May 2018 - dovecot - dovecot.org

Race condition when setting flags (\Deleted) + expunge quickly, leaving mails not deleted
by Tomas Forsman 12 Jun '18

12 Jun '18

Hello. I seem to have found a race condition, when setting flags on multiple emails rapidly. 5 commands including login to reproduce. Problem found using mutt in real world usage. Seems to happen both with UID STORE 1:3 and UID STORE 1,2,3 .. I have tried with the following packages, with a minimized config in a throwaway vm: https://packages.debian.org/stretch/dovecot-core Package: dovecot-core (1:2.2.27-3+deb9u2) https://packages.debian.org/stretch-backports/dovecot-core Package: dovecot-core (1:2.2.34-2~bpo9+1) Also tested pristine: https://dovecot.org/releases/2.3/dovecot-2.3.0.1.tar.gz https://dovecot.org/releases/2.2/dovecot-2.2.34.tar.gz https://dovecot.org/releases/2.2/dovecot-2.2.35.tar.gz .. built on Debian9 with: # apt-get build-dep dovecot-core # bash configure --prefix=/scratch/dovecot-test Tried with both mailbox over NFS (backed by ext4), and also mailbox on ext4. On my test systems, I am only able to reproduce it with mbox_lazy_writes=no, but with production settings on a Ubuntu 14.04 with 1:2.2.9-1ubuntu2.4 that has mbox_lazy_writes=yes it is still reproducable with mutt. I have found two different ways to reproduce it with netcat + cut'n'paste, below is done with 2.2.34/35. All servers are VMs in Ganeti/kvm clusters (Debian hosts in one cluster&organization, Ubuntu host in another cluster&organization). Debian minimized config: > doveconf -n # 2.2.34 (874deae): /etc/dovecot/dovecot.conf # Pigeonhole version 0.4.22 (22940fb7) # OS: Linux 4.9.0-6-amd64 x86_64 Debian 9.4 # Hostname: mail2-vm.cs.umu.se auth_mechanisms = plain login auth_verbose = yes debug_log_path = syslog info_log_path = syslog mail_location = mbox:~/Mail:INBOX=/var/mail/%u mail_privileged_group = mail mbox_lazy_writes = no namespace inbox { inbox = yes location = prefix = } passdb { driver = pam } protocols = imap service auth { client_limit = 2500 unix_listener auth-client { group = postfix mode = 0660 } } service imap-login { inet_listener imap { port = 143 } } service imap { executable = imap } ssl = no syslog_facility = local1 userdb { driver = passwd } Attached is a mailbox that can be used to reproduce the problem, called "error-box", anonymized with mbox-anonymize.pl. It was created by sending test messages with mutt 1.7.2 through Dovecot/Postfix/Amavis. **** Method 1, with IDLE: Setup a local account, MY-USER / MY-PASSWORD (replace below). > mkdir Mail > cp error-box Mail/error1 > nc -v localhost 143 Paste the following 3 lines in a go: a0000 LOGIN "MY-USER" "MY-PASSWORD" a0010 SELECT "error1" a0020 IDLE then wait a second.. then paste the following 3 in a go: DONE a0030 UID STORE 1:3 +FLAGS (\Deleted) a0040 EXPUNGE Notice that the results from UID STORE + EXPUNGE gives: * 2 FETCH (UID 2 FLAGS (\Recent)) * 3 FETCH (UID 3 FLAGS (\Recent)) * 1 EXPUNGE * 3 RECENT a0030 OK Store completed (0.004 + 0.000 + 0.003 secs). a0040 OK Expunge completed (0.004 + 0.000 + 0.003 secs). Verify what's left in the mailbox with: a0050 UID FETCH 1:4 FLAGS * 1 FETCH (UID 2 FLAGS (\Recent)) * 2 FETCH (UID 3 FLAGS (\Recent)) * 3 FETCH (UID 4 FLAGS (\Recent)) a0050 OK Fetch completed (0.001 + 0.000 secs). Since 1 to 3 should be \Deleted and then EXPUNGEd, we should only have UID 4 left, not uid 2-4 without Deleted flag. Try same again, but with 1 inserted delay (use a new mailbox name for every test, otherwise UID numbering will vary, indexes etc): > cp error-box Mail/error2 > nc -v localhost 143 Paste the following 3 lines