- dovecot - dovecot.org

Using both starttls and ssl in passdb on proxy results in timeouts
by Filias Heidt 18 Sep '18

18 Sep '18

Hi List, I have a dovecot which proxies to different backends depending on an entry in a mysql-database. The mysql-query sets ‚ssl‘ to ‚any-cert‘ and this works fine. But this causes me a problem: sieve-backends only support STARTTLS and if I set ‚ssl‘ to ‚any-cert‘ (or yes), it will attempt a TLS-connection to the sieve-backends, which fails. My attempt was to alter the query to include %{real_lport} and return ‚ssl=no‘ and ‚starttls=any-cert‘ if the port matches the sieve-port. It works as expected in that it returns the correct values and proxies to the correct backend. However it seems that TLS is no longer working and I get timeouts from the backends. Debug: client passdb out: OK 1 user=someuser(a)example.com proxy proxy_nopipelining=y host=backend1.example.com nodelay=y nologin starttls=no ssl=any-cert hostip=so.me.i.p pass=<hidden> results in: Sep 17 11:08:47 imapproxy1 dovecot: imap-login: Error: proxy(someuser(a)example.com): Login for so.me.i.p:993 timed out in state=/none (after 30 secs, local=lo.cal.i.p:60524): user=<someuser(a)example.com>, method=PLAIN, rip=re.mo.te.ip, lip=lo.cal.i.p, TLS, session=<OySXgw12auwgARYIAAYABwAAAAAAAwAU> My query looks like this: password_query = SELECT host from proxy_domain, NULL as password, 'y' as nopassword, 'y' as proxy, NULL as destuser, 'y' as proxy_nopipelining, 'y' as nodelay, 'y' as nologin, IF(%{real_lport}=4190, 'any-cert', 'no') as 'starttls', IF(%{real_lport}<>4190, 'any-cert', 'no') as 'ssl‘; As soon as I remove the starttls-part and the passdb only returns ssl=any-cert (without starttls=no) it works flawlessly. Is it possible that I am attacking the problem the wrong way? Or is it not possible to set both starttls and ssl to some values in passdb and enable/disable them as needed? Thanks for any input :) Cheers, Filias

2 2

How to send mail to mailbox with disabled domain?
by Kai Schaetzl 17 Sep '18

17 Sep '18

Given the following: mailboxes: user1(a)example1.com user2(a)example1.com user3(a)example1.com etc. aliases: whatever(a)example1.com -> user1(a)example1.com whatever(a)example2.com -> user1(a)example1.com whatever(a)example3.com -> user1(a)example1.com Now the problem: example1.com MX goes elsewhere (doesn't point to this server anymore). Domains example2.com and example3.com still point to that server and should be able to accept mail. I have to disable mail acceptance for example1.com. If not, mail sent *from* that server (e.g. from a web form) to that domain will not leave the server. However, if I disable example1.com for mail dovecot lmtp will not deliver mail to this mail box anymore, although the mailbox still exists. How can I solve this? Is there a way of solving this, but keeping the domain example1.com in the name for these mailboxes? Or is there a way to tell dovecot to ignore domains for mailbox names? e.g. deliver to "user1"? (All user localparts are unique.) Thanks for any hints. Kai

4 6

Re: Proxy secured incoming POP3/IMAP4 to unsecure backend?
by Alexander Chekalin 17 Sep '18

17 Sep '18

Thank you! Ok, so I can omit ssl=no and startssl=no, and this results in default settings for ssl which is 'off'? Or the defaults are 'on' anyway? Can I somehow specify ports on remote hosts that proxy will use to connect to? Like (just image): 'proxy host_imap=10.1.1.1:143 host_pop=10.1.1.1:110' or somehow? On Mon, Sep 17, 2018 at 4:33 PM Aki Tuomi <aki.tuomi(a)dovecot.fi> wrote: > Due to certain design issues, the ssl=no is actually same as ssl=yes, same > goes for starttls=no. So there is no support actually for "ssl=no" at this > moment. > > Aki > > > On 17 September 2018 at 15:32 Alexander Chekalin < > alexander.chekalin(a)gmail.com> wrote: > > > > > > Surely. > > > > Here it is: > > > > # doveadm auth user(a)domain.com > > Password: > > passdb: chekalin_krg(a)ascon.ru auth succeeded > > extra fields: > > user=user(a)domain.com > > proxy > > host=10.10.14.131 > > ssl=no > > startssl=no > > source_ip=10.10.14.2 > > proxy > > proxy > > pass=password > > > > Two "proxy" are from two "proxy" and "proxy=yes" settings passed from > > passdb. > > > > On Mon, Sep 17, 2018 at 3:03 PM Aki Tuomi <aki.tuomi(a)dovecot.fi> wrote: > > > > > Can you provide output of > > > > > > doveadm auth test some-user > > > Aki > > > > > > On 17.09.2018 14:58, Alexander Chekalin wrote: > > > > > > Dear Aki, > > > > > > we keep our users in LDAP so I when I even return 'proxy > host=backend_ip > > > tls=no' it won't use non-TLS connection. The same is when I remove > 'tls=no' > > > part. May there be any extra things I need to pass when I use LDAP? > > > > > > On Mon, Sep 17, 2018 at 2:07 PM Aki Tuomi <aki.tuomi(a)dovecot.fi> > wrote: > > > > > >> > > >> > > >> On 17.09.2018 13:59, Alexander Chekalin wrote: > > >> > Hi, > > >> > > > >> > I try to set up dovecot as a proxy server, to proxy requests to > > >> > several dovecot-based backend servers. I wand external clients who > > >> > connects to this proxy Dovecot to use TLS (this is easy to set up) > > >> > while want to have unsecured (plain IMAP/POP) connections to > backends. > > >> > > > >> > You see, links to backends are over LAN so no TLS needed, and these > > >> > backends are poor old machines (with old Docecots like 2.0.6) this > is > > >> > why I don't want to use TLS to acces backends. > > >> > > > >> > But as I did the test setup I can see proxy Dovecot uses TLS to > > >> > connect to backends. Is there any way I can specify this aspect of > > >> > Dovecot proxy? > > >> > > > >> > Please advice! > > >> > > > >> > Yours, > > >> > Alexander > > >> > > >> Dovecot does not use TLS/SSL when connecting to a backend server by > > >> default, you are probably specifying this in your proxy config or > > >> password database. > > >> > > >> Aki > > >> > > > > > > >

