This documentation is for Dovecot v1.x, see wiki2 for v2.x documentation.
Differences between revisions 84 and 85
Revision 84 as of 2007-12-11 19:48:01
Size: 41339
Editor: generator
Revision 85 as of 2008-02-16 14:48:01
Size: 41565
Editor: generator
Deletions are marked like this. Additions are marked like this.
Line 174: Line 174:
 mail_extra_groups =:: Grant access to these extra groups for mail processes. Typical use would be to give "mail" group write access to /var/mail to be able to create dotlocks.  mail_extra_groups =:: Grant access to these extra groups for mail processes. Typical use would be to give "mail" group write access to /var/mail to be able to create dotlocks. WARNING: If your users can create symlinks, this will allow the users to read any files that are group-readable by one of these groups! Make sure at least all the common mailboxes have 0600 permissions (or a different group).

Dovecot configuration file

TableOfContents If you're in a hurry, see [:QuickConfiguration:QuickConfiguration]

"dovecot -n" command gives a clean output of the changed settings. Use it instead of copy&pasting this file when posting to the Dovecot mailing list.

'#' character and everything after it is treated as comments. Extra spaces and tabs are ignored. If you want to use either of these explicitly, put the value inside quotes, eg.: key = "# char and trailing whitespace "

Default values are shown for each setting, it's not required to uncomment any of the lines. Exception to this are paths, they're just examples with the real defaults being based on configure options. The paths listed here are for configure --prefix=/usr --sysconfdir=/etc --localstatedir=/var --with-ssldir=/etc/ssl

base_dir = /var/run/dovecot/
Base directory where to store runtime data.
protocols = imap imaps
Protocols we want to be serving: imap imaps pop3 pop3s If you only want to use dovecot-auth, you can set this to "none".
listen = *

IP or host address where to listen in for connections. It's not currently possible to specify multiple addresses. "*" listens in all IPv4 interfaces. "[::]" listens in all IPv6 interfaces, but may also listen in all IPv4 interfaces depending on the operating system.

If you want to specify ports for each service, you will need to configure these settings inside the protocol imap/pop3 { ... } section, so you can specify different ports for IMAP/POP3. For example:

#   protocol imap {
#     listen = *:10143
#     ssl_listen = *:10943
#     ..
#   }
#   protocol pop3 {
#     listen = *:10100
#     ..
#   } 
disable_plaintext_auth = yes
Disable LOGIN command and all other plaintext authentications unless SSL/TLS is used (LOGINDISABLED capability). Note that if the remote IP matches the local IP (ie. you're connecting from the same computer), the connection is considered secure and plaintext authentication is allowed.
shutdown_clients = yes
Should all IMAP and POP3 processes be killed when Dovecot master process shuts down. Setting this to "no" means that Dovecot can be upgraded without forcing existing client connections to close (although that could also be a problem if the upgrade is eg. because of a security fix). This however means that after master process has died, the client processes can't write to log files anymore.


log_path =
Log file to use for error messages, instead of sending them to syslog. /dev/stderr can be used to log into stderr.
info_log_path =
Log file to use for informational and debug messages. Default is the same as log_path.
log_timestamp = "%b %d %H:%M:%S "
Prefix for each line written to log file. % codes are in strftime(3) format.
syslog_facility = mail
Syslog facility to use if you're logging to syslog. Usually if you don't want to use "mail", you'll use local0..local7. Also other standard facilities are supported.

SSL settings

ssl_listen =
IP or host address where to listen in for SSL connections. Defaults to above if not specified.
ssl_disable = no
Disable SSL/TLS support.
ssl_cert_file = /etc/ssl/certs/dovecot.pem
PEM encoded X.509 SSL/TLS certificate and private key. They're opened before dropping root privileges, so keep the key file unreadable by anyone but root. Included doc/ can be used to easily generate self-signed certificate, just make sure to update the domains in dovecot-openssl.cnf
ssl_key_file = /etc/ssl/private/dovecot.pem
ssl_key_password =
If key file is password protected, give the password here. Alternatively give it when starting dovecot with -p parameter.
ssl_ca_file =
File containing trusted SSL certificate authorities. Set this only if you intend to use ssl_verify_client_cert=yes. The CAfile should contain the CA-certificate(s) followed by the matching CRL(s).
ssl_verify_client_cert = no
Request client to send a certificate. If you also want to require it, set ssl_require_client_cert=yes in auth section.
ssl_parameters_regenerate = 168
How often to regenerate the SSL parameters file. Generation is quite CPU intensive operation. The value is in hours, 0 disables regeneration entirely.
ssl_cipher_list = ALL:!LOW
SSL ciphers to use
verbose_ssl = no
Show protocol level SSL errors.