in a go: a0000 LOGIN "MY-USER" "MY-PASSWORD" a0010 SELECT "error2" a0020 IDLE then wait a second.. then paste 1 line: DONE then [newly introduced] wait a second.. then paste 2 lines: a0030 UID STORE 1:3 +FLAGS (\Deleted) a0040 EXPUNGE Notice that the results from UID STORE + EXPUNGE gives: * 1 FETCH (UID 1 FLAGS (\Deleted \Recent)) * 2 FETCH (UID 2 FLAGS (\Deleted \Recent)) * 3 FETCH (UID 3 FLAGS (\Deleted \Recent)) a0030 OK Store completed (0.003 + 0.000 + 0.002 secs). * 3 EXPUNGE * 2 EXPUNGE * 1 EXPUNGE * 1 RECENT a0040 OK Expunge completed (0.003 + 0.000 + 0.002 secs). Verify what's left in the mailbox with: a0050 UID FETCH 1:4 FLAGS * 1 FETCH (UID 4 FLAGS (\Recent)) a0050 OK Fetch completed (0.001 + 0.000 secs). So if I wait some between DONE and UID STORE, it seems to be ok. If I, in the problem-triggering method, change a0030 UID STORE 1:3 +FLAGS (\Deleted) a0040 EXPUNGE to a0030 UID STORE 1:3 +FLAGS (\Deleted) a0035 UID FETCH 1:4 FLAGS a0040 EXPUNGE The untagged responses become: * 1 FETCH (UID 1 FLAGS (\Deleted \Recent)) * 2 FETCH (UID 2 FLAGS (\Deleted \Recent)) * 3 FETCH (UID 3 FLAGS (\Deleted \Recent)) * 4 FETCH (UID 4 FLAGS (\Recent)) * 2 FETCH (UID 2 FLAGS (\Recent)) * 3 FETCH (UID 3 FLAGS (\Recent)) * 1 EXPUNGE * 3 RECENT Full transcript (in&out from screen) from first, failing, run: localhost [127.0.0.1] 143 (imap2) open * OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE AUTH=PLAIN AUTH=LOGIN] Dovecot (Debian) ready. a0000 LOGIN "MY-USERNAME" "MY-PASSWORD" a0010 SELECT "error1" a0020 IDLE a0000 OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE SORT SORT=DISPLAY THREAD=REFERENCES THREAD=REFS THREAD=ORDEREDSUBJECT MULTIAPPEND URL-PARTIAL CATENATE UNSELECT CHILDREN NAMESPACE UIDPLUS LIST-EXTENDED I18NLEVEL=1 CONDSTORE QRESYNC ESEARCH ESORT SEARCHRES WITHIN CONTEXT=SEARCH LIST-STATUS BINARY MOVE SNIPPET=FUZZY] Logged in * FLAGS (\Answered \Flagged \Deleted \Seen \Draft) * OK [PERMANENTFLAGS (\Answered \Flagged \Deleted \Seen \Draft \*)] Flags permitted. * 4 EXISTS * 4 RECENT * OK [UNSEEN 1] First unseen. * OK [UIDVALIDITY 1520944853] UIDs valid * OK [UIDNEXT 5] Predicted next UID a0010 OK [READ-WRITE] Select completed (0.003 + 0.000 + 0.004 secs). + idling DONE a0030 UID STORE 1:3 +FLAGS (\Deleted) a0040 EXPUNGE a0020 OK Idle completed (7.805 + 7.804 + 7.804 secs). * 2 FETCH (UID 2 FLAGS (\Recent)) * 3 FETCH (UID 3 FLAGS (\Recent)) * 1 EXPUNGE * 3 RECENT a0030 OK Store completed (0.004 + 0.000 + 0.003 secs). a0040 OK Expunge completed (0.004 + 0.000 + 0.003 secs). a0050 UID FETCH 1:4 FLAGS * 1 FETCH (UID 2 FLAGS (\Recent)) * 2 FETCH (UID 3 FLAGS (\Recent)) * 3 FETCH (UID 4 FLAGS (\Recent)) a0050 OK Fetch completed (0.001 + 0.000 secs). Full transcript (in&out from screen) from the second, slower, run: localhost [127.0.0.1] 143 (imap2) open * OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE AUTH=PLAIN AUTH=LOGIN] Dovecot (Debian) ready. a0000 LOGIN "MY-USERNAME" "MY-PASSWORD" a0010 SELECT "error2" a0020 IDLE a0000 OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE SORT SORT=DISPLAY THREAD=REFERENCES THREAD=REFS THREAD=ORDEREDSUBJECT MULTIAPPEND URL-PARTIAL CATENATE UNSELECT CHILDREN NAMESPACE UIDPLUS LIST-EXTENDED I18NLEVEL=1 CONDSTORE QRESYNC ESEARCH ESORT SEARCHRES WITHIN CONTEXT=SEARCH LIST-STATUS BINARY MOVE SNIPPET=FUZZY] Logged in * FLAGS (\Answered \Flagged \Deleted \Seen \Draft) * OK [PERMANENTFLAGS (\Answered \Flagged \Deleted \Seen \Draft \*)] Flags permitted. * 4 EXISTS * 4 RECENT * OK [UNSEEN 1] First unseen. * OK [UIDVALIDITY 1520944957] UIDs valid * OK [UIDNEXT 5] Predicted next UID a0010 OK [READ-WRITE] Select completed (0.003 + 0.000 + 0.004 secs). + idling DONE a0020 OK Idle completed (2.830 + 2.829 + 2.829 secs). a0030 UID STORE 1:3 +FLAGS (\Deleted) a0040 EXPUNGE * 1 FETCH (UID 1 FLAGS (\Deleted \Recent)) * 2 FETCH (UID 2 FLAGS (\Deleted \Recent)) * 3 FETCH (UID 3 FLAGS (\Deleted \Recent)) a0030 OK Store completed (0.003 + 0.000 + 0.002 secs). * 3 EXPUNGE * 2 EXPUNGE * 1 EXPUNGE * 1 RECENT a0040 OK Expunge completed (0.003 + 0.000 + 0.002 secs). a0050 UID FETCH 1:4 FLAGS * 1 FETCH (UID 4 FLAGS (\Recent)) a0050 OK Fetch completed (0.001 + 0.000 secs). Transcript from .muttdebug0 when using mutt 1.7.2 against the production server with NFS backed storage, marking message 1-3 for deletion and hitting $ for sync. This is how the problem was first found. [2018-03-12 13:46:28] 5> a0020 IDLE^M [2018-03-12 13:46:28] 5< + idling ... [2018-03-12 13:46:29] Marking 3 messages deleted... [2018-03-12 13:46:29] Expunging messages from server... [2018-03-12 13:46:29] 5> DONE^M a0021 UID STORE 1:3 +FLAGS.SILENT (\Deleted)^M a0022 EXPUNGE^M [2018-03-12 13:46:29] 5< a0020 OK Idle completed (0.001 + 1.634 + 1.633 secs). [2018-03-12 13:46:29] 5< * 2 FETCH (FLAGS (\Recent)) [2018-03-12 13:46:29] Handling FETCH [2018-03-12 13:46:29] FETCH response ignored for this message [2018-03-12 13:46:29] 5< * 3 FETCH (FLAGS (\Recent)) [2018-03-12 13:46:29] Handling FETCH [2018-03-12 13:46:29] FETCH response ignored for this message [2018-03-12 13:46:29] 5< * 1 EXPUNGE [2018-03-12 13:46:29] Handling EXPUNGE [2018-03-12 13:46:29] 5< * 3 RECENT [2018-03-12 13:46:29] 5< a0021 OK Store completed (0.001 + 0.000 + 0.001 secs). [2018-03-12 13:46:29] 5< a0022 OK Expunge completed (0.001 + 0.000 secs). [2018-03-12 13:46:29] IMAP queue drained [2018-03-12 13:46:29] imap_cmd_finish: Expunging mailbox *** Method 2, without IDLE (seems to require that I fetch flags before the wait/store) Copy'n'paste the following: a0000 LOGIN "MY-USER" "MY-PASSWORD" a0010 SELECT "error20" a0015 FETCH 1:4 (UID FLAGS) <wait a second, then paste the next two in one go> a0020 UID STORE 1:3 +FLAGS (\Deleted) a0025 EXPUNGE Gives the following transcript: mail3-vm:~/Mail> nc -v localhost 143 localhost [127.0.0.1] 143 (imap2) open * OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE AUTH=PLAIN AUTH=LOGIN] Dovecot ready. a0000 LOGIN "MY-USER" "MY-PASSWORD" a0010 SELECT "error20" a0015 FETCH 1:4 (UID FLAGS) a0000 OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE SORT SORT=DISPLAY THREAD=REFERENCES THREAD=REFS THREAD=ORDEREDSUBJECT MULTIAPPEND URL-PARTIAL CATENATE UNSELECT CHILDREN NAMESPACE UIDPLUS LIST-EXTENDED I18NLEVEL=1 CONDSTORE QRESYNC ESEARCH ESORT SEARCHRES WITHIN CONTEXT=SEARCH LIST-STATUS BINARY MOVE SNIPPET=FUZZY] Logged in * FLAGS (\Answered \Flagged \Deleted \Seen \Draft) * OK [PERMANENTFLAGS (\Answered \Flagged \Deleted \Seen \Draft \*)] Flags permitted. * 4 EXISTS * 4 RECENT * OK [UNSEEN 1] First unseen. * OK [UIDVALIDITY 1521552107] UIDs valid * OK [UIDNEXT 5] Predicted next UID a0010 OK [READ-WRITE] Select completed (0.001 + 0.000 + 0.001 secs). * 1 FETCH (UID 1 FLAGS (\Recent)) * 2 FETCH (UID 2 FLAGS (\Recent)) * 3 FETCH (UID 3 FLAGS (\Recent)) * 4 FETCH (UID 4 FLAGS (\Recent)) a0015 OK Fetch completed (0.001 + 0.000 secs). a0020 UID STORE 1:3 +FLAGS (\Deleted) a0025 EXPUNGE * 2 FETCH (UID 2 FLAGS (\Recent)) * 3 FETCH (UID 3 FLAGS (\Recent)) * 1 EXPUNGE * 3 RECENT a0020 OK Store completed (0.001 + 0.000 secs). a0025 OK Expunge completed (0.001 + 0.000 secs). /Tomas -- Tomas Forsman, stric(a)cs.umu.se, http://people.cs.umu.se/stric/ `- SysAdmin at Computing Science, University of Umeå

