- dovecot - dovecot.org

Huge difference between the lucene index size created by v2.1 and v2.2
by Akash 20 Aug '14

20 Aug '14

Hi everyone, While examining dovecot versions v2.1 and v2.2 for their lucene search performances, I have noticed a huge difference in the index sizes created by them. Both versions were compiled on same system, against same libclucene, same configure options and were used with same dovecot.conf during run-time. I deleted the lucene-indexes folder and dovecot* files in the Maildir prior to indexing with both versions. The tests were performed on a untouched mail archive folder containing 300000 mails without any dovecot* files in it: root@server:/home/admin/mails/.Archive# ls -l total 78640 drwx------ 2 2500 2500 14098432 Aug 17 22:16 cur drwx------ 2 2500 2500 12435456 Jul 30 09:46 new drwx------ 2 2500 2500 4096 Aug 2 13:02 tmp The command used was: doveadm -v index -u admin Archive After search indexing on v2.1 resulted in: root@server:/home/admin/mails# ls -lh lucene-indexes total 390M -rw------- 1 2500 2500 390M Aug 18 07:03 _25.cfs -rw------- 1 2500 2500 20 Aug 18 07:03 segments.gen -rw------- 1 2500 2500 46 Aug 18 07:03 segments_4d Whereas dovecot v2.2 resulted in: root@server:/home/admin/mails# ls -lh lucene-indexes total 1.5G -rw------- 1 2500 2500 1.5G Aug 18 06:41 _5g.cfs -rw------- 1 2500 2500 20 Aug 18 06:41 segments.gen -rw------- 1 2500 2500 46 Aug 18 06:41 segments_az 390M vs 1.5G. That is a huge difference in size. Why is that? Thanks in advance. -Regards, Akash

2 5

Problems with dovecot 2.2.13 and monit
by Hanno Böck 20 Aug '14

20 Aug '14

Hello, When I upgraded my servers to dovecot 2.2.13 the monitoring tool monit started to send out warnings that it couldn't reach my imap/pop3 servers through ssl any more. The same problem didn't happen on non-ssl-connections. According to people on the monit list this is likely a dovecot issue: https://lists.gnu.org/archive/html/monit-general/2014-06/msg00031.html Let me quote: > the root cause of the error is, that dovecot 2.2.13 closes the > connection if SSL is used in response to LOGOUT command instead of > sending usual response. When no SSL is enabled, dovecot responses to > LOGOUT command normally. [...] > According to RFC 3501 (http://tools.ietf.org/html/rfc3501) LOGOUT is > any-state command, where the server MUST send response before closing > the connection: http://tools.ietf.org/html/rfc3501#section-3.4 > > => the problem is caused by dovecot 2.2.13 bug ... its behaviour is > inconsistent (LOGOUT in non-authenticated state works per RFC > requirement if no SSL is used and doesn't conform to RFC if SSL is > used). It is possible that the problem is related to their DoS-attack > modification, which has most probably unexpected side-effect. Maybe this is related to the DDoS-protection measures that have been added in dovecot 2.2.13. Would apprechiate if someone could have a look. cu, -- Hanno Böck http://hboeck.de/ mail/jabber: hanno(a)hboeck.de GPG: BBB51E42

3 3

Designing a Dovecot Replicate cluster
by Minhas Listas 19 Aug '14

19 Aug '14

Hi all! Do I need one Dovecot Proxy/Director for each Dovecot Replicate Cluster ? Isn't possible to put a Dovecot Proxy/Director on the border of multiple clusters ? Thanks! -- Thiago Henrique www.adminlinux.com.br

1 0

Health monitoring of backend servers
by Minhas Listas 19 Aug '14

19 Aug '14

Does Dovecot Director health monitoring of backend servers ? Or is poolmon (http://www.dovecot.org/list/dovecot/2010-August/051946.html) the best option for this? Maybe Ldirectord? Thanks! -- Thiago Henrique www.adminlinux.com.br

2 4

[SOLVED] Re: Trying Dovecot Replication with dsync
by Minhas Listas 19 Aug '14

19 Aug '14

> > 1) > Aug 14 15:39:34 ceph1 dovecot: imap(account=domaintest.com): Error: > open(/var/run/dovecot/replication-notify-fifo) failed: Permission denied > > What is the correct perms for /var/run/dovecot/replication-notify-fifo ? > Currently: > # ll /var/run/dovecot/replication-notify-fifo > prw------- 1 root root 0 Aug 14 15:57 > /var/run/dovecot/replication-notify-fifo| > > Solved with: # chown dovemail:dovemail /var/run/dovecot/replication-notify-fifo dovemail is the user for imap processes. > 2) > Aug 14 16:02:34 ceph1 dovecot: dsync-server(account=domaintest.com): > Error: Couldn't create lock */var/lib/imap/account=domaintest.com > <http://domaintest.com>/*.dovecot-sync.lock: No such file or directory > Aug 14 16:02:34 ceph1 dovecot: dsync-local(account=domaintest.com): > Error: Couldn't create lock */var/lib/imap/account=domaintest.com > <http://domaintest.com>/*.dovecot-sync.lock: No such file or directory > > In my production cluster, it is automatically created by dovecot when the > user performs the first login: > */var/lib/imap/*<account> for INDEX > */var/spool/imap/*<account> for Mdbox > This was fixed with correct permission on /var/run/dovecot/replication-notify-fifo Thanks! -- Thiago Henrique www.adminlinux.com.br