Login processes


login_dir = /var/run/dovecot/login
Directory where authentication process places authentication UNIX sockets which login needs to be able to connect to. The sockets are created when running as root, so you don't have to worry about permissions. Note that everything in this directory is deleted when Dovecot is started.
login_chroot = yes

chroot login process to the login_dir. Only reason not to do this is if you wish to run the whole Dovecot without roots. <doc/wiki/Rootless.txt>

login_user = dovecot

User to use for the login process. Create a completely new user for this, and don't use it anywhere else. The user must also belong to a group where only it has access, it's used to control access for authentication process. Note that this user is NOT used to access mails. <doc/wiki/UserIds.txt>

login_process_size = 64
Set max. process size in megabytes. If you don't use login_process_per_connection you might need to grow this.
login_process_per_connection = yes
Should each login be processed in it's own process (yes), or should one login process be allowed to process multiple connections (no)? Yes is more secure, espcially with SSL/TLS enabled. No is faster since there's no need to create processes all the time.
login_processes_count = 3
Number of login processes to keep for listening new connections.
login_max_processes_count = 128
Maximum number of login processes to create. The listening process count usually stays at login_processes_count, but when multiple users start logging in at the same time more extra processes are created. To prevent fork-bombing we check only once in a second if new processes should be created - if all of them are used at the time, we double their amount until the limit set by this setting is reached.
login_max_connections = 256
Maximum number of connections allowed per each login process. This setting is used only if login_process_per_connection=no. Once the limit is reached, the process notifies master so that it can create a new login process. You should make sure that the process has at least 16 + login_max_connections * 2 available file descriptors.
login_greeting = Dovecot ready.
Greeting message for clients.
login_log_format_elements = user=<%u> method=%m rip=%r lip=%l %c
Space-separated list of elements we want to log. The elements which have a non-empty variable value are joined together to form a comma-separated string.
login_log_format = %$: %s
Login log format. %$ contains login_log_format_elements string, %s contains the data we want to log.

Mailbox locations and namespaces

mail_location =
Location for users' mailboxes. This is the same as the old default_mail_env setting. The default is empty, which means that Dovecot tries to find the mailboxes automatically. This won't work if the user doesn't have any mail yet, so you should explicitly tell Dovecot the full location.

If you're using mbox, giving a path to the INBOX file (eg. /var/mail/%u) isn't enough. You'll also need to tell Dovecot where the other mailboxes are kept. This is called the "root mail directory", and it must be the first path given in the mail_location setting.

There are a few special variables you can use, eg.:

#   %u - username
#   %n - user part in user@domain, same as %u if there's no domain
#   %d - domain part in user@domain, empty if there's no domain
#   %h - home directory 

See doc/wiki/Variables.txt for full list. Some examples:

#   mail_location = maildir:~/Maildir
#   mail_location = mbox:~/mail:INBOX=/var/mail/%u
#   mail_location = mbox:/var/mail/%d/%1n/%n:INDEX=/var/indexes/%d/%1n/%n 


namespace private {
If you need to set multiple mailbox locations or want to change default namespace settings, you can do it by defining namespace sections. NOTE: Namespaces currently work ONLY with IMAP! POP3 and LDA currently ignore namespaces completely, they use only the mail_location setting.

You can have private, shared and public namespaces. The only difference between them is how Dovecot announces them to client via NAMESPACE extension. Shared namespaces are meant for user-owned mailboxes which are shared to other users, while public namespaces are for more globally accessible mailboxes.

REMEMBER: If you add any namespaces, the default namespace must be added explicitly, ie. mail_location does nothing unless you have a namespace without a location setting. Default namespace is simply done by having a namespace with empty prefix.

separator =
Hierarchy separator to use. You should use the same separator for all namespaces or some clients get confused. '/' is usually a good one. The default however depends on the underlying mail storage format.
prefix =
Prefix required to access this namespace. This needs to be different for all namespaces. For example "Public/".
location =
Physical location of the mailbox. This is in same format as mail_location, which is also the default for it.
inbox = yes
There can be only one INBOX, and this setting defines which namespace has it.
hidden = yes
If namespace is hidden, it's not advertised to clients via NAMESPACE extension or shown in LIST replies. This is mostly useful when converting from another server with different namespaces which you want to depricate but still keep working. For example you can create hidden namespaces with prefixes "~/mail/", "~%u/mail/" and "mail/".
mail_extra_groups =
Grant access to these extra groups for mail processes. Typical use would be to give "mail" group write access to /var/mail to be able to create dotlocks. WARNING: If your users can create symlinks, this will allow the users to read any files that are group-readable by one of these groups! Make sure at least all the common mailboxes have 0600 permissions (or a different group).
mail_full_filesystem_access = no
Allow full filesystem access to clients. There's no access checks other than what the operating system does for the active UID/GID. It works with both maildir and mboxes, allowing you to prefix mailboxes names with eg. /path/ or ~user/.