2 3

Re: 2.3.1 Replication is throwing scary errors
by Reuben Farrelly 09 Jun '18

09 Jun '18

Hi Andy, Funny you say that - I've been testing this very problem for some time today to narrow down the commits where I think the problem started occurring. So far I have been able to roll forward to commit from git: 74ef8506dcb946acc42722683e49fdcb964eed19 >> "doveadm: Unref header search context after use" . So far I have been running on that code for a few hours and I'm not yet seeing the problem, which indicates the issue is probably subsequent to this. I'm of the view that the problem started occurring somewhere between that commit above on 6 March and the final 2.3.1 release on 27 March. There are a lot of changes between those two dates, and testing to narrow down to one specific commit is extremely time consuming. It would be really helpful if even one of the Dovecot developers could comment on this and give us some direction (or candidate commits to investigate) or just let us know if the root cause has been found. This bug is a showstopper and has stopped me tracking master-2.3 for over 3 months now, as I can't test later builds or even upgrade to the 2.3.1 release while replication is so broken. Reuben > Message: 1 > Date: Sun, 6 May 2018 13:21:57 +1000 > From: Andy Weal <andy(a)bizemail.com.au> > To: dovecot(a)dovecot.org > Subject: 2.3.1 Replication is throwing scary errors > Message-ID: <f4274128-350b-9a74-a344-8e6caeffa112(a)bizemail.com.au> > Content-Type: text/plain; charset="utf-8"; Format="flowed" > > Hi all, > > New to the mailing lists but have joined up because of above */2.3.1 > Replication is throwing scary errors > > > /*Brief system configuration > ??? MX1 - Main > ??? ??? Freebsd 11.1-Release-p9 > ??? ??? Hosted on a Vultr VM in Sydney AU > ??? ??? MTA = Postfix 3.4-20180401 > ??? ??? Dovecot = 2.3.1 > ??? ??? File system = ufs > ??? MX2 - Backup > ??? ??? Freebsd 11.1-Release-p9 > ??? ???? Running on bare metal - no VM or jails > ??? ??? MTA = Postfix 3.4-20180401 > ??? ??? Dovecot = 2.3.1 > ??? ??? File system = ufs ( on SSD) > > */ > /*Brief sequence of events > > 1. ??? apx 10 days back upgraded both mx1 and mx2 to dovecot 2.3.1_2 > from 2.3.0 ? (service dovecot stop, portmaster upgrade, service > dovecot start) > 2. ??? both systems ran ok with no errors for 10 days. > 3. ??? Last night I shutdown mx2 and restarted it a few hours later > 4. ??? within minutes i was getting the following types of errors on mx2 > > ???????? May 06 12:56:29 doveadm: Error: Couldn't lock > /var/mail/vhosts/example.net/user1/.dovecot-sync.lock: > fcntl(/var/mail/vhosts/example.net/user1/.dovecot-sync.lock, write-lock, > F_SETLKW) locking failed: Timed out after 30 seconds (WRITE lock held by > pid 1960) > > ? Before i venture down the rabbit hole of fault finding and excess > coffee consumption I was wondering if any of you had any updates on the > problems discussed below. > > > Cheers for now, > Andy > > > > Hi, > > [Formatting is a bit rough, replying from a trimmed digest email] > >> /Message: 1 />/Date: Fri, 6 Apr 2018 15:04:35 +0200 />/From: Michael Grimm <trashcan at ellael.org > <https://dovecot.org/mailman/listinfo/dovecot>> />/To: Dovecot Mailing List <dovecot at dovecot.org > <https://dovecot.org/mailman/listinfo/dovecot>> />/Subject: Re: 2.3.1 Replication is throwing scary errors />/Message-ID: <E7E7A927-68F8-443F-BA59-E66CED8FE878 at ellael.org > <https://dovecot.org/mailman/listinfo/dovecot>> />/Content-Type: text/plain; charset=utf-8 />//>/Reuben Farrelly wrote: />>/From: Michael Grimm <trashcan at ellael.org > <https://dovecot.org/mailman/listinfo/dovecot>> />//>>>/[This is Dovecot 2.3.1 at FreeBSD STABLE-11.1 running in two jails at > distinct servers.] />>>/I did upgrade from 2.2.35 to 2.3.1 today, and I do become pounded by > error messages at server1 (and vice versa at server2) as follows: />>>/| Apr 2 17:12:18 <mail.err> server1.lan dovecot: doveadm: Error: > dsync(server2.lan): I/O has stalled, \ />>>/no activity for 600 seconds (last sent=mail_change, last > recv=mail_change (EOL)) />>>/| Apr 2 17:12:18 <mail.err> server1.lan dovecot: doveadm: Error: Timeout > during state=sync_mails \ />>>/(send=changes recv=mail_requests) />>>/[?] />>>/| Apr 2 18:59:03 <mail.err> server1.lan dovecot: doveadm: Error: > dsync(server2.lan): I/O has stalled, \ />>>/no activity for 600 seconds (last sent=mail, last recv=mail (EOL)) />>>/| Apr 2 18:59:03 <mail.err> server1.lan dovecot: doveadm: Error: Timeout > during state=sync_mails \ />>>/(send=mails recv=recv_last_common) />>>/I cannot see in my personal account any missing replications, *but* I > haven't tested this thoroughly enough. I do have customers being > serviced at these productive servers, *thus* I'm back to 2.2.35 until I > do understand or have learned what is going on. />//>/In my reply to this statement of mine I mentioned that I have seen those > timeouts quite some times during the past year. Thus, I upgraded to > 2.3.1 again, and boom: after some hours I ended up in hanging processes > [1] like (see Remko's mail in addition) ... />//>/doveadm-server: [IP4/6 <USER1> SOME/MAILBOX import:0/0] (doveadm-server) />//>/? at server2 paired with a file like ? />//>/-rw------- 1 vmail dovecot uarch 0 Apr 3 16:52 > /home/to/USER1/.dovecot-sync.lock />//>/Corresponding logfile entries at server2 are like ? />//>/Apr 3 17:10:49 <mail.err> server2.lan dovecot: doveadm: Error: Couldn't > lock /home/to/USER1/.dovecot-sync.lock: \ />/fcntl(/home/to/USER1/.dovecot-sync.lock, write-lock, F_SETLKW) locking > failed: Timed out after 30 seconds \ />/(WRITE lock held by pid 51110) />//>/[1] Even stopping dovecot will not end those processes. One has to > manually kill those before restarting dovecot. />//>/After one day of testing 2.3.1 with a couple of those episodes of > locking/timeout, and now missing mails depending with server your MUA > will connect to, I went back to 2.2.35. After two days at that version I > never had such an episode again. />//>>/It's not just you. This issue hit me recently, and it was impacting />>/replication noticeably. I am following git master-2.3 . />/[...] />>/There is also a second issue of a long standing race with replication />>/occurring somewhere whereby if a mail comes in, is written to disk, is />>/replicated and then deleted in short succession, it will reappear />>/again to the MUA. I suspect the mail is being replicated back from />>/the remote. A few people have reported it over the years but it's not />>/reliable or consistent, so it has never been fixed. />>/And lastly there has been an ongoing but seemingly minor issue />>/relating to locking timing out after 30s particularly on the remote />>/host that is being replicated to. I rarely see the problem on my />>/local disk where almost all of the mail comes in, it's almost always />>/occurring on the replicate/remote system. />//>/It might be time to describe our setups in order to possibly find co > mmon > grounds that might trigger this issue you describe and Rimko and myself > ran into as well. />//>/Servers: Cloud Instances (both identical), around 25ms latency apart. />/Intel Core Processor (Haswell, no TSX) (3092.91-MHz K8-class CPU) />/Both servers are connected via IPsec/racoon tunnels />/OS: FreeBSD 11.1-STABLE (both servers) />/Filesystem: ZFS />/MTA: postfix 3.4-20180401 (postfix delivers via dovecot's LMTP) />/IMAP: dovecot running in FreeBSD jails (issues with 2.3.1, fine with > 2.2.35) />/Replication: unsecured tcp / master-master />/MUA: mainly iOS or macOS mail.app, rarely roundcube / > For me: > > Servers: Main: VM on a VMWare ESXi local system (light > load), local SSD disks (no NFS) > Redundant: Linode VM in Singapore, around 250ms away > Also no NFS. Linode use SSDs for IO. > There is native IPv6 connectivity between both VMs. > As I am using TCPs I don't have a VPN between them - > just raw IPv6 end to end. > OS: Gentoo Linux x86_64 kept well up to date > Filesystem: EXT4 for both > MTA: Postfix 3.4-x (same as you) > IMAP: Dovecot running natively on the machine (no jail/chroot) > Replication: Secured TCP / master-master (but only one master is > actively used for SMTP and MUA access) > MUA: Thunderbird, Maildroid, MS Mail, Gmail client, > Roundcube, LG Mail client on handset etc > >> /I believe it is worthwhile to mention here that I run a poor man's > fail-over approach (round-robin DNS) as follows: />//>/DNS: mail.server.tld resolves to one IP4 and one IP6 address of each > server, thus 4 IP addresses in total / > For me I have a simple pair of servers that replicate mail to each > other. Clients connect to the main server only, which is where mail > comes in. The second one that experiences the problem with locking is > only a redundant server which is a secondary MX, that only gets > occasional SMTP deliveries from time to time. So the amount of > replication back from the "secondary" is very minimal; the vast bulk of > the replication is the primary replicating changes out. > > Reuben >

