Search K
Appearance
Appearance
DANGER
Mbox format is deprecated and should not be used in production setups at all.
Mbox is no longer being maintained for write fixes, for any new or advanced features, nor for optimization improvements.
It still exists solely to read old mail storages, and for backwards utility purposes (specifically for archival purposes, as mbox allows multiple messages to be natively stored in a single file).
In a production system, a more modern mailbox format should be used, e.g., dbox (or Maildir).
Usually UNIX systems are configured by default to deliver mails to /var/mail/username
or /var/spool/mail/username
mboxes. In the IMAP world, these files are called INBOX mailboxes. IMAP protocol supports multiple mailboxes , so there needs to be a place for them as well. Typically they're stored in ~/mail/
or ~/Mail/
directories.
The mbox file contains all the messages of a single mailbox. Because of this, the mbox format is typically thought of as a slow format. However with Dovecot's indexing this isn't true. Only expunging messages from the beginning of a large mbox file is slow with Dovecot, most other operations should be fast. Also because all the mails are in a single file, searching is much faster (if FTS is not used) than with Maildir.
Modifications to mbox may require moving data around within the file, so interruptions (eg. power failures) can cause the mbox to break more or less badly. Although Dovecot tries to minimize the damage by moving the data in a way that data should never get lost (only duplicated), mboxes still aren't recommended to be used for important data.
The history of mbox format, and a discussion of its historical use and generally agreed-upon conventions, can be found in RFC 4155.
Additionally, see the mbox Wikipedia page.
Locking is a mess with mboxes. There are multiple different ways to lock a mbox, and software often uses incompatible locking.
The only standard way to lock an mbox is using a method called "dotlock". This means that a file named <mailbox-name>.lock
is created in the same directory as the mailbox being locked. This works pretty well when the mbox is locked for writing, but for reading it's very inefficient. That's why other locking methods have been used.
It's important that all software that's reading or writing to mboxes use the same locking settings. If they use different methods, they might read/write to an mbox while another process is modifying it, and see corrupted mails. If they use the same methods but in a different order, they can both end up in a deadlock.
There are at least four different ways to lock a mbox:
mailboxname.lock
file created by almost all software when writing to mboxes. This grants the writer an exclusive lock over the mbox, so it's usually not used while reading the mbox so that other processes can also read it at the same time. So while using a dotlock typically prevents actual mailbox corruption, it doesn't protect against read errors if mailbox is modified while a process is reading.
Another problem with dotlocks is that if the mailboxes exist in /var/mail/
, the user may not have write access to the directory, so the dotlock file can't be created. There are a couple of ways to work around this:
Give a mail group write access to the directory and then make sure that all software requiring access to the directory runs with the group's privileges. This may mean making the binary itself setgid-mail, or using a separate dotlock helper program which is setgid-mail. With Dovecot this can be done by setting mail_privileged_group = mail
.
Set sticky bit to the directory (chmod +t /var/mail
). This makes it somewhat safe to use, because users can't delete each others mailboxes, but they can still create new files (the dotlock files). The downside to this is that users can create whatever files they wish in there, such as a mbox for newly created user who hadn't yet received mail.
flock()
system call is quite commonly used for both read and write locking. The read lock allows multiple processes to obtain a read lock for the mbox, so it works well for reading as well. The downside is that it doesn't work if mailboxes are stored in NFS.
Very similar to flock, also commonly used by software. In some systems this fcntl()
system call is compatible with flock()
, but in other systems it's not, so you shouldn't rely on it. fcntl works with NFS if you're using lockd daemon in both NFS server and client.
POSIX lockf()
locking. Because it allows creating only exclusive locks, it's somewhat useless so Dovecot doesn't support it. With Linux lockf()
is internally compatible with fcntl()
locks, but again you shouldn't rely on this.
If multiple lock methods are used, which is usually the case since dotlocks aren't typically used for read locking, the order in which the locking is done is important. Consider if two programs were running at the same time, both use dotlock and fcntl locking but in different order:
Now both of them are waiting for each others locks. Finally after a couple of minutes they time out and fail the operation.
For Dovecot you can configure locking using the mbox_read_locks
and mbox_write_locks
settings. The defaults are:
mbox_read_locks = fcntl
mbox_write_locks = dotlock fcntl
Here's a list of how to find out the locking settings for other software:
$ procmail -v 2>&1|grep Locking
Locking strategies: dotlocking, fcntl()
Postfix has two different ways to deliver to mboxes. One is the "mailbox" transport and another one is the "virtual" transport.