Mail processes

mail_debug = no
Enable mail process debugging. This can help you figure out why Dovecot isn't finding your mails.
mail_log_prefix = "%Us(%u): "
Log prefix for mail processes. See doc/wiki/Variables.txt for list of possible variables you can use.
mail_log_max_lines_per_sec = 10
Max. number of lines a mail process is allowed to log per second before it's throttled. 0 means unlimited. Typically there's no need to change this unless you're using mail_log plugin, which may log a lot.
mmap_disable = no
Don't use mmap() at all. This is required if you store indexes to shared filesystems (NFS or clustered filesystem).
mmap_no_write = no
Don't write() to mmaped files. This is required for some operating systems which use separate caches for them, such as OpenBSD.
dotlock_use_excl = no
Rely on O_EXCL to work when creating dotlock files. The default is to use hard linking. O_EXCL makes the dotlocking faster, but it doesn't always work with NFS.
fsync_disable = no
Don't use fsync() or fdatasync() calls. This makes the performance better at the cost of potential data loss if the server (or the file server) goes down.
lock_method = fcntl
Locking method for index files. Alternatives are fcntl, flock and dotlock. Dotlocking uses some tricks which may create more disk I/O than other locking methods. NFS users: flock doesn't work, remember to change mmap_disable.
mail_drop_priv_before_exec = no
Drop all privileges before exec()ing the mail process. This is mostly meant for debugging, otherwise you don't get core dumps. It could be a small security risk if you use single UID for multiple users, as the users could ptrace() each others processes then.
verbose_proctitle = no
Show more verbose process titles (in ps). Currently shows user name and IP address. Useful for seeing who are actually using the IMAP processes (eg. shared mailboxes or if same uid is used for multiple accounts).
first_valid_uid = 500
Valid UID range for users, defaults to 500 and above. This is mostly to make sure that users can't log in as daemons or other system users. Note that denying root logins is hardcoded to dovecot binary and can't be done even if first_valid_uid is set to 0.
last_valid_uid = 0
first_valid_gid = 1
Valid GID range for users, defaults to non-root/wheel. Users having non-valid GID as primary group ID aren't allowed to log in. If user belongs to supplementary groups with non-valid GIDs, those groups are not set.
last_valid_gid = 0
max_mail_processes = 1024
Maximum number of running mail processes. When this limit is reached, new users aren't allowed to log in.
mail_process_size = 256
Set max. process size in megabytes. Most of the memory goes to mmap()ing files, so it shouldn't harm much even if this limit is set pretty high.
mail_max_keyword_length = 50
Maximum allowed length for mail keyword name. It's only forced when trying to create new keywords.
valid_chroot_dirs =

':' separated list of directories under which chrooting is allowed for mail processes (ie. /var/mail will allow chrooting to /var/mail/foo/bar too). This setting doesn't affect login_chroot or auth chroot variables. WARNING: Never add directories here which local users can modify, that may lead to root exploit. Usually this should be done only if you don't allow shell access for users. <doc/wiki/Chrooting.txt>

mail_chroot =

Default chroot directory for mail processes. This can be overridden for specific users in user database by giving /./ in user's home directory (eg. /home/./user chroots into /home). Note that usually there is no real need to do chrooting, Dovecot doesn't allow users to access files outside their mail directory anyway. <doc/wiki/Chrooting.txt>

Mailbox handling optimizations

mail_cache_fields =
Space-separated list of fields to initially save into cache file. Currently these fields are allowed:

#  flags, date.sent, date.received, size.virtual, size.physical
#, imap.body, imap.bodystructure 

Different IMAP clients work in different ways, so they benefit from different cached fields. Some do not benefit from them at all. Caching more than necessary generates useless disk I/O, so you don't want to do that either.

Dovecot attempts to automatically figure out what client wants and it keeps only that. However the first few times a mailbox is opened, Dovecot hasn't yet figured out what client needs, so it may not perform optimally. If you know what fields the majority of your clients need, it may be useful to set these fields by hand. If client doesn't actually use them, Dovecot will eventually drop them.

