- dovecot - dovecot.org

[Dovecot] 1.0-test49 FD leak
by Gregory Bond 18 Oct '04

18 Oct '04

I'm running 1.0-test49 on Solaris 2.8. The good news is that this version fixes my problem with corrupted index files (so far). The bad news is it seems to be leaking FDs to do with inbox: imap 10198 gnb 156r VDIR 255,7032 8192 5784160 /home/gnb/Maildir/new imap 10198 gnb 157r VDIR 255,7032 1093632 2917923 /home/gnb/Maildir/cur imap 10198 gnb 158r VDIR 255,7032 8192 5784160 /home/gnb/Maildir/new imap 10198 gnb 159r VDIR 255,7032 1093632 2917923 /home/gnb/Maildir/cur imap 10198 gnb 160r VDIR 255,7032 8192 5784160 /home/gnb/Maildir/new imap 10198 gnb 161r VDIR 255,7032 1093632 2917923 /home/gnb/Maildir/cur imap 10198 gnb 162r VDIR 255,7032 8192 5784160 /home/gnb/Maildir/new imap 10198 gnb 163r VDIR 255,7032 1093632 2917923 /home/gnb/Maildir/cur imap 10198 gnb 164r VDIR 255,7032 8192 5784160 /home/gnb/Maildir/new imap 10198 gnb 165r VDIR 255,7032 1093632 2917923 /home/gnb/Maildir/cur (seems to be a new pair of FDs leaked each time my MUA checks for new mail).

2 2

[Dovecot] 49-latest compile problem and CVS method
by Kim Culhan 18 Oct '04

18 Oct '04

--- Timo Sirainen <tss(a)iki.fi> wrote: > On 17.10.2004, at 21:02, Kim Culhan wrote: > > > The connection is being made with ssl so it looks like > > the plaintext auth preventer is broken. > > Fixed in CVS now (and in next nightly snapshot). Very good, this compile problem though on test49-latest from nightly: ar cru libpassword.a mycrypt.o password-scheme.o password-scheme-md5crypt.o password-scheme-cram-md5.o password-scheme-ntlm.o password-scheme-rpa.o ranlib libpassword.a make: don't know how to make auth-master-interface.h. Stop Also, tried getting the cvs tree with: cvs -d :pserver:anonymous@dovecot.org:/home/cvs co dovecot It doesn't populate the directory with all the files; does not include the build files such as Makefile.in, configure etc Any way to get cvs to do that ? -kim _______________________________ Do you Yahoo!? Express yourself with Y! Messenger! Free. Download now. http://messenger.yahoo.com

1 0

[Dovecot] Index corruption
by Gregory Bond 18 Oct '04

18 Oct '04

I'm getting pretty constant index corruption on (admittedly large) Maildir folders. I'm running Thunderbird 0.7.3 and Mutt as my IMAP clients (on FreeBSD). Server is Dovecot 0.99.10.9 configured thusly: ./configure --disable-ipv6 CC=gcc-3.3 CXX=g++-3.3 --prefix=/opt/dovecot --without-pop3d on a Solaris 2.8 box. Maildirs are NFS-mounted, but the index files are on local UFS filesystem: default_mail_env = maildir:~/Maildir:INDEX=/export/01/imapd/%u MTA is Postfix, LDA is procmail delivering to several Maildir folders. There will be 2 (or more) dovecot instances running for this user (me) - several from Thunderbird and one from the BIFF client (icewm). Turning off the BIFF client doesn't seem to help. Symptoms: when using Thunderbird, and deleting a msg or moving it to another folder, Thunderbird will sometimes not display the next unread message, but just sit there. Changing folders works, but changing back to the previous folder (or quitting/restarting Thunderbird) only shows the first N messages (N is about 500, in a Maildir of about 6000, and it is the SAME email that is "last" every time.) This happens with multiple folders, but (usually) only folders that are delivered to by procmail (this may be related to the problem or a coincidence in that only those folders have unread email!) Accessing the mailboxes using MUTT via IMAP shows the same truncated message list. (I'm the only user using dovecot as I want Maildir, the other users have uw-imap, so dovecot is listening on port 4343) Closing all imap clients, then removing the .index.* files from the appropriate /export/01/imapd/ directory, then restarting the imap client causes the missing messages to reappear (after a suitable wait for the index files to be recreated). It really looks like locking or truncation issues in the index files.... Any clues? Debugging hints? This is happening several times per day so I can get lots o' debug data! Greg.