# postconf mailbox_delivery_lock
mailbox_delivery_lock = fcntl, dotlock
# postconf virtual_mailbox_lock
virtual_mailbox_lock = fcntl
In the above case, if you used the mailbox transport, you'd have to change Dovecot's configuration to mbox_write_locks = fcntl
dotlock or vice versa for Postfix.
If you used the virtual transport, it doesn't really matter if the dotlock is missing, since the fcntl is common with Dovecot and Postfix.
$ mutt -v|grep -i lock
Debian's policy specifies that all software should use "fcntl and then dotlock" locking, but this probably applies only to most commonly used software.
By default, when listing mailboxes, Dovecot simply assumes that all files it sees are mboxes and all directories mean that they contain sub-mailboxes. There are two special cases however which aren't listed:
.subscriptions
file contains IMAP's mailbox subscriptions..imap/
directory contains Dovecot's index files.Because it's not possible to have a file which is also a directory, it's not normally possible to create a mailbox and child mailboxes under it.
However if you really want to be able to have mailboxes containing both messages and child mailboxes under mbox, then Dovecot can be configured to do this, subject to certain provisos; see child folders.
Dovecot uses c-Client (ie. UW-IMAP, Pine) compatible headers in mbox messages to store metadata. These headers are:
Header | Description |
---|---|
X-IMAPbase | Contains UIDVALIDITY, last used UID, and list of used keywords |
X-IMAP | Same as X-IMAPbase but also specifies that the message is a "pseudo-message" |
X-UID | Message's allocated UID |
Status | R (\Seen) and O (non-\Recent) flags |
X-Status | A (\Answered), F (\Flagged), T (\Draft), and D (\Deleted) flags |
X-Keywords | Message's keywords |
Content-Length | Length of the message body in bytes |
Whenever any of these headers exist, Dovecot treats them as its own private metadata. It does sanity checks for them, so the headers may also be modified or removed completely. None of these headers are sent to IMAP/POP3 clients when they read the mail.
WARNING
The MTA, MDA, or LDA should strip all these headers case-insensitively before writing the mail to the mbox.
Only the first message contains the X-IMAP or X-IMAPbase header. The difference is that when all the messages are deleted from mbox file, a pseudo message is written to the mbox which contains X-IMAP header.
This is the "DON'T DELETE THIS MESSAGE -- FOLDER INTERNAL DATA" message which you hate seeing when using non-C-client and non-Dovecot software. This is however important to prevent abuse, otherwise the first mail which is received could contain faked X-IMAPbase header which could cause trouble.
If message contains X-Keywords header, it contains a space-separated list of keywords for the mail. Since the same header can come from the mail's sender, only the keywords are listed in X-IMAP header are used.
The UID for a new message is calculated from last used UID in X-IMAP header +
Content-Length is used as long as another valid mail starts after that many bytes. Because the byte count must be exact, it's quite unlikely that abusing it can cause messages to be skipped (or rather appended to the previous message's body).
Status and X-Status headers are trusted completely, so it's pretty good idea to filter them in LDA if possible.
Updating messages' flags and keywords can be a slow operation since you may have to insert a new header (Status, X-Status, X-Keywords) or at least insert data in the header's value. Some mbox MUAs do this simply by rewriting all of the mbox after the inserted data. If the mbox is large, this can be very slow. Dovecot optimizes this by always leaving some space characters after some of its internal headers. It can use this space to move only minimal amount of data necessary to get the necessary data inserted. Also if data is removed, it just grows these spaces areas.
There are several configuration options that can be used that will affect optimization:
In mboxes a new mail always begins with a "From " line, commonly referred to as From_
-line. To avoid confusion, lines beginning with "From " in message bodies are usually prefixed with '>' character while the message is being written to in mbox.
Dovecot doesn't currently do this escaping however. Instead it prevents this confusion by adding Content-Length headers so it knows later where the next message begins. Dovecot also doesn't remove the '>' characters before sending the data to clients.
There are a few minor variants of this format:
An original mbox format originated with Unix System V. Messages are stored in a single file, with each message beginning with a line containing "From SENDER DATE". If "From " (case-sensitive, with the space) occurs at the beginning of a line anywhere in the email, it is escaped with a greater-than sign (to ">From "). Lines already quoted as such, for example ">From " or ">>>From " are not quoted again, which leads to irrecoverable corruption of the message content.