Usually you should just leave this field alone. The potential benefits are typically unnoticeable.

mail_never_cache_fields =
Space-separated list of fields that Dovecot should never save to cache file. Useful if you want to save disk space at the cost of more I/O when the fields needed.
mail_cache_min_mail_count = 0
The minimum number of mails in a mailbox before updates are done to cache file. This allows optimizing Dovecot's behavior to do less disk writes at the cost of more disk reads.
mailbox_idle_check_interval = 30
When IDLE command is running, mailbox is checked once in a while to see if there are any new mails or other changes. This setting defines the minimum time in seconds to wait between those checks. Dovecot can also use dnotify, inotify and kqueue to find out immediately when changes occur.
mail_save_crlf = no
Save mails with CR+LF instead of plain LF. This makes sending those mails take less CPU, especially with sendfile() syscall with Linux and FreeBSD. But it also creates a bit more disk I/O which may just make it slower. Also note that if other software reads the mboxes/maildirs, they may handle the extra CRs wrong and cause problems.

Maildir-specific settings

maildir_stat_dirs = no

By default LIST command returns all entries in maildir beginning with a dot. Enabling this option makes Dovecot return only entries which are directories. This is done by stat()ing each entry, so it causes more disk I/O. (For systems setting struct dirent->d_type, this check is free and it's done always regardless of this setting)

maildir_copy_with_hardlinks = no
When copying a message, do it with hard links whenever possible. This makes the performance much better, and it's unlikely to have any side effects.
maildir_copy_preserve_filename = no
When copying a message, try to preserve the base filename. Only if the destination mailbox already contains the same name (ie. the mail is being copied there twice), a new name is given. The destination filename check is done only by looking at dovecot-uidlist file, so if something outside Dovecot does similar filename preserving copies, you may run into problems. NOTE: This setting requires maildir_copy_with_hardlinks = yes to work.

mbox-specific settings

mbox_read_locks = fcntl
Which locking methods to use for locking mbox. There are four available:

#  dotlock: Create <mailbox>.lock file. This is the oldest and most NFS-safe
#           solution. If you want to use /var/mail/ like directory, the users
#           will need write access to that directory.
#  fcntl  : Use this if possible. Works with NFS too if lockd is used.
#  flock  : May not exist in all systems. Doesn't work with NFS.
#  lockf  : May not exist in all systems. Doesn't work with NFS. 

You can use multiple locking methods; if you do the order they're declared in is important to avoid deadlocks if other MTAs/MUAs are using multiple locking methods as well. Some operating systems don't allow using some of them simultaneously.

mbox_write_locks = dotlock fcntl
mbox_lock_timeout = 300
Maximum time in seconds to wait for lock (all of them) before aborting.
mbox_dotlock_change_timeout = 120
If dotlock exists but the mailbox isn't modified in any way, override the lock file after this many seconds.
mbox_dirty_syncs = yes
When mbox changes unexpectedly we have to fully read it to find out what changed. If the mbox is large this can take a long time. Since the change is usually just a newly appended mail, it'd be faster to simply read the new mails. If this setting is enabled, Dovecot does this but still safely fallbacks to re-reading the whole mbox file whenever something in mbox isn't how it's expected to be. The only real downside to this setting is that if some other MUA changes message flags, Dovecot doesn't notice it immediately. Note that a full sync is done with SELECT, EXAMINE, EXPUNGE and CHECK commands.
mbox_very_dirty_syncs = no
Like mbox_dirty_syncs, but don't do full syncs even with SELECT, EXAMINE, EXPUNGE or CHECK commands. If this is set, mbox_dirty_syncs is ignored.
mbox_lazy_writes = yes
Delay writing mbox headers until doing a full write sync (EXPUNGE and CHECK commands and when closing the mailbox). This is especially useful for POP3 where clients often delete all mails. The downside is that our changes aren't immediately visible to other MUAs.
mbox_min_index_size = 0
If mbox size is smaller than this (in kilobytes), don't write index files. If an index file already exists it's still read, just not updated.

dbox-specific settings

dbox_rotate_size = 2048
Maximum dbox file size in kilobytes until it's rotated.
dbox_rotate_min_size = 16
Minimum dbox file size in kilobytes before it's rotated (overrides dbox_rotate_days)
dbox_rotate_days = 0
Maximum dbox file age in days until it's rotated. Day always begins from midnight, so 1 = today, 2 = yesterday, etc. 0 = check disabled.

IMAP specific settings

protocol imap {
login_executable = /usr/libexec/dovecot/imap-login
Login executable location.
mail_executable = /usr/libexec/dovecot/imap
IMAP executable location. Changing this allows you to execute other binaries before the imap process is executed.

