March 2007 - dovecot - dovecot.org

[Dovecot] convert plugin
by FiL ＠ Kpoxa 07 Mar '07

07 Mar '07

Was about to try convert plugin (as opposed to mb2md.pl) and noticed, that it doesn't convert subfolders. I have following structure: ~/ | mail/ | Archive | Help | Sent | Done/ | Cardiology | Childrens | Orkin when I use convert-tool - #convert-tool username /home/username mbox:/home/username/mail/ maildir:/home/username/Maildir/ Error: Mailbox conversion: Couldn't create mailbox Done/Cardiology Error: Internal failure If I try "on the fly" conversion (by putting convert_mail = mbox:%h/mail & convert_skip_broken_mailboxes=yes in dovecot.conf) I get this in the log - Mar 6 12:23:33 enders2 dovecot: imap-login: Login: user=<username>, method=PLAIN, rip=::ffff:127.0.0.1, lip=::ffff:127.0.0.1, secured Mar 6 12:23:43 enders2 dovecot: IMAP(username): Mailbox conversion: Couldn't create mailbox Done/Cardiology Both methods convert Archive and Logs mailboxes fine. Can anybody confirm? P.S. Linux enders2.tchlab.org 2.6.18-1.2200.fc5smp #1 SMP Sat Oct 14 17:15:35 EDT 2006 i686 i686 i386 GNU/Linux dovecot-1.0.rc22 (I know this is not the latest version, but I didn't see any changes to convert tool/plugin). FiL

2 4

[Dovecot] Config questionsabout various login parms
by Stewart Dean 07 Mar '07

07 Mar '07

I'm currently runing UW-IMAP (compiled vanilla) with procmail and mbox and am making my usual careful crawl (at turtle speed) through figuring out DC config. Some questions: a) I have something like 3-4 root imapd sessions so I presume that the default *login_processes_count* = 3 will be OK. I'm less clear about login_max_processes_count and login_max_connections = 256. A common enough heavy load has something like 275 unique users accessing imap and 600 imapd processes. Suggestions? I'm beginning with no SSL, KISS. b) If I set login_process_per_connection =no, would the number of users accessing imap be the same as the number of imapd sessions? c) login_process_size is confusing/intriguing. If there are 500 processes, the default of 32MB means 16GB (the box has 12GB). OTOH, what about the big players with a 400 MB inbox...will they be able to open their inbox with the default d) login_max_connections ditto...I shouldn't think that an individual user would conceivably have the default of 256, except pathologically....or is this applied to the listening imapd root processes. Thanks in advance for your help and patience......... -- ==== Stewart Dean, Unix System Admin, Henderson Computer Resources Center of Bard College, Annandale-on-Hudson, New York 12504 sdean(a)bard.edu voice: 845-758-7475, fax: 845-758-7035

3 5

[Dovecot] Nokia 6680 and Dovecot
by Boris Ivanov 07 Mar '07

07 Mar '07

Hi Guys Im using Nokia 6680 to connect to dovecot (dovecot-1.0.b7_1 from FreeBSD ports) Where ever it trying to login it failed despite it works well with PC clients. * OK [CAPABILITY IMAP4rev1 SORT THREAD=REFERENCES MULTIAPPEND UNSELECT LITERAL+ IDLE CHILDREN NAMESPACE LOGIN-REFERRALS AUTH=PLAIN AUTH=LOGIN] Dovecot ready. 1 CAPABILITY * CAPABILITY IMAP4rev1 SORT THREAD=REFERENCES MULTIAPPEND UNSELECT LITERAL+ IDLE CHILDREN NAMESPACE LOGIN-REFERRALS AUTH=PLAIN AUTH=LOGIN 1 OK Capability completed. 2 LOGIN sj sj 2 NO Authentication failed. I relized that problem is 1 character message tag - 1,2. Is it bug in phone or in dovecot? -- Yours Boris

2 2

[Dovecot] Newbie questions: Load-balanced Dovecot with NFS storage
by Mike Brudenell 07 Mar '07

