Re: [Dovecot] New server missing auth-worker entries in logs

2013-12-21 Thread Timo Sirainen
On 21.12.2013, at 23.24, Charles Marcus wrote: > Ok, I'm making progress on my new server, but there is an entry in the logs > on the old/current server that is not showing p on the new one... lines like: > > dovecot: auth-worker(20284): mysql(localhost): Connected to database > my_auth_db >

Re: [Dovecot] Eliminate legacy INBOX namespace - how?

2013-12-21 Thread Timo Sirainen
On 21.12.2013, at 19.12, Charles Marcus wrote: >> I don't think it can be done easily with just a single config. Whatever you >> do to new users might break existing setups. So the only good way I think >> would be to use two different IPs. One for the new setups, one for the old. >> For examp

[Dovecot] New server missing auth-worker entries in logs

2013-12-21 Thread Charles Marcus
Hi all, Ok, I'm making progress on my new server, but there is an entry in the logs on the old/current server that is not showing p on the new one... lines like: dovecot: auth-worker(20284): mysql(localhost): Connected to database my_auth_db I've looked everywhere, compared settings, etc,

Re: [Dovecot] Eliminate legacy INBOX namespace - how?

2013-12-21 Thread Eugene
Hello, It has been a while since I migrated from Courier to Dovecot, but I don't remember any special problems. They both were using $HOME/Maildir for storage so I did not even had to move anything, just remove that INBOX root prefix from the client config and refresh the folder list. Best w

Re: [Dovecot] File descriptor leak in sieve-extprograms

2013-12-21 Thread Kirill A. Shutemov
On Sat, Dec 21, 2013 at 03:18:31PM +0100, Stephan Bosch wrote: > On 12/14/2013 5:39 PM, Kirill A. Shutemov wrote: > > Stephan Bosch wrote: > >> On 12/13/2013 11:14 PM, Kirill A. Shutemov wrote: > >>> Stephan Bosch wrote: > This probably fixes it, although I am not entirely sure. I am mainly a

Re: [Dovecot] Eliminate legacy INBOX namespace - how?

2013-12-21 Thread Charles Marcus
On 2013-12-21 10:58 AM, Timo Sirainen wrote: On 21.12.2013, at 17.49, Charles Marcus wrote: Our mailstore was converted from a Courier-IMAP system last year, but I am planning a migration to a shiny new VM, and would really like to clean things up and do away with the legacy INBOX namespace

Re: [Dovecot] [SOLVED] Errors in log file

2013-12-21 Thread Christian Rößner
Hi, This is why it's not a good idea to use home as the Maildir root directory. You can work around this by setting maildir_stat_dirs=yes (minor performance hit). A bigger problem comes if a user creates a "dovecot/sieve" folder. > >>> I fully agree. The mailboxes are histo

Re: [Dovecot] Eliminate legacy INBOX namespace - how?

2013-12-21 Thread Timo Sirainen
On 21.12.2013, at 17.49, Charles Marcus wrote: > Our mailstore was converted from a Courier-IMAP system last year, but I am > planning a migration to a shiny new VM, and would really like to clean things > up and do away with the legacy INBOX namespace. > > My goal is to simplify IMAP client s

[Dovecot] Eliminate legacy INBOX namespace - how?

2013-12-21 Thread Charles Marcus
Hi all, Our mailstore was converted from a Courier-IMAP system last year, but I am planning a migration to a shiny new VM, and would really like to clean things up and do away with the legacy INBOX namespace. My goal is to simplify IMAP client setup - no special 'INBOX' namespace needed, the

Re: [Dovecot] [SOLVED] Errors in log file

2013-12-21 Thread Charles Marcus
On 2013-12-20 6:47 PM, Christian Rößner wrote: This is why it's not a good idea to use home as the Maildir root directory. You can work around this by setting maildir_stat_dirs=yes (minor performance hit). A bigger problem comes if a user creates a "dovecot/sieve" folder. I fully agree. The

Re: [Dovecot] File descriptor leak in sieve-extprograms

2013-12-21 Thread Stephan Bosch
On 12/14/2013 5:39 PM, Kirill A. Shutemov wrote: > Stephan Bosch wrote: >> On 12/13/2013 11:14 PM, Kirill A. Shutemov wrote: >>> Stephan Bosch wrote: This probably fixes it, although I am not entirely sure. I am mainly a bit puzzled on how this causes the fd leak; the fd_in and fd_out are