This would write rawlogs into ~/dovecot.rawlog/ directory:

#   mail_executable = /usr/libexec/dovecot/rawlog /usr/libexec/dovecot/imap 

This would attach gdb into the imap process and write backtraces into /tmp/gdbhelper.* files:

#   mail_executable = /usr/libexec/dovecot/gdbhelper /usr/libexec/dovecot/imap 
imap_max_line_length = 65536
Maximum IMAP command line length in bytes. Some clients generate very long command lines with huge mailboxes, so you may need to raise this if you get "Too long argument" or "IMAP command line too large" errors often.
mail_plugins =
Support for dynamically loadable plugins. mail_plugins is a space separated list of plugins to load.
mail_plugin_dir = /usr/lib/dovecot/imap
login_greeting_capability = no
Send IMAP capabilities in greeting message. This makes it unnecessary for clients to request it with CAPABILITY command, so it saves one round-trip. Many clients however don't understand it and ask the CAPABILITY anyway.
imap_capability =
Override the IMAP CAPABILITY response.
imap_client_workarounds = outlook-idle
Workarounds for various client bugs:

#   delay-newmail:
#     Send EXISTS/RECENT new mail notifications only when replying to NOOP
#     and CHECK commands. Some clients ignore them otherwise, for example OSX
#     Mail (<v2.1). Outlook Express breaks more badly though, without this it
#     may show user "Message no longer in server" errors. Note that OE6 still
#     breaks even with this workaround if synchronization is set to
#     "Headers Only".
#   outlook-idle:
#     Outlook and Outlook Express never abort IDLE command, so if no mail
#     arrives in half a hour, Dovecot closes the connection. This is still
#     fine, except Outlook doesn't connect back so you don't see if new mail
#     arrives.
#   netscape-eoh:
#     Netscape 4.x breaks if message headers don't end with the empty "end of
#     headers" line. Normally all messages have this, but setting this
#     workaround makes sure that Netscape never breaks by adding the line if
#     it doesn't exist. This is done only for FETCH BODY{{{[HEADER.FIELDS..]

# commands. Note that RFC says this shouldn't be done. # tb-extra-mailbox-sep: # With mbox storage a mailbox can contain either mails or submailboxes, # but not both. Thunderbird separates these two by forcing server to # accept '/' suffix in mailbox names in subscriptions list. }}}

The list is space-separated.


POP3 specific settings

protocol pop3 {
login_executable = /usr/libexec/dovecot/pop3-login
Login executable location.
mail_executable = /usr/libexec/dovecot/pop3
POP3 executable location. See IMAP's mail_executable above for examples how this could be changed.
pop3_no_flag_updates = no
Don't try to set mails non-recent or seen with POP3 sessions. This is mostly intended to reduce disk I/O. With maildir it doesn't move files from new/ to cur/, with mbox it doesn't write Status-header.
pop3_enable_last = no
Support LAST command which exists in old POP3 specs, but has been removed from new ones. Some clients still wish to use this though. Enabling this makes RSET command clear all \Seen flags from messages.
pop3_reuse_xuidl = no
If mail has X-UIDL header, use it as the mail's UIDL.
pop3_lock_session = no
Keep the mailbox locked for the entire POP3 session.
pop3_uidl_format =
POP3 UIDL (unique mail identifier) format to use. You can use following variables:

#  %v - Mailbox's IMAP UIDVALIDITY
#  %u - Mail's IMAP UID
#  %m - MD5 sum of the mailbox headers in hex (mbox only)
#  %f - filename (maildir only) 

If you want UIDL compatibility with other POP3 servers, use:

#  UW's ipop3d         : %08Xv%08Xu
#  Courier version 0   : %f
#  Courier version 1   : %u
#  Courier version 2   : %v-%u
#  Cyrus (<= 2.1.3)    : %u
#  Cyrus (>= 2.1.4)    : %v.%u
#  Older Dovecots      : %v.%u
#  tpop3d              : %Mf 

Note that Outlook 2003 seems to have problems with %v.%u format which was Dovecot's default, so if you're building a new server it would be a good idea to change this. %08Xu%08Xv should be pretty fail-safe.

NOTE: Nowadays this is required to be set explicitly, since the old default was bad but it couldn't be changed without breaking existing installations. %08Xu%08Xv will be the new default, so use it for new installations.

pop3_logout_format = top=%t/%p, retr=%r/%b, del=%d/%m, size=%s
POP3 logout format string:

#  %t - number of TOP commands
#  %p - number of bytes sent to client as a result of TOP command
#  %r - number of RETR commands
#  %b - number of bytes sent to client as a result of RETR command
#  %d - number of deleted messages
#  %m - number of messages (before deletion)
#  %s - mailbox size in bytes (before deletion) 
mail_plugins =
Support for dynamically loadable plugins. mail_plugins is a space separated list of plugins to load.
mail_plugin_dir = /usr/lib/dovecot/pop3
pop3_client_workarounds =
Workarounds for various client bugs:

#   outlook-no-nuls:
#     Outlook and Outlook Express hang if mails contain NUL characters.
#     This setting replaces them with 0x80 character.
#   oe-ns-eoh:
#     Outlook Express and Netscape Mail breaks if end of headers-line is
#     missing. This option simply sends it if it's missing. 

