- dovecot - dovecot.org

[Dovecot] mails delivered to the wrong user when using lmtp_proxy and reject_unverified_recipient
by Peer Heinlein 02 Aug '13

02 Aug '13

Hi, looks like we detected a serious bug in dovecot's lmtp proxying where e-mails are delivered to the wrong user. The setup is: *) Dovecot is configured with "lmtp_proxy=yes" # Support proxying to other LMTP/SMTP servers by performing passdb lookups. lmtp_proxy = yes *) Postfix uses "dynamic recipient verification", so Postfix starts sending a (verify) mail by LMTP to dovecot, but quits the lmtp-session right after the RCPT TO:. No DATA-stage is reached in the protocol and no real e-mail is sent. But Postfix had a LMTP-connection for "user1". *) Just some seconds later a "real" e-mail to "user2" has to be delivered to dovecot by LMTP. But Dovecot will deliver this mail to the wrong "user1" instead of "user2". Looks like dovecot re-uses the (still opened?) lmtp-proxy-connection from "user1" to deliver an e-mail to "user2". Have a log at the protocol: 1) There's a verify call to user1 from Postfix: Jul 19 13:49:49 mailms postfix/lmtp[9842]: DE653280C51: to=<user1(a)example.com>, relay=localhost[127.0.0.1]:24, conn_use=2, delay=120, delays=117/0.45/0/2.5, dsn=2.1.5, status=deliverable (250 2.1.5 OK) 2) Just five seconds later the e-mail to user2 (see Postfix' point of view in the last line) is delivered to user2 (see result from Dovecot in the last line): Jul 19 13:50:04 mailms dovecot: lmtp(10965, kraemer): save: box=INBOX, uid=49880, msgid=<59798276-E5D1-4053-A570-9901B731DF5D(a)example.come>, size=11020 Jul 19 13:50:04 mailms dovecot: lmtp(10965, kraemer): 1zTeKrMn6VHVKgAAhyqEuA: msgid=<59798276-E5D1-4053-A570-9901B731DF5D(a)example.com>: saved mail to INBOX Jul 19 13:50:04 mailms postfix/lmtp[10953]: C25FC280BE5: to=<user2(a)example.com>, relay=localhost[127.0.0.1]:24, conn_use=19, delay=116, delays=115/0.53/0/0.33, dsn=2.0.0, status=sent (250 2.0.0 <user2> 1zTeKrMn6VHVKgAAhyqEuA Saved) Same with user3 and user4: Jul 19 14:47:53 mailms postfix/lmtp[10845]: C389A2809D7: to=<user3(a)example.com>, relay=localhost[127.0.0.1]:24, delay=4.7, delays=3.7/0.87/0/0.19, dsn=2.1.5, status=deliverable (250 2.1.5 OK) Jul 19 14:47:55 mailms dovecot: lmtp(26546, fs211113): save: box=INBOX, uid=8504, msgid=<928729810.113.1374238063381(a)example.com>, size=233151 Jul 19 14:47:55 mailms dovecot: lmtp(26546, fs211113): MbMvI2816VGyZwAAhyqEuA: msgid=<928729810.113.1374238063381(a)example.com>: saved mail to INBOX Jul 19 14:47:55 mailms postfix/lmtp[22524]: 6F0D2280A6E: to=<user4(a)example.com>, relay=localhost[127.0.0.1]:24, conn_use=2, delay=10, delays=8.4/1/0/0.8, dsn=2.0.0, status=sent (250 2.0.0 <user3> MbMvI2816VGyZwAAhyqEuA Saved) The user itself is quite normal in the user database (but has a mailhost=127.0.0.1 set): root@mailms:/etc/dovecot/conf.d# doveadm user user2(a)example.com userdb: user2(a)example.com uid : 5000 gid : 5000 home : /srv/mail/user2 root@mailms:/etc/dovecot/conf.d# doveadm auth user2(a)example.com Password: passdb: user2(a)example.com auth failed extra fields: user=user2 Peer -- Heinlein Support GmbH Schwedter Str. 8/9b, 10119 Berlin http://www.heinlein-support.de Tel: 030 / 405051-42 Fax: 030 / 405051-19 Zwangsangaben lt. §35a GmbHG: HRB 93818 B / Amtsgericht Berlin-Charlottenburg, Geschäftsführer: Peer Heinlein -- Sitz: Berlin

