- dovecot - dovecot.org

[Dovecot] test76-77 dies at SIGHUP
by Johan Fredin 22 Jul '05

22 Jul '05

Hello people This is on OpenBSD/sparc64. After chickening along with 0.99 for a while, I finally upgraded to 1.0-test. I started out with test73, which worked fine. With the upgrade to test76, I lost the ability to SIGHUP dovecot when I rotated it's logfile. Dovecot would just die abruptly, not a trace in the log. I just upgraded to test77, but the problem is still there. I've put up my logfile and the last bit of a ktrace here: http://legonet.org/~griffin/tmp/dovecot/ The ktrace is partitial mostly because it contains usernames, I can edit those out and upload the hole shebang if that'll help. I'd appreciate any tips/pointers on this one! /Johan

3 2

[Dovecot] 0.99.14 and nfs
by Cyril Feraudet 22 Jul '05

22 Jul '05

Hi, somebody have an idea to make dovecot working safly with nfs without lockd ? It seems working without problem except for files ".customflags" and ".subscriptions" is exists a patch for use the dotlock locking method for these files ? Thanks, Cyril Feraudet

3 4

[Dovecot] test77, yahoo & fetchmail
by Chris Spiegel 22 Jul '05

22 Jul '05

I've run into a rather narrowly scoped problem .. On test77 and test78, mails from Yahoo's webmail are seen, by fetchmail using imaps, to have a 0-length body. When fetching, it claims "0 body octets". Other mail clients (mutt, kmail) can view the messages fine. pop3s retrieval works as well. Every other email I've received so far works fine. It's just those sent by Yahoo's webmail service that have the problem. imaps retrieval does, however, work fine with test76. Let me know if there's anything I can do to help track this down. Thanks, Chris

2 1

[Dovecot] Re: Dovecot 0.99.11 config
by nickr 21 Jul '05

21 Jul '05

----- Original Message ----- From: nickr To: dovecot(a)dovecot.org Sent: Thursday, July 21, 2005 4:19 PM Subject: Dovecot 0.99.11 config I am at wits end...So far it seems dovecot is light years ahead of UW-IMAP. Glad to have UW-IMAP behind me...well almost. My setup = Sendmail 8.13.4 with Procmail delivery to /var/spool/mail/<userfile> openssl 0.9.8 dovecot 0.99.11 I want to continue to use the standard mbox with all the user's mail stored in /var/spool/mail/<userfile>. User's are known to this mail server via NIS and ALL data is mounted accress NFS in read only mode on this machine, including ~home directories. I want to run dovecot like UW-IMAP...trying to keep it simple-->mail delivery to mbox /var/spool/mail/<username>. My /etc/dovecot.conf: I tried both default_ mail_env = INBOX:/var/spool/mail/%u:INBOX:INDEX=MEMORY with and with out the INBOX:INDEX=MEMORY I want to disable the indicies feature. We don't need it we're about 20 engineers or so who want to continue to use the /var/spool/mail mbox-the original Sendmail location. Most users are downloading every piece of mail through OutHouse or Pine, aka OutLook. I keep getting failed to create storage meaages in /var/log/maillog... Jul 21 15:53:26 mailtest pop3-login: Login: nickr [::ffff:172.26.64.2] Jul 21 15:53:26 mailtest pop3(nickr): Failed to create storage with data: INBOX:/var/spool/mail/nickr Jul 21 15:53:26 mailtest dovecot: child 4773 (pop3) returned error 89 Permissions info for /var/spool/mail/nickr: drwxrwxrwt 3 root mail 4096 Jul 21 15:43 mail [root@mailtest mail]# ls -la nickr total 4204 drwxrwxrwx 2 nickr mail 4096 Jul 21 15:19 . drwxrwxrwt 3 root mail 4096 Jul 21 15:43 .. -rwxrwxrwx 1 nickr ftp 0 Jul 21 14:56 inbox -rwxrwxrwx 1 nickr mail 2375 Jul 21 15:12 msg.FAUO -rwxrwxrwx 1 nickr mail 4256870 Jul 20 16:55 nickr As you can see I am spoping just fine.... Thanks in advance for ANY help!!!!!!!!!