07 Mar '07

Hi, I experimented with Dovecot a while back and our site is now taking the plunge to switch to it from the UW IMAP server when we rejig the topology of our mail store and IMAP servers. Currently we have a farm of 5 IMAP servers running the UW server. Each has locally attached disk used to store people's mailboxes. Each user is allocated to one of these five servers and can only access their mail via that server. Our aim is to remove this restriction to provide resilience in the event of a server failing. Our thoughts... * We are envisaging having a farm of N (to be determined) servers running Dovecot. These will be closed systems: users do not have shell- access, and can only reach their mail via IMAP. * A single mailstore common to all N servers will be hosted on our NetApp filer and NFS-mounted on each server. Each server would be running Solaris 10. If we use all-new kit it will be x86-based, or we may have to redeploy some of our existing SPARC boxes to the new service. * We will be converting user mailboxes to be in Maildir format. * Our hardware load balancer, which supports persistent sessions, will be used to distribute users between the servers. This could either direct a user to any one of the N servers, or could be set to 'prefer' a given server for a particular user (only failing over to one of the others if their 'usual' server was unavailable). More on this in my questions below. :-) I have been scouring the Dovecot list, archives and Wiki and think I have come up with a set of issues we need to be aware of and things we have to do in order to make this work reliably. Please would you let me know if I have missed or misunderstood anything? Issues To Watch For ------------------- * For timestamp integrity our servers will be synched with NTP to our local time server. * Because sessions could be on different servers memory mapping of index files doesn't work well with NFS, so set mmap_disable=yes in dovecot.conf. * Dovecot relies on the mtime timestamp of mailbox files so the NFS-mounted mailstore needs to be mounted with these options on each server: actimeo=0 (Are there any other mount options we should use too?) * We will be using filesystem quotas on users mailboxes. We understand that Dovecot's index files are best on no-quota filestores so will store these separately. Questions --------- Q1. Would it be better to store the index files on NFS-shared filestore and direct users to any of the IMAP server machines? Or to store the index files on local disk and direct each user to their 'preferred' IMAP server machine? Q2. Does Dovecot (or "something") clean out old index files that haven't been accessed for a while? Eg, when a user has temporarily come through on a different IMAP server to normal. Or do the index files sit there untouched for evermore? Q3. Storing the index files on the NetApp filer would give us the ability to share them between servers and grow their volume as need be, but at the cost of performance. How big do the index files get? Are they typically a few kBytes per message? Per mailbox? Per user? Or more? Q4. We will be using Exim as the MTA, which can deliver direct to Maildir mailboxes. However I understand that Dovecot's "deliver" LDA adds the benefit of updating the index files as each message is delivered. Is this a significant gain? Or is there little difference in actual use? Q5. We have around 20,000 mail accounts and will therefore be seeing lots of concurrent IMAP sessions, usually secure (SSL) ones. I have seen mention that this can give rise to "Too many open files" errors under Solaris. How do we avoid this when we are likely to have several thousand concurrent IMAP sessions per server machine? With many thanks for your time and advice... Cheers, Mike Brudenell -- The Computing Service, University of York, Heslington, York Yo10 5DD, UK Tel:+44-1904-433811 FAX:+44-1904-433740 * Unsolicited commercial e-mail is NOT welcome at this e-mail address. *

5 9

[Dovecot] Time moving backwards error
by Emmett Pate 07 Mar '07

07 Mar '07

I'm running dovecot on an OpenBSD 3.7 server and after upgrading to RC25, the daemon started exiting with: Fatal: Time just moved backwards (1173221579 -> 1173221578)! This might cause a lot of problems, so I'll just kill myself now. After upgrading to RC26, I've gotten the error: Error: Time just moved backwards by 1 seconds. I'll sleep now until we're back in present. Occasionally, however, it's followed by: Fatal: Sleep interrupted, byebye. The server is running NTP, but I can't find any reason why it would have to roll the clock back a second so often. Until I get this figured out, I've gone back to RC19. Any ideas? Thanks. -- Emmett "Buddy" Pate Chief Technology Officer William E. Wood and Associates, Realtors epate(a)williamewood.com

3 2

[Dovecot] dovecot-ldap.conf - Active directory
by Cristian Salan 07 Mar '07

07 Mar '07

Hello, I have a configuration with dovecot-ldap working with openldap (on FreebSD 5.4 Release). Now I'm trying to do the same thing with an Active Directory (win2k3). I have installed an properly configured SFU. Via nss_ldap-pam_ldap is working fine (FreeBSD 6.2 Release). Because nss_ldap is reported broken with dovecot I'm trying to use dovecot-ldap for both passdb and userdb. My guessed filters/attributes follows: user_filter = (&(objectClass=user)(msSFU30Name=%u)) user_attrs = homeDirectory=msSFU30HomeDirectory, uidNumber=msSFU30UidNumber, gidNumber=msSFU30GidNumber pass_attrs = msSFU30Name,msSFU30Password pass_filter = (&(objectClass=user)(msSFU30Name=%u)) But it doesn't work. Any advice? Cristian Salan