2 5

Proxy secured incoming POP3/IMAP4 to unsecure backend?
by Alexander Chekalin 17 Sep '18

17 Sep '18

Hi, I try to set up dovecot as a proxy server, to proxy requests to several dovecot-based backend servers. I wand external clients who connects to this proxy Dovecot to use TLS (this is easy to set up) while want to have unsecured (plain IMAP/POP) connections to backends. You see, links to backends are over LAN so no TLS needed, and these backends are poor old machines (with old Docecots like 2.0.6) this is why I don't want to use TLS to acces backends. But as I did the test setup I can see proxy Dovecot uses TLS to connect to backends. Is there any way I can specify this aspect of Dovecot proxy? Please advice! Yours, Alexander

3 2

Problem with envelope from in Pigeonhole after upgrade
by David Dodd 17 Sep '18

17 Sep '18

On our FreeBSD 11.2 servers, we recently upgraded from: dovecot2-2.2.26.0_1 dovecot-pigeonhole-0.4.16_1 to: dovecot-2.3.2.1 dovecot-pigeonhole-0.5.2_3 Prior to the upgrade the envelope sender was as expected. After the upgrade, the envelope sender for a user with a sieve script configured is now being set to: >From MAILER-DAEMON Tue Sep 04 13:31:39 2018 The same user without a sieve script in place gets: >From someone(a)gmail.com Tue Sep 04 11:06:04 2018 This happens when manually testing by directly executing: /usr/local/libexec/dovecot/dovecot-lda -d dgdtst -f someone(a)gmail.com It happens with an empty sieve script i.e. just using the implicit keep functionality. I have groped around looking for a configuration setting that might fix the behaviour but nothing I have tried helps. Can anyone suggest what to set to correct the problem ? -- Dave

2 7

Quotas not working with roundcube
by Mik J 17 Sep '18

17 Sep '18

Hello, I'm trying to have quotas working with roundcube. When I'm logged into roundcube I see 0% quota which seems to correspond to disabled. However I did everything that I could to enable quotas https://wiki.dovecot.org/Quota # doveconf -n auth_mechanisms = cram-md5 loginauth_verbose = yes debug_log_path = /tmp/dovecot.log first_valid_uid = 1000 log_path = /var/log/dovecot.log mail_gid = vmail mail_location = maildir:/home/mail/%d/%n/Maildir mail_plugins = notify replication quota mail_uid = vmail mbox_write_locks = fcntl mmap_disable = yes 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 { driver = bsdauth } passdb { args = /etc/dovecot/dovecot.passwd driver = passwd-file } plugin { quota_grace = 10%% quota_rule = *:storage=2G quota_rule2 = Trash:storage=+200M quota_rule3 = SPAM:ignore quota_status_nouser = DUNNO quota_status_overquota = 552 5.2.2 Mailbox is full quota_status_success = DUNNO quota_warning = storage=95%% quota-warning 95 %u quota_warning2 = storage=80%% quota-warning 80 %u } ssl_ca = /etc/ssl/certs/CA_Intermed_Lets_Encrypt.crt ssl_cert = </etc/ssl/certs/imap.mydomain.org.crt ssl_key = # hidden, use -P to show it ssl_key_password = # hidden, use -P to show it userdb { driver = passwd } userdb { driver = static } protocol lda { mail_plugins = notify replication quota } protocol imap { mail_plugins = notify replication quota imap_quota }