1 0

[Dovecot] linux 2.4 vs 2.6 kernel
by vmstech 20 Jul '05

20 Jul '05

hi I read ( I think in dovecot wiki) that ext2 or ext3 fs on a 2.4 kernel can get slow due to the number of files in a directory. Does anyone know if kernel 2.6 has similar limitation? I'm not at that point now, but before the box goes live, I'm wondering if I should resolve that. Glenn

14 25

[Dovecot] Debian has moved to 1.0stable
by Jaldhar H. Vyas 20 Jul '05

20 Jul '05

Now that sarge has been released, I have as requested moved Debians' Dovecot packages to the 1.0stable branch. Most of the work was actually done by Fabio Tranchitella who is joining me as co-maintainer. The new packages should be available in sid later today. A backport to sarge will also be provided as soon as I can get around to it. Please report all problems to the Debian bug tracking system (see http://bugs.debian.org for instructions or use the reportbug package) not to me directly so that Fabio can also see. -- Jaldhar H. Vyas <jaldhar(a)debian.org> La Salle Debain - http://www.braincells.com/debian/

1 0

Re: [Dovecot] linux 2.4 vs 2.6 kernel
by Paul Vatta 20 Jul '05

20 Jul '05

> > >> ------------------------------------------------------------------------ > We were posed with the same dilemma, but it was due to the need to stick with RHEL supplied packages (ReiserFS does not come with RHEL) that we stayed with ext3. Our old production mail server, running our in-house flavour of Linux, was based on the 2.4 kernel with ReiserFS. We load tested Dovecot with maildir under RHEL3 (2.4 kernel) using ext3 and found it was unable to standup to our requirements. In part this was due to an inordinate imap-login / pop-login processes being created and hanging around under RHEL3. Using RHEL4 (2.6 kernel), we found we were able to handle the required load - this is the configuration we went live with. Processes ended cleanly too (result of the improved schedular and how it interacted with our IBM x445 and its NUMA architecture?) and the I/O issues seemed to have been resolved - ext3 was fine. >> Subject: >> Re: [Dovecot] linux 2.4 vs 2.6 kernel >> From: >> Marc Perkel <marc(a)perkel.com> >> Date: >> Mon, 18 Jul 2005 18:33:54 -0700 >> >> CC: >> dovecot(a)dovecot.org >> >> >> >> >> Curtis Maloney wrote: >> >>> Marc Perkel wrote: >>> >>>> Actually - yes. Same problem. But the solution is to use the Reiser >>>> Filesystem. It doesn't have the ext3 problem. >>> >>> >>> >>> Now, I don't want to start a religious war, but... >>> >>> Ext2 has the advantages that 1) it can fall back to ext2 in recovery >>> cases, and 2) because of that, it has all the well-tested ext2 >>> recovery tools available. >>> >>> People I know who've used Reiser say it's wonderfuly fast, but if it >>> corrupts, well... save your time, and go straight to restoring your >>> backups. >>> >>> Also, take a look at the tune2fs options. I understand -O dir_index >>> will set it to use a b-tree index for faster handling of large >>> directories. >>> >>> -- >>> Curtis Maloney >>> cmaloney(a)cardgate.net >> >> >> I've been using it for 5 years and it works great. It specifically >> eliminates the speed problems of Maildir where you have thousands of >> files in a single directory. And it has infinite inodes so you never >> run out of them. Reiser is ideal for email systems. >> >>------------------------------------------------------------------------ >> >>_______________________________________________ >>dovecot mailing list >>dovecot(a)dovecot.org >>http://dovecot.org/cgi-bin/mailman/listinfo/dovecot >>