8 20

offtopic: Solr compatible IMAP client
by Andy Smith 04 Jun '18

04 Jun '18

Hi, not sure if I'm allowed to go this off topic on here, sorry if not! I've just installed Solr FTS for Dovecot and its great, as is Dovecot! ;) It works perfectly for fast searches using Roundcube, but can anyone suggest any Windows and/or Linux mail clients that work well with IMAP searches rather than doing their own searches on local copies of mails? I've had a look at Thunderbird and you can use IMAP search, buy you have to manually select it from an advanced search dialogue and it seems you cannot make this the default behaviour. I also tried OperaMail but it doesn't appear to work at all (only header is searchable without downloading local copies). I'm pretty sure Outlook will only use its own search engine on its local copy of data. thanks a lot, Andy. PS this guide was really helpful for setting up Solr FTS with a recent release of Solr (I'm using 6.5), might be worth including a link somewhere on the Dovecot wiki?: http://mor-pah.net/2016/08/15/dovecot-2-2-with-solr-6-or-5

4 7

Re: SSL error after upgrading to 2.31
by Aki Tuomi 01 Jun '18

01 Jun '18

You forgot to cc the list. ssl_ca is used only for validating client certificates. ---Aki TuomiDovecot oy -------- Original message --------From: Marc Perkel <marc(a)perkel.com> Date: 21/05/2018 18:25 (GMT+02:00) To: Aki Tuomi <aki.tuomi(a)dovecot.fi> Subject: Re: SSL error after upgrading to 2.31 On 05/21/2018 07:54 AM, Aki Tuomi wrote: Does ssl_cert file contain intermediates? No - but the ssl_ca does. --- Aki Tuomi Dovecot oy -------- Original message -------- From: Marc Perkel <marc(a)perkel.com> Date: 21/05/2018 16:32 (GMT+02:00) To: dovecot(a)dovecot.org Subject: SSL error after upgrading to 2.31 After upgrading to 2.31 I'm getting this error. Not sure what I'm doing wrong. No (No signatures could be verified because the chain contains only one certificate and it is not self signed.) ssl = yes ssl_cert = </etc/exim/certs/ctyme.com.crt ssl_key = </etc/exim/certs/ctyme.com.key ssl_ca = </etc/exim/certs/ca.crt local mail.ctyme.com { protocol imap { ssl_cert = </etc/exim/certs/ctyme.com.crt ssl_key = </etc/exim/certs/ctyme.com.key ssl_ca = </etc/exim/certs/ca.crt } protocol pop3 { ssl_cert = </etc/exim/certs/ctyme.com.crt ssl_key = </etc/exim/certs/ctyme.com.key ssl_ca = </etc/exim/certs/ca.crt } } local mail.junkemailfilter.com { protocol imap { ssl_cert = </etc/exim/certs/junkemailfilter.com.crt ssl_key = </etc/exim/certs/junkemailfilter.com.key ssl_ca = </etc/exim/certs/ca.crt } protocol pop3 { ssl_cert = </etc/exim/certs/junkemailfilter.com.crt ssl_key = </etc/exim/certs/junkemailfilter.com.key ssl_ca = </etc/exim/certs/ca.crt } }

3 9

Struggling with sieve, fileinto and variables.
by Barry Pearce 31 May '18

31 May '18

Hi, Im on Manjaro linux fully up to date running dovecot 2.3.1-2 pigeonhole 0.5.1-2 All is running well except I am having problems with fileinto - it refuses to use variables, and mailboxexists is also failing for me. Im just dealing with plus addressing - should be simple but the behaviour Im experiencing isnt right. require ["variables", "envelope", "fileinto", "subaddress", "mailbox"]; if envelope :matches :detail "to" "*" { set :lower "name" "${1}"; } fileinto :create "folder3"; This works, but changing the last line to: fileinto :create "${name}"; fails and files into inbox. As does: fileinto :create "${1}"; It makes no difference if the fileinto is placed inside the if statement or otherwise. Using sieve-test suggests the basic script it should work and traces show the right action. On a related matter the script Im actually trying to get to work is: require ["variables", "envelope", "fileinto", "subaddress", "mailbox"]; if envelope :matches :detail "to" "*" { set :lower "name" "${1}"; } if mailboxexists "${name}" { fileinto :create "${name}"; } But fails and files in the inbox. Modifying the script to: require ["variables", "envelope", "fileinto", "subaddress", "mailbox"]; if envelope :matches :detail "to" "*" { set :lower "name" "${1}"; } if mailboxexists "${name}" { fileinto :create "folder7"; } else { fileinto :create "folder8"; } Files into folder8. So the mailboxexists is failing. ## Started executing script 'plus-addressing' 3: envelope test 3: starting `:matches' match with `i;ascii-casemap' comparator: 3: getting `to' part from message envelope 3: extracting `detail' part from address <test+folder4(a)test.net> 3: matching value `folder4' 3: with key `*' => 1 3: finishing match with result: matched 3: jump if result is false 3: not jumping 4: set command 7: modify :lower "folder4" => "folder4" 7: assign `name' [0] = "folder4" 7: mailboxexists test 7: mailbox `folder4' cannot be opened 7: some mailboxes are unavailable 7: jump if result is false 7: jumping to line 9 10: fileinto action 10: store message in mailbox `folder8' ## Finished executing script 'plus-addressing' Here folder4 actually does exist - so the sieve-test confirms that issue. Im at a loss as to whats going on. Im particularly perplexed by the difference in behaviour between sieve-test and the result under the live server. Any ideas?

