proxied to the new backend on the next connection. However, My question is, Why does this appear in the logs? "dovecot[1963]: imap-login: Debug: Ignoring unknown passdb extra field: director_tag" How could this be an unknown passdb extra field, when the documentation states that you may return a field containing this key? In the file core/src/login-common/client-common-auth.c - Starting at line 142, there is no if condition checking for 'director_tag'. This causes that line to be output to debug handle. Is this intentional? Thank you for your time. Full configuration: auth_debug = yes auth_debug_passwords = yes auth_mechanisms = plain login auth_verbose = yes auth_verbose_passwords = yes director_mail_servers = 192.168.0.1@foo 192.168.0.2@bar 192.168.0.3@foobar director_servers = x.x.x.x y.y.y.y director_user_expire = 5 mins disable_plaintext_auth = no doveadm_port = 24245 haproxy_trusted_networks = x.x.x.x y.y.y.y 127.0.0.1 login_greeting = Dovecot At Your Service 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 passdb { args = /etc/dovecot/dovecot-sql.conf driver = sql } protocols = imap pop3 service director { fifo_listener login/proxy-notify { mode = 0666 user = $default_login_user } inet_listener { port = 9090 } unix_listener director-userdb { mode = 0600 } unix_listener login/director { mode = 0666 } } service imap-login { executable = imap-login director inet_listener imap { haproxy = yes port = 143 } inet_listener imaps { haproxy = yes port = 993 ssl = yes } } service ipc { unix_listener ipc { user = dovecot } } service pop3-login { executable = pop3-login director inet_listener pop3 { haproxy = yes port = 110 } inet_listener pop3s { haproxy = yes port = 995 ssl = yes } } ssl_cert = </etc/dovecot/private/dovecot.pem ssl_dh = # hidden, use -P to show it ssl_key = # hidden, use -P to show it userdb { args = /etc/dovecot/dovecot-sql.conf driver = sql } protocol doveadm { auth_socket_path = director-userdb }