3 5

Dsync deleting mailboxes due to duplicate UIDs
by Andreas Thienemann 16 Sep '18

16 Sep '18

Hi, I am attempting to migrate a mailspool from a cyrus server to a dovecot server using the dsync backup approach as described in the wiki at <https://wiki.dovecot.org/Migration/Dsync>. The first attempt works great. Everything copies over and a quick glance over the spool looks good. Running doveadm backup again though results in the following error: dsync(eggs): Warning: Deleting mailbox 'INBOX.MailRestore': UID=39 already exists locally for a different mail: Headers hashes don't match (18d567fc7e258a67e47b629c8bb16500 vs 230354b2d5cad21ebbb4a7440b977adb) As promised, the folder MailRestore is gone after dsync finishes. Running doveadm backup again for the third time copies the folder again. Running it a fourth time gives the same error and the folder is gone. Trying to figure this out I initially had a few mails where the header hashes were 68b329da9893e34099c7d8ad5cb9c940, the md5sum of an empty string. Turns out that the source mailspool had a few broken emails and cleaning these out fixed most my issues. But in this case, I am stumped. UID 39 on the server is a valid mail with correct headers and everything. Is there a good way of identifying the two mails that seem to clash? I had a quick look at the dovecot code but did not see how the header_stream gets hashed into the hdr_hash used for comparing mails. Thanks, Andreas

3 5

dovecot_deleted_to_trash no longer compiles with dovecot 2.3.2.1
by Robert S 16 Sep '18

16 Sep '18

I have just upgraded to the latest version of dovecot (Gentoo amd64). When I try to compile the dovecot_deleted_to_trash plugin it won't compile. Is this plugin going to be updated (from https://github.com/lexbrugman/dovecot_deleted_to_trash - same error from https://github.com/NethServer/dovecot-deleted-to-trash)? # make cc \ -fPIC -shared -Wall \ -I/usr/include/dovecot \ -I/usr/include/dovecot/src \ -I/usr/include/dovecot/src/lib \ -I/usr/include/dovecot/src/lib-storage \ -I/usr/include/dovecot/src/lib-mail \ -I/usr/include/dovecot/src/lib-imap \ -I/usr/include/dovecot/src/lib-index \ -DHAVE_CONFIG_H \ src/deleted-to-trash-plugin.c -o lib_deleted_to_trash_plugin.so src/deleted-to-trash-plugin.c: In function ‘copy_deleted_mail_to_trash’: src/deleted-to-trash-plugin.c:117:16: error: too few arguments to function ‘mail box_transaction_begin’ dest_trans = mailbox_transaction_begin(trash_box, MAILBOX_TRANSACTION_FLAG_EX TERNAL); ^~~~~~~~~~~~~~~~~~~~~~~~~ In file included from /usr/include/dovecot/mail-storage-private.h:7:0, from src/deleted-to-trash-plugin.h:5, from src/deleted-to-trash-plugin.c:4: /usr/include/dovecot/mail-storage.h:656:1: note: declared here mailbox_transaction_begin(struct mailbox *box, ^~~~~~~~~~~~~~~~~~~~~~~~~ make: *** [Makefile:32: lib_deleted_to_trash_plugin.so] Error 1 Some further info: # dovecot -n # 2.3.2.1 (0719df592): /etc/dovecot/dovecot.conf # OS: Linux 4.14.8-gentoo-r1 x86_64 Gentoo Base System release 2.4.1 # Hostname: basement-vm.schmidli.com.au doveconf: Warning: please set ssl_dh=</etc/dovecot/dh.pem doveconf: Warning: You can generate it with: dd if=/var/lib/dovecot/ssl-parameters.dat bs=1 skip=88 | openssl dhparam -inform der > /etc/dovecot/dh.pem dict { expire = db:/var/lib/dovecot/expire.db } disable_plaintext_auth = no listen = * mail_location = maildir:~/.maildir mail_plugins = deleted_to_trash mailbox_alias 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 } plugin { deleted_to_trash_folder = Trash expire = Trash 7 Spam 30 expire_dict = proxy::expire

1 0

Re: auth_policy in a non-authenticating proxy chain
by Peter Mogensen 15 Sep '18

15 Sep '18

On 09/15/2018 10:41 AM, Aki Tuomi wrote: > Point of sending the success ones is to maintain whitelist as well as > blacklist so you know which ones you should not tarpit anymore. We > know it does scale as we have very large deployments using the whole > three request per login model. > > "Success" in a proxy which is not it self authenticating is only whether it know where to proxy the requested username to. I'm not sure whether this would be input to a whitelist. I'm not doubting that 3 req/login scales. /Peter

2 1

auth_policy in a non-authenticating proxy chain
by Peter Mogensen 15 Sep '18

15 Sep '18

Hi, I was looking into the new Authentication Policy feature: https://wiki2.dovecot.org/Authentication/Policy I had kinda hoped that I would be able to enfore this in a proxy running in front of several backends. This proxy does not authenticate. It use "nopassword". But I realize that the "succes" reported in the final authpolicy req. (command=report) is not what is actaully happening on the IMAP protocol level, but rather the result of the passdb chain in the proxy. (I should probably have predicted this, it's kinda reasonable). However... since the proxy use "nopassword", ALL passdb lookups result in "success", so the proxy will never report an authentication failure to the authpolicy server. This, of course, forces me to do the authpolicy check on the backend with a shared state, but It would still have been nice to have the proxy being able to do the first "command=allow" req. and reject attemps already there even though the backend does "command=report". /Peter

3 4