3 6

[Dovecot] test49 Authenticate PLAIN disabled stops pop3s
by Kim Culhan 17 Oct '04

17 Oct '04

Unlike test47, test49 refuses to authenticate pop3s if disable_plaintext_auth = yes Checked the crypto with: snort -vde tcp port 995 The connection is being made with ssl so it looks like the plaintext auth preventer is broken. -kim __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 1

[Dovecot] UIDL bug
by dovecot＠spam.turbolink.net 16 Oct '04

16 Oct '04

I noticed a problem with handling UIDL msgnumber in test46. If you do "UIDL 1" on a POP3 connection it returns the UID of the message followed by a dot and an error message. I made 2 changes: I added the following line to line 527 of src/pop3/commands.c ctx->message = message; I changed line 559 from: if (list_uids_iter(client, ctx)) to: if (!list_uids_iter(client, ctx)) Next time I'll submit an actual diff, I've made lots of other changes and comments in my source though so it's harder. - Mike

3 11

[Dovecot] 1.0-test49
by Timo Sirainen 16 Oct '04

16 Oct '04

http://dovecot.org/test/ Fixes the extra data appearing in POP3 mails (and possibly some other random problems). Created lib-sql API with asynchronous PostgreSQL and synchronous MySQL implementations. Also means I dropped mysql and pgsql userdb and passdb and replaced them with one generic sql userdb and passdb. If you were using the old ones, you'll have to fix your configuration. See included doc/dovecot-sql.conf for more information. Note that the password_query must now return "password" named column, before it was assumed that the first returned column was password.

1 0

[Dovecot] errors in logfile
by Dan Sully 15 Oct '04

15 Oct '04

I just upgraded to -test47 from 0.99.11-test4, and am getting a bunch of these messages in the logfile: dovecot: Oct 14 07:30:21 Info: imap-login: Login: andrew [64.36.1.74] dovecot: Oct 14 07:30:22 Error: IMAP(andrew): sigaction(SIGRTMIN) failed: Invalid argument dovecot: Oct 14 07:30:22 Error: IMAP(andrew): fcntl(F_SETSIG) failed: Invalid argument dovecot: Oct 14 07:30:22 Error: IMAP(andrew): read(event_pipe) failed: Is a directory dovecot: Oct 14 07:30:22 Error: child 7047 (imap) returned error 89 dovecot: Oct 14 07:30:22 Error: child 6631 (auth) killed with signal 11 Any ideas? How can I get more information about what's happening there? Thanks. -D -- <Nigel> Please refrain from fearing the reaper.

2 4

[Dovecot] LDAP with Unix Sockets
by Paolo Basenghi 15 Oct '04

15 Oct '04

Is it possible tu specify a LDAP URL like ldapi://%2fvar%2frun%2fldapi/ in the dovecot-ldap.conf file instead host:port? Best regards -- ___________________________________________________________________ Ing. PAOLO BASENGHI :::: Systems & Networking Engineer p.basenghi (at) netribe.it ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ NETRIBE srl :: Collaborative E-Business 42100 :: Reggio Emilia :: Italy :: Via della Costituzione, 27/4 ph. +39 0522 232378 :: fax +39 0522 232386 :: http://www.netribe.it ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

2 1

[Dovecot] source rpm compile
by kakula 14 Oct '04

14 Oct '04

Hi, I was compiling the source rpm but had a problem. rpm -bb dovecot-xxx.spec The process stoped and showed: File must begin with "/": %{_initrddir}/dovecot

3 4

[Dovecot] 1.0-test48
by Timo Sirainen 14 Oct '04

14 Oct '04

http://dovecot.org/test/ - dovecot-auth talks now ASCII based protocol. this makes it easy to write complete replacement auth server if someone wants to do something weird. I also thought maybe Postfix could be made to talk that (or similiar) protocol as well, but no-one has answered my mail about it so far. Perhaps I should just write binary compatible libsasl2 which actually talks to dovecot-auth. - other authentication code cleanups - don't insert extra spaces at the beginning of FETCH replies (I guess most clients ignored them) - mbox fixes - some index fixes - Hopefully fixes _send_istream() assert crash with FreeBSD (sendfile() was used wrong) The authentication code changes haven't been tested too much. I only tried that PLAIN authentication works. Others might not.

2 1