Named for Raul Dhesi in June 1995, though several people came up with the same idea around the same time. An issue with the mboxo format was that if the text ">From " appeared in the body of an email (such as from a reply quote), it was not possible to distinguish this from the mailbox format's quoted ">From ". mboxrd fixes this by always quoting already quoted "From " lines (e.g. ">From ", ">>From ", ">>>From ", etc.) as well, so readers can just remove the first ">" character. This format is used by qmail and getmail (>=4.35.0).
Originated with Unix System V Release 4 mail tools. It adds a Content-Length field which indicates the number of bytes in the message. This is used to determine message boundaries. It still quotes "From " as the original mboxo format does (and not as mboxrd does it).
Like mboxcl but does away with the "From " quoting. Dovecot uses this format internally.
(Multi-channel Memorandum Distribution Facility mailbox format) originated with the MMDF daemon. The format surrounds each message with lines containing four control-A's. This eliminates the need to escape From: lines.
From_
lines.From_
line marks the beginning of a message.From_
line.From_
line or scans till the end of file, whenever From_
comes first.>From_
lines and >>From_
lines and so on.In general Dovecot doesn't mind if you modify the mbox file externally. It's fine if external software expunges messages or appends new ones. However moving around existing messages, inserting messages in the middle of the file or modifying existing messages isn't allowed.
Especially modifying existing messages (eg. removing attachments) may cause all kinds of problems. If you do that, at the minimum go and delete dovecot.index.cache
file from the mailbox, otherwise weird things may happen. However IMAP protocol guarantees that messages don't change at all, and deleting Dovecot's cache file doesn't clear clients' local caches, so it still may not work right.
If you insert messages, or if you "undelete" messages (eg. replace mbox from a backup), you may see errors in Dovecot's logs:
mbox sync: UID inserted in the middle of mailbox /home/tss/mail/inbox (817 > 787, seq=18, idx_msgs=32)
This is normal. Dovecot just assigned new UIDs for the messages. See below for other reasons why UID insertions could happen.
The above error message can be read as: "18th message in the mbox file contained X-UID: 787 header, however the index file at that position told the message was supposed to have UID 817. There are 32 messages currently in the index file."
There are four possibilities why the error message could happen:
Message with a X-UID: 787 header really was inserted in the mbox file. For example you replaced mbox from a backup.
Something changed the X-UID headers. Very unlikely.
The message was expunged from the index file, but for some reason it wasn't expunged from the mbox file. The index file is updated only after a successful mbox file modification, so this shouldn't really happen either.
If this problem happens constantly, it could mean that you're sharing the same index file for multiple different mboxes!
This could happen if you let Dovecot do mailbox autodetection and it sometimes uses /var/mail/%u
(when it exists) and other times ~/mail/inbox
. Use an explicit mail_location
setting to make sure the same INBOX is used.
Another possibility is that you're sharing index files between multiple users. Each user must have their own home directory.
It's possible that broken X-UID headers in mails and mbox_lazy_writes = yes
combination has some bugs. If you're able to reproduce such an error, please let us know how.
UIDVALIDITY is stored in X-IMAPbase: or X-IMAP: header of the first message in mbox file. This is done by both Dovecot and UW-IMAP (and Pine). It's also stored in dovecot.index
file. It shouldn't normally change, because if it does it means that client has to download all the messages for the mailbox again.
If the UIDVALIDITY in mbox file doesn't match the one in dovecot.index
file, Dovecot logs an error:
UIDVALIDITY changed (1100532544 -> 1178155834) in mbox file /home/user/mail/mailbox
This can happen when the following happens:
Dovecot accesses the mailbox saving the current UIDVALIDITY to dovecot.index
file.
The UIDVALIDITY gets lost from the mbox file
X-IMAP: or X-IMAPbase: header gets lost because something else than Dovecot or UW-IMAP deletes the first message
The whole file gets truncated
Something else than Dovecot deletes or renames the mbox
The mailbox is accessed (or created if necessary) by UW-IMAP or Pine. It notices that the mailbox is missing UIDVALIDITY, so it assigns a new UIDVALIDITY and writes the X-IMAPbase: or X-IMAP: header.
Dovecot accesses again the mailbox. UIDVALIDITY in the mbox file's header doesn't match the one in dovecot.index
file. It logs an error and updates the UIDVALIDITY in the index file to the new one.
Dovecot's mbox code is a bit fragile because of the way it works. However instead of just corrupting the mbox file, it usually assert-crashes whenever it notices an inconsistency. You may see crashes such as:
Panic: mbox /home/user/mail/mailbox: seq=2 uid=45 uid_broken=0 originally needed 12 bytes, now needs 27 bytes
This is a bit difficult problem to fix. Usually this crash has been related to Dovecot rewriting some headers that were broken. If you see these crashes, it would really help if you were able to reproduce the crash.