The list is space-separated.


LDA specific settings

protocol lda {
postmaster_address =
Address to use when sending rejection mails.
hostname =
Hostname to use in various parts of sent mails, eg. in Message-Id. Default is the system's real hostname.
mail_plugins =
Support for dynamically loadable plugins. mail_plugins is a space separated list of plugins to load.
mail_plugin_dir = /usr/lib/dovecot/lda
sendmail_path = /usr/lib/sendmail
Binary to use for sending mails.
auth_socket_path = /var/run/dovecot/auth-master
UNIX socket path to master authentication server to find users.

Authentication processes

auth_executable = /usr/libexec/dovecot/dovecot-auth
Executable location
auth_process_size = 256
Set max. process size in megabytes.
auth_cache_size = 0
Authentication cache size in kilobytes. 0 means it's disabled. Note that bsdauth, PAM and vpopmail require cache_key to be set for caching to be used.
auth_cache_ttl = 3600
Time to live in seconds for cached data. After this many seconds the cached record is no longer used, *except* if the main database lookup returns internal failure. We also try to handle password changes automatically: If user's previous authentication was successful, but this one wasn't, the cache isn't used. For now this works only with plaintext authentication.
auth_realms =
Space separated list of realms for SASL authentication mechanisms that need them. You can leave it empty if you don't want to support multiple realms. Many clients simply use the first one listed here, so keep the default realm first.
auth_default_realm =
Default realm/domain to use if none was specified. This is used for both SASL realms and appending @domain to username in plaintext logins.
auth_username_chars = abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ01234567890.-_@
List of allowed characters in username. If the user-given username contains a character not listed in here, the login automatically fails. This is just an extra check to make sure user can't exploit any potential quote escaping vulnerabilities with SQL/LDAP databases. If you want to allow all characters, set this value to empty.
auth_username_translation =

Username character translations before it's looked up from databases. The value contains series of from -> to characters. For example "#@/@" means that '#' and '/' characters are translated to '@'.

auth_username_format =
Username formatting before it's looked up from databases. You can use the standard variables here, eg. %Lu would lowercase the username, %n would drop away the domain if it was given, or "%n-AT-%d" would change the '@' into "-AT-". This translation is done after auth_username_translation changes.
auth_master_user_separator =

If you want to allow master users to log in by specifying the master username within the normal username string (ie. not using SASL mechanism's support for it), you can specify the separator character here. The format is then <username><separator><master username>. UW-IMAP uses "*" as the separator, so that could be a good choice.

auth_anonymous_username = anonymous
Username to use for users logging in with ANONYMOUS SASL mechanism
auth_verbose = no
More verbose logging. Useful for figuring out why authentication isn't working.
auth_debug = no
Even more verbose logging for debugging purposes. Shows for example SQL queries.
auth_debug_passwords = no
In case of password mismatches, log the passwords and used scheme so the problem can be debugged. Requires auth_debug=yes to be set.
auth_worker_max_count = 30
Maximum number of dovecot-auth worker processes. They're used to execute blocking passdb and userdb queries (eg. MySQL and PAM). They're automatically created and destroyed as needed.
auth_gssapi_hostname =
Host name to use in GSSAPI principal names. The default is to use the name returned by gethostname().
auth_krb5_keytab =
Kerberos keytab to use for the GSSAPI mechanism. Will use the system default (usually /etc/krb5.keytab) if not specified.
auth default {
mechanisms = plain
Space separated list of wanted authentication mechanisms:

#   plain login digest-md5 cram-md5 ntlm rpa apop anonymous gssapi 

NOTE: See also disable_plaintext_auth setting.

Password database is used to verify user's password (and nothing more). You can have multiple passdbs and userdbs. This is useful if you want to allow both system users (/etc/passwd) and virtual users to login without duplicating the system users into virtual database.