2 1

Can't Get SASL to Work
by Christopher Wooley 25 Sep '19

25 Sep '19

I have been fighting with this for almost 2 days. I can't seem to get the SMTP server to authenticate through SASL. I have turned so many things off and on, I don't remember what I've done, anymore. Can someone look at these and tell me what I'm missing. I get this in the log, anytime postfix tries to authenticate. Sep 24 19:48:14 ubuntu postfix/smtpd[14464]: warning: SASL: Connect to smtpd failed: No such file or directory Sep 24 19:48:14 ubuntu postfix/smtpd[14464]: fatal: no SASL authentication mechanisms Sep 24 19:48:15 ubuntu postfix/master[14048]: warning: process /usr/lib/postfix/sbin/smtpd pid 14464 exit status 1 srw-rw-rw- 1 postfix postfix 0 Sep 24 19:10 auth 10-auth.conf disable_plaintext_auth = no auth_ssl_require_client_cert = no auth_mechanisms = plain !include auth-system.conf.ext -------------- 10-master.conf service imap-login { inet_listener imap { } inet_listener imaps { } } service pop3-login { inet_listener pop3 { } inet_listener pop3s { } } service lmtp { unix_listener /var/spool/postfix/private/dovecot-lmtp { mode = 0666 user = postfix group = postfix } } service imap { } service pop3 { } service auth { unix_listener /var/spool/postfix/private/auth { mode = 0666 user = postfix group = postfix } unix_listener auth-userdb { mode = 0666 user = postfix group = postfix } } service auth-worker { } service dict { unix_listener dict { } } -------------- main.cf smtpd_banner = $myhostname ESMTP $mail_name (Ubuntu) biff = no append_dot_mydomain = no readme_directory = no compatibility_level = 2 smtpd_tls_cert_file = /etc/ssl/certs/ssl-cert-snakeoil.pem smtpd_tls_key_file = /etc/ssl/private/ssl-cert-snakeoil.key smtpd_use_tls=yes smtpd_tls_session_cache_database = btree:${data_directory}/smtpd_scache smtp_tls_session_cache_database = btree:${data_directory}/smtp_scache smtpd_relay_restrictions = permit_mynetworks permit_sasl_authenticated defer_unauth_destination myhostname = ubuntu.local.acr-live.com alias_maps = hash:/etc/aliases alias_database = hash:/etc/aliases mydestination = $myhostname, ubuntu, localhost.localdomain, xyz.com , localhost mynetworks = 127.0.0.0/8 [::ffff:127.0.0.0]/104 [::1]/128 mailbox_size_limit = 0 recipient_delimiter = + inet_interfaces = all inet_protocols = all smtpd_sasl_type = dovecot smtpd_sasl_auth_enable = yes smtp_sasl_auth_enable = yes broken_sasl_auth_clients = yes smtpd_tls_security_level = may ---------------------------------- master.cf smtp inet n - y - - smtpd submission inet n - y - - smtpd -o syslog_name=postfix/submission -o smtpd_tls_security_level=encrypt -o smtpd_sasl_auth_enable=yes -o smtpd_tls_auth_only=yes -o smtpd_reject_unlisted_recipient=no -o smtpd_sasl_type=dovecot -o smtpd_sasl_path=private/auth -o smtpd_recipient_restrictions= -o smtpd_relay_restrictions=permit_sasl_authenticated,reject -o milter_macro_daemon_name=ORIGINATING pickup unix n - y 60 1 pickup cleanup unix n - y - 0 cleanup qmgr unix n - n 300 1 qmgr tlsmgr unix - - y 1000? 1 tlsmgr rewrite unix - - y - - trivial-rewrite bounce unix - - y - 0 bounce defer unix - - y - 0 bounce trace unix - - y - 0 bounce verify unix - - y - 1 verify flush unix n - y 1000? 0 flush proxymap unix - - n - - proxymap proxywrite unix - - n - 1 proxymap smtp unix - - y - - smtp relay unix - - y - - smtp -o syslog_name=postfix/$service_name showq unix n - y - - showq error unix - - y - - error retry unix - - y - - error discard unix - - y - - discard local unix - n n - - local virtual unix - n n - - virtual lmtp unix - - y - - lmtp anvil unix - - y - 1 anvil scache unix - - y - 1 scache maildrop unix - n n - - pipe flags=DRhu user=vmail argv=/usr/bin/maildrop -d ${recipient} uucp unix - n n - - pipe flags=Fqhu user=uucp argv=uux -r -n -z -a$sender - $nexthop!rmail ($recipient) ifmail unix - n n - - pipe flags=F user=ftn argv=/usr/lib/ifmail/ifmail -r $nexthop ($recipient) bsmtp unix - n n - - pipe flags=Fq. user=bsmtp argv=/usr/lib/bsmtp/bsmtp -t$nexthop -f$sender $recipient scalemail-backend unix - n n - 2 pipe flags=R user=scalemail argv=/usr/lib/scalemail/bin/scalemail-store ${nexthop} ${user} ${extension} mailman unix - n n - - pipe flags=FR user=list argv=/usr/lib/mailman/bin/postfix-to-mailman.py ${nexthop} ${user}

2 2

Mail messages are occasionally getting "stuck" behind one message on the server with POP3
by David Salisbury 24 Sep '19