2 3

strange misflagging of incoming mail as trash
by Robb Aley Allan 31 May '18

31 May '18

Setting up postfix/dovecot on a clean FreeBSD cloud server, and have discovered a strange problem resulting from the interaction of Mac OS X Mail and dovecot (using the Maildir file structure). Essentially, under very specific circumstances, Mail somehow tells dovecot that a new incoming message is trash, dovecot marks it as such, and the message effectively vanishes. The specific circumstance is as follows: 1. A message is sent to a server running postfix. It is placed in the Maildir mail structure in the “new” directory. 2. OS X Mail is used to retrieve mail via IMAP, and dovecot transfers the message from “new” to “cur”. 3. OS X Mail has a rule that seeks to transfer an incoming message to a subfolder if the message sender belongs to a group created in the Contacts app (IOW the sender is matched against an OS-internal contacts database). 4. IF the sender is the user, AND the user is in the group against which the rule is compared, the message is marked as trash (suffix “T”) and is never loaded into the mail client, either in the inbox or in the trash folder. For all intents and purposes, it disappears. The connection log shows the following line: "WROTE May 30 23:38:11.960 [kCFStreamSocketSecurityLevelTLSv1_2] -- host:cloud.helical.com -- port:993 -- socket:0x6080004b6c80 -- thread:0x604000678e40 12.3 UID STORE 125 +FLAGS.SILENT (\Deleted)” which appears to be the client instructing dovecot to mark the message as deleted. I have the log files from the connection on both sides, and the console log from the Mac, but can’t understand 1. how Mail decides that the message should be deleted, and 2. why Mail “loses” the message (doesn’t appear in the Trash mailbox or anywhere else). _____ Robb Aley Allan

1 0

imapsieve: mailbox Trash: FLAG event
by Peter Schiffer 30 May '18

30 May '18

Hi all, after upgrade from Dovecot 2.2 to 2.3 I started to see that emails from Trash folder are automatically deleted after ~7 days. Email clients are not configured to automatically expunge or empty trash on log out (I see this behavior for multiple mailboxes, with different clients). With mail_debug enabled, I see this log message for every email that gets deleted from Trash: May 29 08:02:17 hostname dovecot: imap(some@mail)<7725><xxb6+VFtL+JT0CE 8>: Debug: imapsieve: mailbox Trash: FLAG event (changed flags: \Deleted) I have Dovecot 2.3.1 (8e2f634) and Pigeonhole 0.5.1 for sieve. My configuration can be found here: https://github.com/pschiffe/mailcow-dockerized/tree/master/data/conf/dovecot There is enabled support for sieve config in sql, but all sieve tables are empty, and there are no sieve filters for Trash flagging. Do you know how to find the source of that Trash flag event? I need to disable it or to configure it so it deletes emails at least couple of months old. Thanks all, peter

2 1

Re: Fatal: nfs flush requires mail_fsync=always
by Juan C. Blanco 30 May '18

30 May '18

Hello, any news about the attached error? I'm preparing the 2.2 to 2.3 upgrade and having the same error. We have the mail stores in an NFS filer. Regards > On 19.01.2018 11:55, Søren Skou wrote: >> Hiya all, >> >> I'm seeing this "Fatal: nfs flush requires mail_fsync=always" error on >> my testbed. The issue is that from what I can see, mail_fsync is set >> to always : >> >> # doveconf -n | grep mail_fs >> mail_fsync = always >> >> The result is that the client does not connect at all, which is not >> really what I wanted to happen :) >> >> Any idea what is going wrong here? >> >> Best regards >> Søren P. Skou >> >> doveconf -n >> >> # 2.3.1.alpha0 (bdfa22623) [XI:2:2.3.1~alpha0-1~auto+14]: >> /etc/dovecot/dovecot.conf >> # Pigeonhole version 0.5.1.alpha0 (d5f710e0) >> # OS: Linux 4.9.0-4-amd64 x86_64 Debian 9.3 nfs >> auth_worker_max_count = 200 >> dict { >> expire = mysql:/etc/dovecot/dovecot-dict-sql.conf.ext >> quota = mysql:/etc/dovecot/dovecot-dict-sql.conf.ext >> sqlquota = mysql:/etc/dovecot/dovecot-dict-sql.conf.ext >> } >> disable_plaintext_auth = no >> lock_method = dotlock >> mail_fsync = always >> mail_location = maildir:/mnt/virtual_mail/%d/%n >> mail_nfs_index = yes >> mail_nfs_storage = yes >> mail_plugins = quota >> mailbox_list_index = no >> metric imap_select_no { >> event_name = imap_command_finished >> filter { >> name = SELECT >> tagged_reply_state = NO >> } >> } >> mmap_disable = yes >> namespace inbox { >> inbox = yes >> location = >> mailbox Drafts { >> special_use = \Drafts >> } >> mailbox Junk { >> auto = subscribe >> 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-sql.conf.ext >> driver = sql >> } >> plugin { >> quota = dict:User quota::proxy::sqlquota >> quota_grace = 10%% >> quota_rule = *:storage=1000m:messages=300000 >> quota_rule2 = Trash:storage=+10%%:messages=+10%% >> quota_rule3 = Junk:storage=+20%%:messages=+20%% >> quota_status_nouser = DUNNO >> quota_status_overquota = 552 5.2.2 Mailbox is full >> quota_status_success = DUNNO >> quota_warning = storage=75%%:messages=75%% quota-warning 75 %u >> quota_warning2 = storage=95%%:messages=95%% quota-warning 95 %u >> quota_warning3 = -storage=100%%:messages=100%% quota-warning below %u >> sieve = /etc/dovecot/sieve/default.sieve >> sieve_global_dir = /etc/dovecot/sieve >> } >> protocols = " imap pop3" >> service dict { >> unix_listener dict { >> mode = 0600 >> user = vmail >> } >> } >> service imap { >> executable = imap >> } >> service quota-status { >> client_limit = 1000 >> executable = quota-status -p postfix >> inet_listener { >> address = 127.0.0.1 >> port = 12340 >> } >> } >> service quota-warning { >> executable = script /usr/local/bin/quota-warning.sh >> user = vmail >> } >> ssl_ca = /etc/ssl/certs/ca-root.crt >> ssl_cert = </etc/ssl/certs/stjerne.danskkabeltv.dk-2019.crt >> ssl_cipher_list = TLSv1+HIGH !SSLv2 !RC4 !aNULL !eNULL !3DES-CBC !3DES @STRENGTH >> ssl_dh = # hidden, use -P to show it >> ssl_key = # hidden, use -P to show it >> userdb { >> args = uid=2000 gid=2000 home=/mnt/virtual_mail/%d/%n >> driver = static >> } >> protocol lmtp { >> mail_plugins = quota >> } >> protocol lda { >> mail_plugins = quota >> } >> protocol imap { >> mail_plugins = quota imap_quota >> rawlog_dir = /tmp/rawlog/%u >> } > > Hi! > > Thanks, we'll look into it. > > Aki -- +-------------------------------------------------------------------+ | Juan C. Blanco | | | | Centro de Calculo | | | E.T.S. Ingenieros Informáticos | E-mail: jcblanco(a)fi.upm.es | | Universidad Politécnica de Madrid | | | Campus de Montegancedo | | | Boadilla del Monte | Tel.: (+34) 91 067 2771 | | 28660 MADRID (Spain) | Fax : (+34) 91 336 7412 | +-------------------------------------------------------------------+