5 7

[Dovecot] Compiling under Fedora Core 4 - Problem
by Marc Perkel 19 Jul '05

19 Jul '05

OK - trying to migrate to dovecot and I like what I see so far, but having a hard time getting it to work. I decided to go with the 1.0 version because I need to get away from the ~/Mail namespace. I'm trying to port from a Linuxconf virtual WU-IMAP type config. So - I compiled but then decided I wanted mysql so I tried to reconfigure and now getting compile errors. Looks like I'm missing something? /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x25): In function `report_errors': : undefined reference to `ERR_get_error_line_data' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0xa4): In function `vio_ssl_read': : undefined reference to `SSL_read' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0xc5): In function `vio_ssl_read': : undefined reference to `SSL_get_error' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0xf9): In function `vio_ssl_write': : undefined reference to `SSL_write' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x22d): In function `vio_ssl_close': : undefined reference to `SSL_shutdown' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x23d): In function `vio_ssl_close': : undefined reference to `SSL_free' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x35c): In function `sslaccept': : undefined reference to `SSL_new' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x372): In function `sslaccept': : undefined reference to `SSL_clear' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x380): In function `sslaccept': : undefined reference to `SSL_get_session' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x38f): In function `sslaccept': : undefined reference to `SSL_SESSION_set_timeout' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x3a3): In function `sslaccept': : undefined reference to `SSL_set_fd' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x3b1): In function `sslaccept': : undefined reference to `SSL_set_accept_state' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x3bf): In function `sslaccept': : undefined reference to `SSL_do_handshake' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x3d5): In function `sslaccept': : undefined reference to `SSL_get_peer_certificate' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x3e3): In function `sslaccept': : undefined reference to `X509_get_subject_name' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x3fb): In function `sslaccept': : undefined reference to `X509_NAME_oneline' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x40b): In function `sslaccept': : undefined reference to `X509_get_issuer_name' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x423): In function `sslaccept': : undefined reference to `X509_NAME_oneline' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x433): In function `sslaccept': : undefined reference to `X509_free' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x453): In function `sslaccept': : undefined reference to `SSL_get_shared_ciphers' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x4c1): In function `sslaccept': : undefined reference to `SSL_free' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x53b): In function `sslconnect': : undefined reference to `SSL_new' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x54d): In function `sslconnect': : undefined reference to `SSL_clear' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x55b): In function `sslconnect': : undefined reference to `SSL_get_session' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x56a): In function `sslconnect': : undefined reference to `SSL_SESSION_set_timeout' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x57e): In function `sslconnect': : undefined reference to `SSL_set_fd' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x58c): In function `sslconnect': : undefined reference to `SSL_set_connect_state' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x59a): In function `sslconnect': : undefined reference to `SSL_do_handshake' /usr/lib/mysql/libmysqlclient.a(viossl.o)(.text+0x607): In function `sslconnect': : undefined reference to `SSL_free' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x8): In function `get_dh512': : undefined reference to `DH_new' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x2a): In function `get_dh512': : undefined reference to `BN_bin2bn' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x49): In function `get_dh512': : undefined reference to `BN_bin2bn' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x67): In function `get_dh512': : undefined reference to `DH_free' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x9a): In function `report_errors': : undefined reference to `ERR_get_error_line_data' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0xcd): In function `vio_set_cert_stuff': : undefined reference to `SSL_CTX_use_certificate_file' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0xf1): In function `vio_set_cert_stuff': : undefined reference to `SSL_CTX_use_PrivateKey_file' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0xfd): In function `vio_set_cert_stuff': : undefined reference to `SSL_CTX_check_private_key' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x14e): In function `vio_set_cert_stuff': : undefined reference to `ERR_print_errors_fp' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x1c1): In function `vio_set_cert_stuff': : undefined reference to `ERR_print_errors_fp' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x1e9): In function `vio_verify_callback': : undefined reference to `X509_STORE_CTX_get_current_cert' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x1f3): In function `vio_verify_callback': : undefined reference to `X509_STORE_CTX_get_error' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x1fb): In function `vio_verify_callback': : undefined reference to `X509_STORE_CTX_get_error_depth' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x209): In function `vio_verify_callback': : undefined reference to `X509_get_subject_name' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x223): In function `vio_verify_callback': : undefined reference to `X509_NAME_oneline' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x271): In function `vio_verify_callback': : undefined reference to `X509_get_issuer_name' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x285): In function `vio_verify_callback': : undefined reference to `X509_NAME_oneline' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x2e7): In function `new_VioSSLConnectorFd': : undefined reference to `TLSv1_client_method' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x2f2): In function `new_VioSSLConnectorFd': : undefined reference to `SSL_CTX_new' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x30c): In function `new_VioSSLConnectorFd': : undefined reference to `SSL_CTX_set_cipher_list' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x326): In function `new_VioSSLConnectorFd': : undefined reference to `SSL_CTX_set_verify' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x350): In function `new_VioSSLConnectorFd': : undefined reference to `SSL_CTX_load_verify_locations' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x379): In function `new_VioSSLConnectorFd': : undefined reference to `SSL_CTX_ctrl' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x381): In function `new_VioSSLConnectorFd': : undefined reference to `DH_free' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x394): In function `new_VioSSLConnectorFd': : undefined reference to `SSL_CTX_set_default_verify_paths' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x3bc): In function `new_VioSSLConnectorFd': : undefined reference to `SSL_load_error_strings' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x3cd): In function `new_VioSSLConnectorFd': : undefined reference to `OPENSSL_add_all_algorithms_noconf' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x422): In function `new_VioSSLAcceptorFd': : undefined reference to `TLSv1_server_method' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x42d): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_CTX_new' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x447): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_CTX_set_cipher_list' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x469): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_CTX_ctrl' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x483): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_CTX_set_verify' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x49c): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_CTX_set_session_id_context' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x4c6): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_CTX_load_verify_locations' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x4ef): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_CTX_ctrl' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x4f7): In function `new_VioSSLAcceptorFd': : undefined reference to `DH_free' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x50a): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_CTX_set_default_verify_paths' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x532): In function `new_VioSSLAcceptorFd': : undefined reference to `SSL_load_error_strings' /usr/lib/mysql/libmysqlclient.a(viosslfactories.o)(.text+0x543): In function `new_VioSSLAcceptorFd': : undefined reference to `OPENSSL_add_all_algorithms_noconf' collect2: ld returned 1 exit status make[3]: *** [dovecot-auth] Error 1 make[3]: Leaving directory `/ctyme/linuxapps/dovecot/dovecot-1.0-test78/src/auth' make[2]: *** [all-recursive] Error 1 make[2]: Leaving directory `/ctyme/linuxapps/dovecot/dovecot-1.0-test78/src' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/ctyme/linuxapps/dovecot/dovecot-1.0-test78' make: *** [all] Error 2 -- Marc Perkel - marc(a)perkel.com Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com

2 1

[Dovecot] dovecot(pam_unix)[13823]: authentication failure; logname= uid=0 euid=0 tty= ruser= rhost
by Dovecot 19 Jul '05

19 Jul '05

I installed Dovecot on our FC3 sendmail server and get this from our Dallas office ( out of state to me ). dovecot(pam_unix)[13823]: authentication failure; logname= uid=0 euid=0 tty= ruser= rhost Only one Dallas user can get mail (pop3 or imap). All local users work fine. I ended upgoing back to UW and all works fine. What should I look for? Thanks

1 0

Re: [Dovecot] linux 2.4 vs 2.6 kernel
by Brian Bartholomew 19 Jul '05

19 Jul '05

I can heartily recommend reiserfs on kernel 2.4 for Intel and Sparc. My spam filtering setup currently has 27K files in a quarantine dir, and it works just fine. In a year and a half over 50 machines I've had three corrupted reiserfs filesystems, none of them on production machines, at least one of which I can blame on a bad disk. I would not be surprised to learn they were all bad disks, because the disks are five+ years old. For the last one I moved the bad dir aside where it sat peacefully for a week, and when I fsck'ed it the more aggressive mode of fsck coredumped partway through and it became unaccessable. I then restored it from the tar I had made prior. Brian

1 0