4 3

[Dovecot] Auth-woker log lines
by Grzegorz Staniak 02 Aug '13

02 Aug '13

Hi, Are log lines like the following: Jul 28 15:30:50 mx1.somewhere dovecot: auth-worker(18980): sql(user(a)domain,217.67.x.x): unknown user Jul 28 15:32:56 mx1.somewhere dovecot: auth-worker(18980): sql(user(a)domain,212.182.x.x): Password mismatch written every time any authetntication phase fails, in SASL as well as in POP3/IMAP mailbox access? I have a legacy app that would gain a lot if I could distinguish between failed SASL authentication and failed POP3/IMAP authentication. I know I can use the "auth failed" lines for the latter case, and postfix log lines in the former, but they don't contain user info. Is there a way for log monitoring software to get failed login lines with both user info, and the reason (SASL, POP3, IMAP)? Regards, -- Grzegorz Staniak <gstaniak [at] gmail _dot_ com>

2 1

[Dovecot] 2.2.4 - quota-status changing the user it is running as
by Axel Luttgens 02 Aug '13

02 Aug '13

Hello, I'm currently experimenting with this quota-status service configuration: service quota-status { client_limit = 1 executable = quota-status -p postfix # Let's make the default explicit. user = root unix_listener /var/spool/postfix/private/quota-policyd { user = postfix } } The idea is to run the service as root during the preliminary tests (at worst, since the service is going to be used as a policy daemon by Postfix only, it shouldn't be very problematic to have it running as root anyway). A user, with address john.doe(a)example.com and identified as john.doe, is known to be over-quota. A "doveadm quota get" returns: $ sudo doveadm quota get -u john.doe Password: Quota name Type Value Limit % Quota utilisateur STORAGE 5 5 100 Quota utilisateur MESSAGE 9 - 0 and messages for that user are correctly rejected by lmtp: dovecot[4989]: lmtp(5069, john.doe(a)example.com) QWSWLgrP4lF7FAAA5Q0ykw: msgid=<20130714161643.9085DF176F2(a)ALMba.local>: save failed to INBOX: Quota exceeded (mailbox for user is full) Let's now simulate a connection from Postfix: $ sudo -u postfix telnet /_ROOT/var/spool/postfix/private/quota-policyd Trying /_ROOT/var/spool/postfix/private/quota-policyd... Connected to (null). Escape character is '^]'. A look at the output of top (excerpt) confirms that quota-status is running as root: PID COMMAND UID 5100 quota-status 0 Going on with our telnet session: recipient=john.doe(a)example.com size=10000 action=OK Hmmm... OK, this may be a config problem of mine which may require further investigation. Anyway, looking at top's output: PID COMMAND UID 5100 quota-status 999 it appears that quota-status is now running as the mail_uid/mail_gid user; the switch happens immediately after having entered the empty line in the telnet session. Let's then try to go further within the telnet session: recipient=john.doe(a)example.com size=10000 action=DEFER_IF_PERMIT Internal error occurred. Refer to server log for more information. ^] telnet> quit Connection closed. and a look at the log indeed reveals that quota-status doesn't have sufficient privileges anymore: dovecot[4989]: quota-status(john.doe(a)example.com): Error: user john.doe(a)example.com: Error reading configuration: net_connect_unix(/_ROOT/var/run/dovecot/config) failed: Permission denied Is this the expected behavior, to have quota-status switch to another user? TIA, Axel

3 8

[Dovecot] [PATCH] quota-status: allow different action for messages that are too large (over maximum quota)
by Ulrich Zehl 02 Aug '13