2 2

Second rule isn't apply when first rule matches
by daniel_1983＠protonmail.com 30 May '18

30 May '18

Dear list, I want to define two concurrent rules : 1. I want to flag an e-mail containing the word "cloud" in the body 2. I want to move mail sent explicitly to me (as opposed to mail sent to an alias I am part of) going to "INOBX.I must answere this" If I put rule (1) first, everything works as expected. If put rule (2) first, only that rule applies. Here's a small test case you should be able to reproduce on your setup. Let's work on this fake test e-mail : $ cat testmail.eml Date: Mon, 28 May 2018 12:41:53 +0100 From: friend(a)otherdomain.tld To: me(a)mydomain.tld Subject: dur dur la vie mon frère, le cloud c'est le top du top Here's the first, broken version of the script $ cat test.sieve require ["body","fileinto","imap4flags","vacation"]; # rule:[Mail about me] if anyof (header :contains "to" "me(a)mydomain.tld", body :text :contains "ahmed") { fileinto "INBOX.I must answere this"; } # rule:[Mail about the Cloud] if body :text :contains "cloud" { addflag "\\Flagged"; } Let's test it out, the two rules should be applied : $ sieve-test test.sieve testmail.eml Performed actions: * store message in folder: INBOX.I must answere Implicit keep: (none) sieve-test(root): Info: final result: success $ Notice that the last rule isn't applied although it matches. Now we invert the order of the rules and apply the script on the same test e-mail : $ cat test.sieve require ["body","fileinto","imap4flags","vacation"]; # rule:[Mail about the Cloud] if body :text :contains "cloud" { addflag "\\Flagged"; } # rule:[Mail about me] if anyof (header :contains "to" "me(a)mydomain.tld", body :text :contains "ahmed") { fileinto "INBOX.I must answere this"; } Running the test again : $ sieve-test test.sieve testmail.eml Performed actions: * store message in folder: INBOX.I must answere + add IMAP flags: \flagged Implicit keep: (none) sieve-test(root): Info: final result: success $ The IMAP flag was set ! Any ideas ?

2 3

Problem in Pigeonhole sievec
by Thorsten Hater 30 May '18

30 May '18

Dear all, I stumbled upon the following behaviour of Pigeonhole, which I consider to be problematic. A user deployed a Sieve script similar to the following snippet if not anyof (address :is ["from","cc"] ["...", ..., "...@... GARBAGE", ...] { fileinto "inbox.Trash"; stop; } Note the extra line break before GARBAGE. This script is obviously broken, but gets accepted by sievec and only fails later, at runtime with line X: error: found stray carriage-return (CR) character in quoted string started at line X. So, the question is whether line breaks in strings are allowed in general and the runtime error is unavoidable, or should sievec return an error? Best regards, Thorsten

3 3