By adding master=yes setting inside a passdb you make the passdb a list of "master users", who can log in as anyone else. Unless you're using PAM, you probably still want the destination user to be looked up from passdb that it really exists. This can be done by adding pass=yes setting to the master passdb. <doc/wiki/Authentication.MasterUsers.txt>

Users can be temporarily disabled by adding a passdb with deny=yes. If the user is found from that database, authentication will fail. The deny passdb should always be specified before others, so it gets checked first. Here's an example:

passdb passwd-file {
args = /etc/dovecot.deny
File contains a list of usernames, one per line
deny = yes
passdb pam {

PAM authentication. Preferred nowadays by most systems. Note that PAM can only be used to verify if user's password is correct, so it can't be used as userdb. If you don't want to use a separate user database (passwd usually), you can use static userdb. REMEMBER: You'll need /etc/pam.d/dovecot file created for PAM authentication to actually work. <doc/wiki/PasswordDatabase.PAM.txt>

args = dovecot

[blocking=yes] [session=yes] [setcred=yes] [cache_key=<key>] [<service name>]

By default a new process is forked from dovecot-auth for each PAM lookup. Setting blocking=yes uses the alternative way: dovecot-auth worker processes do the PAM lookups.

session=yes makes Dovecot open and immediately close PAM session. Some PAM plugins need this to work, such as pam_mkhomedir.

setcred=yes makes Dovecot establish PAM credentials if some PAM plugins need that. They aren't ever deleted though, so this isn't enabled by default.

cache_key can be used to enable authentication caching for PAM (auth_cache_size also needs to be set). It isn't enabled by default because PAM modules can do all kinds of checks besides checking password, such as checking IP address. Dovecot can't know about these checks without some help. cache_key is simply a list of variables (see doc/wiki/Variables.txt) which must match for the cached data to be used. Here are some examples:

#   %u - Username must match. Probably sufficient for most uses.
#   %u%r - Username and remote IP address must match.
#   %u%s - Username and service (ie. IMAP, POP3) must match. 

If service name is "*", it means the authenticating service name is used, eg. pop3 or imap (/etc/pam.d/pop3, /etc/pam.d/imap).

Some examples:

#   args = session=yes *
#   args = cache_key=%u dovecot 
passdb passwd {

System users (NSS, /etc/passwd, or similiar) In many systems nowadays this uses Name Service Switch, which is configured in /etc/nsswitch.conf. <doc/wiki/AuthDatabase.Passwd.txt>

args =

[blocking=yes] - See userdb passwd for explanation

passdb shadow {

Shadow passwords for system users (NSS, /etc/shadow or similiar). Deprecated by PAM nowadays. <doc/wiki/PasswordDatabase.Shadow.txt>

args =

[blocking=yes] - See userdb passwd for explanation

passdb bsdauth {

PAM-like authentication for OpenBSD. <doc/wiki/PasswordDatabase.BSDAuth.txt>

args =

[cache_key=<key>] - See cache_key in PAM for explanation.

passdb passwd-file {

passwd-like file with specified location <doc/wiki/AuthDatabase.PasswdFile.txt>

args =
Path for passwd-file
passdb checkpassword {

checkpassword executable authentication NOTE: You will probably want to use "userdb prefetch" with this. <doc/wiki/PasswordDatabase.CheckPassword.txt>

args =
Path for checkpassword binary
passdb sql {

SQL database <doc/wiki/AuthDatabase.SQL.txt>

args =
Path for SQL configuration file, see doc/dovecot-sql-example.conf
passdb ldap {

LDAP database <doc/wiki/AuthDatabase.LDAP.txt>

args =
Path for LDAP configuration file, see doc/dovecot-ldap-example.conf
passdb vpopmail {

vpopmail authentication <doc/wiki/AuthDatabase.VPopMail.txt>

args =

[cache_key=<key>] - See cache_key in PAM for explanation.


User database specifies where mails are located and what user/group IDs own them. For single-UID configuration use "static".


userdb passwd {

System users (NSS, /etc/passwd, or similiar). In many systems nowadays this uses Name Service Switch, which is configured in /etc/nsswitch.conf. <doc/wiki/AuthDatabase.Passwd.txt>

args =

[blocking=yes] - By default the lookups are done in the main dovecot-auth process. This setting causes the lookups to be done in auth worker proceses. Useful with remote NSS lookups that may block. NOTE: Be sure to use this setting with nss_ldap or users might get logged in as each others!

userdb passwd-file {

passwd-like file with specified location <doc/wiki/AuthDatabase.PasswdFile.txt>

args =
Path for passwd-file
userdb static {

static settings generated from template <doc/wiki/UserDatabase.Static.txt>

args =
Template for the fields. Can return anything a userdb could normally return. For example:

#  args = uid=500 gid=500 home=/var/mail/%u 

If you use deliver, it needs to look up users only from the userdb. This of course doesn't work with static because there is no list of users. Normally static userdb handles this by doing a passdb lookup. This works with most passdbs, with PAM being the most notable exception. If you do the user verification another way, you can add allow_all_users=yes to the args in which case the passdb lookup is skipped.