2 1

[Dovecot] sporadic problem using ldap authentication
by Nuno Lopes 07 Mar '07

07 Mar '07

Hi, recently some users started complaining about sporadic authentication failures. When examining the logs I saw the following: Mar 7 03:20:24 gollum dovecot: auth(default): file auth-request.c: line 472 (auth_request_lookup_credentials_callback): assertion failed: (request->state == AUTH_REQUEST_STATE_PASSDB) Mar 7 03:20:24 gollum dovecot: auth(default): Raw backtrace: dovecot-auth [0x41d7fc] -> dovecot-auth [0x41da6c] -> dovecot-auth [0x40ddf3] -> dovecot-auth [0x416475] -> dovecot-auth [0x416825] -> dovecot-auth [0x41176d] -> dovecot-auth(io_loop_handler_run+0x13c) [0x420cec] -> dovecot-auth(io_loop_run+0x1c) [0x42005c] -> dovecot-auth(main+0x41e) [0x412ade] -> /lib64/tls/libc.so.6(__libc_start_main+0xea) [0x2a959fbaaa] -> dovecot-auth(ldap_msgfree+0x42) [0x4098da] Mar 7 03:20:24 gollum dovecot: child 10883 (auth) killed with signal 6 Mar 7 03:20:24 gollum dovecot: pop3-login: Can't connect to auth server at default: Resource temporarily unavailable Mar 7 03:20:24 gollum dovecot: auth(default): file auth-request.c: line 472 (auth_request_lookup_credentials_callback): assertion failed: (request->state == AUTH_REQUEST_STATE_PASSDB) Mar 7 03:20:24 gollum dovecot: auth(default): Raw backtrace: dovecot-auth [0x41d7fc] -> dovecot-auth [0x41da6c] -> dovecot-auth [0x40ddf3] -> dovecot-auth [0x416475] -> dovecot-auth [0x416825] -> dovecot-auth [0x41176d] -> dovecot-auth(io_loop_handler_run+0x13c) [0x420cec] -> dovecot-auth(io_loop_run+0x1c) [0x42005c] -> dovecot-auth(main+0x41e) [0x412ade] -> /lib64/tls/libc.so.6(__libc_start_main+0xea) [0x2a959fbaaa] -> dovecot-auth(ldap_msgfree+0x42) [0x4098da] Mar 7 03:20:24 gollum dovecot: child 10945 (auth) killed with signal 6 This output is from dovecot-1.0rc26 but this was first seen in 1.0rc7 and 1.0rc24 has similar results. I have seen no problem with the ldap server. I'm using similar setups in several other servers with no trouble at all. The main difference I see is that this machine is 64-bit and on a 2.6 linux kernel versus x86 on a 2.4. Thanks for any help you can provide. -- Nuno Lopes

2 1

[Dovecot] Question about auth-master, deliver, and SQL maps
by Richard Coleman 07 Mar '07

07 Mar '07

I have been running Postfix/Dovecot with a Postgresql backend for several months, and everything has been running great. Now I want to add server side filtering, so I'm attempting to configure deliver. The SQL maps for Postfix are all keyed upon the recipient email address. The SQL map (user_query) for Dovecot is keyed upon the IMAP login name, which is potentially unrelated to the email address. But when I configure deliver to use auth-master, it doesn't work right, because it is passing the recipient email address as a key to user_query, which is wrong. Somehow, I need auth-master to use a different SQL query from user_query when doing lookups for deliver. In essence, I want deliver to use the same SQL queries where were previously used for the Postfix "virtual" delivery agent. Is there any way to do this? Richard Coleman rcoleman(a)criticalmagic.com

2 1

[Dovecot] 1.0.rc25 released
by Timo Sirainen 07 Mar '07

07 Mar '07

http://dovecot.org/releases/dovecot-1.0.rc25.tar.gz http://dovecot.org/releases/dovecot-1.0.rc25.tar.gz.sig Instead of having "Should v1.0 be released already" discussion, how about having "What's still missing from wiki.dovecot.org and how could it be improved" discussion? And what should the wiki exported to doc/ directory in the tarball look like? * If time moves backwards, Dovecot kills itself instead of giving random problems. + Added --with-headers configure option to install .h files. Binary package builders could use this to create some dovecot-dev package to make compiling plugins easier. - PLAIN authentication: Don't crash dovecot-auth with invalid input. - IMAP APPEND: Don't crash if saving fails - IMAP LIST: If prefix.INBOX has children and we're listing under prefix.%, don't drop the prefix. - mbox: Broken X-UID headers still weren't handled correctly. - mail-log plugin: Fixed deleted/undeleted logging.

16 29

[Dovecot] maildirsize corruption
by Dean Manners 07 Mar '07

07 Mar '07

Using maildir++ with LDA and POP/IMAP we have been seeing intermittant maildirsize corruption. Where an operation will be prepended by a handful of ASCI null characters ^@^@. Eg: # cat maildirsize 5120000S -46895 -1 93359 1 60576 1 2858 1 49547 1 -206340 -4 94603 1 84785 1 54463 1 82486 1 97104 1 ^@^@^@^@^@^@^@^@40488 1 73397 1 83662 1 24668 1 42673 1 12081 1 -713606 -12 The instances of this that we have seen have all been deliver operations. Anyone else seen this ? Regards __________________________________________ Dean Manners

3 5