02 Aug '13

This proposed patch to quota-status allows the administrator to return different actions for messages that (a) cannot currently fit in the quota (because there's not enough free space) and (b) are over the maximum quota limit (and could not even be delivered if the mailbox was empty). Configuration works like this: plugin { quota_status_overquota = DEFER_IF_PERMIT 4.2.2 Mailbox full quota_status_toolarge = REJECT 5.2.3 Message length exceeds administrative limit } The patch is backwards compatible: If quota_status_toolarge is not set, quota_status_overquota is returned. No changes are necessary in setups where the new feature is not desired. The reason I developed this patch is: In my environment, I have mostly POP3 mailboxes, where condition (a) changes frequently (usually after each login), but condition (b) will almost never change (users do not get additional mailbox space often). Thus, I'd like to avoid bouncing mails that could be delivered half an hour later. This patch works for 2.2 and 2.1. (This assumes you have already applied the trivial patch I sent earlier; otherwise patch(1) will report some fuzz.) What do you think of this? Ulrich --- a/src/plugins/quota/quota-status.c 2013-08-01 18:05:24.000000000 +0200 +++ b/src/plugins/quota/quota-status.c 2013-08-01 18:03:30.000000000 +0200 @@ -46,13 +46,12 @@ } static int -quota_check(struct mail_user *user, uoff_t mail_size, const char **error_r) +quota_check(struct mail_user *user, uoff_t mail_size, const char **error_r, bool *too_large) { struct quota_user *quser = QUOTA_USER_CONTEXT(user); struct mail_namespace *ns; struct mailbox *box; struct quota_transaction_context *ctx; - bool too_large; int ret; if (quser == NULL) { @@ -64,7 +63,7 @@ box = mailbox_alloc(ns->list, "INBOX", 0); ctx = quota_transaction_begin(box); - ret = quota_test_alloc(ctx, I_MAX(1, mail_size), &too_large); + ret = quota_test_alloc(ctx, I_MAX(1, mail_size), too_large); quota_transaction_rollback(&ctx); mailbox_free(&box); @@ -82,6 +81,7 @@ struct mail_storage_service_user *service_user; struct mail_user *user; const char *value = NULL, *error; + bool too_large; int ret; if (client->recipient == NULL) { @@ -98,14 +98,21 @@ if (ret == 0) { value = nouser_reply; } else if (ret > 0) { - if ((ret = quota_check(user, client->size, &error)) > 0) { + if ((ret = quota_check(user, client->size, &error, &too_large)) > 0) { /* under quota */ value = mail_user_plugin_getenv(user, "quota_status_success"); if (value == NULL) value = "OK"; } else if (ret == 0) { - /* over quota */ - value = mail_user_plugin_getenv(user, "quota_status_overquota"); + if (too_large == TRUE) { + /* even over maximum quota */ + value = mail_user_plugin_getenv(user, "quota_status_toolarge"); + if (value == NULL) /* backwards compatibility */ + value = mail_user_plugin_getenv(user, "quota_status_overquota"); + } else { + /* over quota */ + value = mail_user_plugin_getenv(user, "quota_status_overquota"); + } if (value == NULL) value = t_strdup_printf("554 5.2.2 %s", error); }

2 1

Re: [Dovecot] LDA vs. LMTP
by Joseph Tam 02 Aug '13

02 Aug '13

(Weird: this message digest got dumped into Google's spam folder. Maybe it didn't like the string in a later post (obfuscated here) master(dot)cf, which in the context of this mailing list is a postfix configuration file, but which Gmail interpret as a website. However, that domain is a SURBL/DBL blacklisted URI). Ben Morrow writes: > > > And when it's running as root there is always the danger > > > of privilege escalation. LDA only runs when it's needed and since it > > > uses only user rights it shoudbe more harmless. > > > > ... > > (I'm assuming LMTP stays as root, and not spawning off user > > processes to do the real work.) > > It doesn't stay as root; Dovecot's LMTP switches down to the user's uid > to perform delivery, including sieve scripts. I stand corrected. This removes the other objection that the original poster for running a peristent LMTP process. Joseph Tam <jtam.home(a)gmail.com>

1 0

[Dovecot] misconception in uid and gid
by Felix Rubio Dalmau 02 Aug '13

02 Aug '13

Hi, I'm using a SQL database to store the usernames and passwords of my users. However, I have found that I must explicitly return the uid and gid with every query, even though all virtual users use the same uid and gid. The query, that works, is: user_query = SELECT mail, 'vmail' AS uid, 'vmail' AS gid, home, CONCAT('*:storage=', quota_MB*1024) AS quota_rule \ FROM virtual_users AS V LEFT JOIN virtual_domains AS D ON V.domain_id=D.id WHERE V.user='%n' AND D.name='%d' Nevertheless, if I remove the segments "'vmail' AS uid, 'vmail' AS gid" and modify the file conf.d/10-mail.conf to show mail_uid = vmail mail_gid = vmail I get this error: dovecot: auth: Error: userdb(<mail>): client doesn't have lookup permissions for this user: userdb reply doesn't contain uid (change userdb socket permissions) Does anybody know what is wrong in my set-up? Regards, Felix