24 Sep '19

I'm currently using Dovecot 2.2.33.2, although I've had this particular issue happen periodically over the last few years with other Dovecot versions. It seems to only affect accounts I'm receiving with POP3, not IMAP ones. What happens is the "top-most" (or, oldest) message in the "cur" folder seems to get "stuck". So I'll, say, come back in after a break to check my email (Thunderbird in my case, latest version), and in one of my folders I'll have 30 copies of the same email (all with the same timestamp), and no newer messages delivered since the timestamp of that message. If I go look in the "cur" folder, it has lots of messages in it. The "new" folder is empty. Every time I click "Get Messages" in the mail client, I get another new copy of the same email and no other mail. If I go manually into the cur folder and move the top-most (again, meaning "oldest") message out (which of course corresponds to the message that I'm receiving copies of over-and-over repeatedly) then when I click "Get Messages" all of the other messages come pouring into my mail client just fine and everything becomes "un-stuck". I've done some experiments, and if I copy that same "bad" message back in to cur with its original filename, it gets stuck again. It is an encrypted message, so I tried decrypting it and putting it back in cur with the original name, and it gets stuck in the same fashion even when it is decrypted. I tried putting the bad message into an account that receives with IMAP, just to see if it would maybe keep messages from delivering or something, but new messages come into the IMAP account just fine, even with the bad message present, so the issue seems specific to POP3. Any ideas what might cause something like that? I've had it happen twice over the last two days (which is unusually frequent, hence my message here), and the message sources are varied. Today it was a normal email message from a person, and yesterday it was a Twitch notification. I don't see anything unusual in the logs (and I did turn on debugging to see if it gave some additional info), just normal POP3 logins and connections closings when I'm clicking "Get Messages". -David

1 0

Any need to be worried about occasional dsync errors ?
by Laura Smith 24 Sep '19

24 Sep '19

I am occasionally (maybe every 4 hours or less frequently) seeing the following two errors appear in my logs. Are they any cause for concern ? Error: Timeout during state=sync_mails (send=done recv=mails) I/O has stalled, no activity for 600 seconds (last sent=mail_request (EOL)

1 0

Quota plug-in, do the directors also need it enabled?
by Kristijan Savic - ratiokontakt GmbH 23 Sep '19

23 Sep '19

I am having an issue with the quota display not showing up in e.g. Roundcube. There are multiple Dovecot servers and three directors. A load balancer also does it's thing. The quota plug-in is enabled everywhere, except on the directors. Do those also need it enabled? -- Kristijan Savic

1 0

pigeonhole sieve "standalone"
by Greg Minshall 22 Sep '19

22 Sep '19

hi. i've been thinking (but, for years now!) of a replacement for procmail, and pigeonhold sieve seems like a good candidate. but, i run nmh, not dovecot. as i started install it (on arch linux), i started to worry that it may be foolish of me to try to run it outside of a dovecot system. any wisdom about that would be gratefully received. thanks in advance, and thanks for the software. Greg Minshall

1 0

mdbox + sis attachments. Error stream is larger than expected by 1 byte
by Oleg Krivonosov 20 Sep '19

20 Sep '19

Hello! I use Dovecot 2.3.4.1-5+deb10u1 (Debian 10 x64) with mdbox, sis and imap. I catch many errors with attachments "Stream is larger than expected" by 1 byte in emails from Xerox WorkCentre 3345. Examples: Sep 20 23:20:04 mailserver dovecot: imap(wasya)<606><AG1zywGTVodVrE84>: Error: Mailbox INBOX: UID=5002: read(attachments-connector(/var/mail/virtual/wasya/mdbox/storage/m.210)) failed: read(/var/mail/attachments/b3/56/b3563977baf186928d38f81480b6443cf6d6d7e4-b86d80062ab36f5d7f190000f7ccf497[base64:18 b/l]) failed: Stream is larger than expected (882801 > 882800, eof=0) (FETCH BODY[]) Sep 20 23:32:27 mailserver dovecot: imap(wasya)<4831><n0a39wGTZ8hVyu8l>: Error: Mailbox INBOX: UID=654996: read(attachments-connector(/var/mail/virtual/wasya/mdbox/storage/m.4372)) failed: read(/var/mail/virtual/wasya/mdbox/storage/m.4372 trailer) failed: Stream is larger than expected (38 > 37, eof=0) (FETCH BODY[]). This error breaks getting email in MS Outlook and breaks viewing attachment in the problem email. Please help me to debug and to fix this issue. -- Best regards Oleg

1 0