userdb sql {

SQL database <doc/wiki/AuthDatabase.SQL.txt>

args =
Path for SQL configuration file, see doc/dovecot-sql-example.conf
userdb ldap {

LDAP database <doc/wiki/AuthDatabase.LDAP.txt>

args =
Path for LDAP configuration file, see doc/dovecot-ldap-example.conf
userdb vpopmail {

vpopmail <doc/wiki/AuthDatabase.VPopMail.txt>

userdb prefetch {

"prefetch" user database means that the passdb already provided the needed information and there's no need to do a separate userdb lookup. This can be made to work with SQL and LDAP databases, see their example configuration files for more information how to do it. <doc/wiki/UserDatabase.Prefetch.txt>

user = root
User to use for the process. This user needs access to only user and password databases, nothing else. Only shadow and pam authentication requires roots, so use something else if possible. Note that passwd authentication with BSDs internally accesses shadow files, which also requires roots. Note that this user is NOT used to access mails. That user is specified by userdb above.
chroot =
Directory where to chroot the process. Most authentication backends don't work if this is set, and there's no point chrooting if auth_user is root. Note that valid_chroot_dirs isn't needed to use this setting.
count = 1
Number of authentication processes to create
ssl_require_client_cert = no
Require a valid SSL client certificate or the authentication fails.
ssl_username_from_cert = no

Take the username from client's SSL certificate, using X509_NAME_get_text_by_NID() which returns the subject's DN's CommonName.

socket listen {
It's possible to export the authentication interface to other programs:
master {
path = /var/run/dovecot/auth-master
Master socket provides access to userdb information. It's typically used to give Dovecot's local delivery agent access to userdb so it can find mailbox locations.
mode = 0600
user =
Default user/group is the one who started dovecot-auth (root)
group =
client {
path = /var/run/dovecot/auth-client
The client socket is generally safe to export to everyone. Typical use is to export it to your SMTP server so it can do SMTP AUTH lookups using it.
mode = 0660
auth external {
If you wish to use another authentication server than dovecot-auth, you can use connect sockets. They are assumed to be already running, Dovecot's master process only tries to connect to them. They don't need any other settings than the path for the master socket, as the configuration is done elsewhere. Note that the client sockets must exist in the login_dir.

#  socket connect {
#    master {
#      path = /var/run/dovecot/auth-master
#    }
#  } 

Dictionary server settings

Dictionary can be used by some plugins to store key=value lists. Currently this is only used by dict quota backend. The dictionary can be used either directly or though a dictionary server. The following dict block maps dictionary names to URIs when the server is used. These can then be referenced using URIs in format "proxy:<name>".

dict {
quota = mysql:/etc/dovecot-dict-quota.conf

Plugin settings

plugin {

Here you can give some extra environment variables to mail processes. This is mostly meant for passing parameters to plugins. %variable expansion is done for all values.

quota = maildir
Quota plugin. Multiple backends are supported:

#   dirsize: Find and sum all the files found from mail directory.
#            Extremely SLOW with Maildir. It'll eat your CPU and disk I/O.
#   dict: Keep quota stored in dictionary (eg. SQL)
#   maildir: Maildir++ quota
#   fs: Read-only support for filesystem quota 
acl = vfile:/etc/dovecot-acls
ACL plugin. vfile backend reads ACLs from "dovecot-acl" file from maildir directory. You can also optionally give a global ACL directory path where ACLs are applied to all users' mailboxes. The global ACL directory contains one file for each mailbox, eg. INBOX or sub.mailbox.
convert_mail = mbox:%h/mail

Convert plugin. If set, specifies the source storage path which is converted to destination storage (mail_location) when the user logs in. The existing mail directory is renamed to <dir>-converted.

convert_skip_broken_mailboxes = no
Skip mailboxes which we can't open successfully instead of aborting.
trash = /etc/dovecot-trash.conf

Trash plugin. When saving a message would make user go over quota, this plugin automatically deletes the oldest mails from configured mailboxes until the message can be saved within quota limits. The configuration file is a text file where each line is in format: <priority> <mailbox name> Mails are first deleted in lowest -> highest priority number order

Lazy expunge plugin. Currently works only with maildirs. When a user expunges mails, the mails are moved to a mailbox in another namespace (1st). When a mailbox is deleted, the mailbox is moved to another namespace (2nd) as well. Also if the deleted mailbox had any expunged messages, they're moved to a 3rd namespace. The mails won't be counted in quota, and they're not deleted automatically (use a cronjob or something).

None: MainConfig (last edited 2011-04-25 16:53:54 by generator)