3 4

Re: [Dovecot] dovecot Digest, Vol 124, Issue 4
by Felix Rubio Dalmau 02 Aug '13

02 Aug '13

Hello Axel, but then I don't get it: I thought that "uid" and "gid" in the user_query where used to access the local FS, whereas the "unix_listener auth-userdb" are used to indicate under which owner/group must be auth-userdb run... although maybe I'm wrong :-S :-) What I'm looking forward to is to eliminate the need for returning these two fixed items, as long as all the virtual_users will be using the same uid and gid. Does anybody know how can I do it?? Regards, and thank you! Felix > Date: Thu, 1 Aug 2013 17:50:33 +0200 > From: Axel Luttgens <AxelLuttgens(a)swing.be> > To: Dovecot Mailing List <dovecot(a)dovecot.org> > Subject: Re: [Dovecot] misconception in uid and gid > Message-ID: <89AC3212-EFE6-41E5-A22D-97DC8E771218(a)swing.be> > Content-Type: text/plain; charset=iso-8859-1 > > Le 1 ao?t 2013 ? 15:43, Felix Rubio Dalmau a ?crit : > > > [...] > > > > Nevertheless, if I remove the segments "'vmail' AS uid, 'vmail' AS gid" and modify the file conf.d/10-mail.conf to show > > > > mail_uid = vmail > > mail_gid = vmail > > > > I get this error: > > > > dovecot: auth: Error: userdb(<mail>): client doesn't have lookup permissions for this user: userdb reply doesn't contain uid (change userdb socket permissions) > > > > Does anybody know what is wrong in my set-up? > > Hello Felix, > > Yes and no... > This still remains a bit unclear to me, but you could try something like this: > > service auth { > > unix_listener auth-userdb { > > # default: user = $default_internal_user > group = vmail # default: group = > mode = 0660 # default: mode = 0666 > } > } > > The code has some provisions to avoid the auth-userdb to be too widely open, in spite of that default mode 0666. Changing that mode short-circuits those provisions, and the above is the most secure setting I could think of in the case of a single uid/gid setup. > > HTH, > Axel

2 1

[Dovecot] Wrong remote IP (rip) in mail.log using IMAP login
by tyli 02 Aug '13

02 Aug '13

Dear list users While trying to secure our dovecot server with fail2ban I came across the following problem: We use dovecot (1.2.9, ubuntu package) behind a NAT, and failed login attempts are logged with our firewall as the remote ip. Example: Apr 15 08:36:26 mail dovecot: imap-login: Disconnected (auth failed, 6 attempts): user=<xy>, method=PLAIN, rip=192.168.0.1, lip=192.168.0.3 Therefore I would ban 192.168.0.1 which means that I ban EVERY user. Funny thing is that POP3 login attempts are logged correctly: Apr 13 11:05:50 mail dovecot: pop3-login: Disconnected (auth failed, 1 attempts): user=<sgvyniwx>, method=PLAIN, rip=217.81.27.55, lip=192.168.0.3 Any ideas how to change this? Thanks in advance tyli