2 1

IMAP on 993/SSL or 143/STARTTLS?
by Alex 18 Aug '14

18 Aug '14

Hi, I have a postfix+dovecot-2.2.13 system and have configured it to support IMAPS on 993 with SSL/TLS. I'm noticing with users using Thunderbird, the autodetect defaults to IMAPS on 143 with STARTTLS. Which is preferred? Which is more secure? Which is more common? Why would someone choose one over the other? Can I ask the same question about SMTP and submission? Why would one choose 587 with STARTTLS versus 465 with SSL/TLS? Thanks, Alex

4 3

Segmentation fault while indexing a large mailbox using doveadm..
by Akash 18 Aug '14

18 Aug '14

Hi, I am having the same issue as: http://www.dovecot.org/list/dovecot/2014-April/095681.html although I am using the latest dovecot-core and dovecot-lucene packages from the debian sid repo which already has the patch (http://hg.dovecot.org/dovecot-2.2/rev/5e42a8c18b51) mentioned in the above thread implemented. doveadm(admin): Info: Old.All: Caching mails seq=1..300332 9400/300332 Program received signal SIGSEGV, Segmentation fault. (gdb) bt #0 __wcslen ( s=s@entry=0xb93887a8 L" btnToggle: \"Change size\"}</script>\r\n<link href=\"http://s26 ucoz net/src/ulightbox/ulightbox css\" rel=\"stylesheet\" type=\"text/css\" /><script type=\"text/javascript\" src=\"http://s26 ucoz net/src/ulightb"...) at wcslen.c:36 #1 0xb7c026fd in wcsdup ( s=0xb93887a8 L" btnToggle: \"Change size\"}</script>\r\n<link href=\"http://s26 ucoz net/src/ulightbox/ulightbox css\" rel=\"stylesheet\" type=\"text/css\" /><script type=\"text/javascript\" src=\"http://s26 ucoz net/src/ulightb"...) at wcsdup.c:29 #2 0xb7a65660 in lucene::document::Field::Field(wchar_t const*, wchar_t const*, int, bool) () from /usr/lib/i386-linux-gnu/libclucene-core.so.1 #3 0xb7b5cd48 in lucene_index_build_more () from /usr/lib/dovecot/modules/lib21_fts_lucene_plugin.so #4 0xb7b5b7b9 in ?? () from /usr/lib/dovecot/modules/lib21_fts_lucene_plugin.so #5 0xb7b6ac2a in fts_backend_update_build_more () from /usr/lib/dovecot/modules/lib20_fts_plugin.so #6 0xb7b6b674 in ?? () from /usr/lib/dovecot/modules/lib20_fts_plugin.so #7 0xb7b6bbf4 in fts_build_mail () from /usr/lib/dovecot/modules/lib20_fts_plugin.so #8 0xb7b713cc in ?? () from /usr/lib/dovecot/modules/lib20_fts_plugin.so #9 0xb7e4969b in mail_precache () from /usr/lib/dovecot/libdovecot-storage.so.0 #10 0xb7fa1c7f in ?? () #11 0xb7f9d133 in _start () (gdb) # dovecot --version 2.2.13 Please help.. -Thanks, Akash

2 2

Dovecot failing with exit code 19200 in cron job
by Murray Trainer 18 Aug '14

18 Aug '14

Hi All, I am synchronizing a few hundred dovecot mailboxes with a perl script via a crontab. It works for a while (ie exit status=0) but the rest then start failing with exit code 19200. I tried redirecting STDERR but showed nothing. I suspect the cron job is hitting some sort of resource limit? If I run the perl script from the command line it appears to dsync all the accounts OK. The relevant part of my Perl code is below. Any ideas? Thanks, Murray foreach (@userlist) { my $email = $_; chomp ($email); # get rid of CR logit("Started sync of mailbox $email") if $debug; my $dsync_command="dsync -v -D -u $email mirror ssh -C root@$mailstore dsync -u $email"; logit("dsync_command=$dsync_command"); # Run dsync of mailbox logit("Started secondary sync of $mailbox to $mailstore") if $debug; my $exit_status = system ( $dsync_command ); logit("Dsync exit status: $exit_status") if $debug; logit("Finished secondary sync of $mailbox to $mailstore") if $debug; }

2 1

NFS INTR mount option
by Murray Trainer 18 Aug '14

18 Aug '14

Hi again, We have our mailboxes on several NFS filesystems mounted with the options below. 10.11.0.238:/mailbox_store_01 on /home1 type nfs4 (rw,noatime,vers=4.0,rsize=65536,wsize=65536,namlen=255,hard,nordirplus,proto=tcp,port=0,timeo=600,retrans=2,sec=sys,clientaddr=10.11.0.95,local_lock=none,addr=10.110.238) On the rare occasions that the NFS mounts hang, the system load goes through the roof and I often have to reboot the mailstore server as dovecot process are hung and can't be killed. Will the NFS INTR mount option help in this situation? Thanks Murray Send EMAIL [1] Links: ------ [1] https://webmail.westnet.com.au/index.php/mail#

1 0

Manage Sieve Security?
by Chris 18 Aug '14

18 Aug '14

Dear All, can I expose Dovecot's Manage Sieve port to the internet or should I restrict access (e.g. by IP address or VPN)? -- Christian

2 1