5 5

[Dovecot] Dovecot v2.2.4 (ebbe00999da8) doveadm crash
by e-frog 01 Aug '13

01 Aug '13

Hello, Current Dovecot v2.2 HG crashes using the following command: $ doveadm mailbox status -A -t unseen 'virtual/unread' Please see backtrace below and dovecot -n output attached. Thanks, e-frog gdb /usr/bin/doveadm 24023 GNU gdb (GDB) 7.5.91.20130417-cvs-ubuntu Copyright (C) 2013 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "i686-linux-gnu". For bug reporting instructions, please see: <http://www.gnu.org/software/gdb/bugs/>... Reading symbols from /usr/bin/doveadm...Reading symbols from /usr/lib/debug/usr/bin/doveadm...done. done. Attaching to program: /usr/bin/doveadm, process 24023 warning: unable to open /proc file '/proc/24023/status' warning: unable to open /proc file '/proc/24023/status' ptrace: No such process. [New LWP 24023] warning: Can't read pathname for load map: Input/output error. [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/i386-linux-gnu/libthread_db.so.1". Core was generated by `doveadm mailbox status -A -t unseen virtual/unread'. Program terminated with signal 11, Segmentation fault. #0 i_stream_next_line (stream=0x0) at istream.c:382 382 istream.c: No such file or directory. (gdb) bt full #0 i_stream_next_line (stream=0x0) at istream.c:382 _stream = <optimized out> pos = <optimized out> #1 0xb754656b in auth_master_user_list_next (ctx=0xb91bfbd8) at auth-master.c:689 line = <optimized out> #2 0xb7634a0b in mail_storage_service_all_next (ctx=0xb91c4ba8, username_r=username_r@entry=0xbf8e6630) at mail-storage-service.c:1260 __FUNCTION__ = "mail_storage_service_all_next" #3 0xb7752d46 in doveadm_mail_cmd_get_next_user (ctx=0xb91be478, username_r=0xbf8e6630) at doveadm-mail.c:416 No locals. #4 0xb7753e95 in doveadm_mail_all_users (wildcard_user=0x0, argv=0xb91ba1ec, ctx=0xb91be478) at doveadm-mail.c:368 ret = <optimized out> input = {module = 0x0, service = 0xb77918f3 "doveadm", username = 0xb91bf8a0 "test1(a)local.lan", session_id = 0x0, local_ip = {family = 0, u = { ip6 = {__in6_u = {__u6_addr8 = '\000' <repeats 15 times>, __u6_addr16 = {0, 0, 0, 0, 0, 0, 0, 0}, __u6_addr32 = {0, 0, 0, 0}}}, ip4 = { s_addr = 0}}}, remote_ip = {family = 0, u = {ip6 = {__in6_u = {__u6_addr8 = '\000' <repeats 15 times>, __u6_addr16 = {0, 0, 0, 0, 0, 0, 0, 0}, __u6_addr32 = {0, 0, 0, 0}}}, ip4 = {s_addr = 0}}}, local_port = 0, remote_port = 0, userdb_fields = 0x0, flags_override_add = (unknown: 0), flags_override_remove = (unknown: 0), no_userdb_lookup = 0} user_idx = 0 user = 0xb91bf8a0 "test1(a)local.lan" error = 0x10 <Address 0x10 out of bounds> #5 doveadm_mail_cmd (argv=0xb91ba1ec, argc=-1189371412, cmd=0xbf8e668c) at doveadm-mail.c:523 ctx = 0xb91be478 getopt_args = <optimized out> ret = <optimized out> c = <optimized out> wildcard_user = 0x0 error = <optimized out> #6 doveadm_mail_try_run (cmd_name=cmd_name@entry=0xb91ba200 "mailbox", argc=5, argc@entry=6, argv=argv@entry=0xb91ba1dc) at doveadm-mail.c:608 cmd = 0xbf8e668c #7 0xb7752903 in main (argc=6, argv=0xb91ba1dc) at doveadm.c:398 cmd_name = 0xb91ba200 "mailbox" i = <optimized out> quick_init = false c = <optimized out> (gdb)

2 4

[Dovecot] [PATCH] quota-status: remove unnecessary newlines
by Ulrich Zehl 01 Aug '13

01 Aug '13

Here's a trivial patch for quota-status.c that removes unecessary newlines present in the default overquota message. It applies to 2.2 and 2.1. Ulrich --- a/src/plugins/quota/quota-status.c 2013-08-01 14:35:35.000000000 +0200 +++ b/src/plugins/quota/quota-status.c 2013-08-01 17:58:26.000000000 +0200 @@ -107,7 +107,7 @@ /* over quota */ value = mail_user_plugin_getenv(user, "quota_status_overquota"); if (value == NULL) - value = t_strdup_printf("554 5.2.2 %s\n\n", error); + value = t_strdup_printf("554 5.2.2 %s", error); } value = t_strdup(value); /* user's pool is being freed */ mail_user_unref(&user);

1 0