If you have such a mailbox which crashes every time when it's tried to be opened, please put the mbox through mbox anonymizer and send it, the mailbox's dovecot.index
and dovecot.index.log
files to dovecot@dovecot.org.
None of those files contain any actual message contents so it's safe to send them.
Since the problems usually have been related to broken headers, you should be able to avoid them by filtering out all the Dovecot's internal metadata headers. This is a good idea to do in any case.
If you use LDA it does this filtering automatically. Otherwise you could do this in your SMTP server. The headers that you should filter out are:
pop3_reuse_xuidl = yes
)mbox_dirty_syncs
Default | yes |
---|---|
Value | boolean |
See Also |
Enable optimized mbox syncing?
For larger mbox files, it can take a long time to determine what has changed when the file is altered unexpectedly. Since the change in most cases consists solely of newly appended mail, Dovecot can operate more quickly if it starts off by simply reading the new messages, then falls back to reading the entire mbox file if something elsewhere in it isn't as expected.
Dovecot assumes that external mbox file changes only mean that new messages were appended to it. Without this setting Dovecot re-reads the whole mbox file whenever it changes. There are various safeguards in place to make this setting safe even when other changes than appends were done to the mbox. The downside to this setting is that external message flag modifications may not be visible immediately.
When this setting is enabled, Dovecot tries to avoid re-reading the mbox
every time something changes. Whenever the mbox changes (i.e. timestamp or
size), Dovecot first checks if the mailbox's size changed. If it didn't, it
most likely meant that only message flags were changed so it does a full
mbox read to find it. If the mailbox shrunk, it means that mails were
expunged and again Dovecot does a full sync. Usually however the only thing
besides Dovecot that modifies the mbox is the LDA which appends new mails
to the mbox. So if the mbox size was grown, Dovecot first checks if the
last known message is still where it was last time. If it is, Dovecot reads
only the newly added messages and goes into "dirty mode". As long as
Dovecot is in dirty mode, it can't be certain that mails are where it
expects them to be, so whenever accessing some mail, it first verifies that
it really is the correct mail by finding its X-UID header. If the X-UID
header is different, it fallbacks to a full sync to find the mail's correct
position. The dirty mode goes away after a full sync. If
mbox_lazy_writes
was enabled and the mail didn't yet
have an X-UID header, Dovecot uses the MD5 sum of a couple of headers to
compare the mails.
mbox_dotlock_change_timeout
Default | 2 mins |
---|---|
Value | time |
See Also |
Override a lockfile after this amount of time if a dot-lock exists but the mailbox hasn't been modified in any way.
mbox_lazy_writes
Default | yes |
---|---|
Value | boolean |
See Also |
If enabled, mbox headers (e.g., metadata updates, such as writing X-UID headers or flag changes) are not written until a full write sync is performed (triggered via IMAP EXPUNGE or CHECK commands and/or when the mailbox is closed). mbox rewrites can be costly, so this may avoid a lot of disk writes.
Enabling this setting is especially useful with POP3, in which clients often delete all mail messages.
One negative consequence of enabling this setting is that the changes aren't immediately visible to other MUAs.
C-Client works the same way. The upside of this is that it reduces writes because multiple flag updates to same message can be grouped, and sometimes the writes don't have to be done at all if the whole message is expunged. The downside is that other processes don't notice the changes immediately (but other Dovecot processes do notice because the changes are in index files).
mbox_lock_timeout
Default | 5 mins |
---|---|
Value | time |
See Also |
The maximum time to wait for all locks to be released before aborting.
mbox_md5
Default | apop3d |
---|---|
Value | string |
Allowed Values | apop3d all |
See Also | |
Advanced Setting; this should not normally be changed. |
The mail-header selection algorithm to use for MD5 POP3 UIDLs when the
setting pop3_uidl_format = %m
is applied.
mbox_min_index_size
Default | [None] |
---|---|
Value | size |
See Also | |
Advanced Setting; this should not normally be changed. |
For mboxes smaller than this size, index files are not written.
If an index file already exists, it gets read but not updated.
The default should not be changed for most installations.
mbox_read_locks
Default | fcntl |
---|---|
Value | string |
Allowed Values | dotlock dotlock_try fcntl flock lockf |
See Also |
Specify which locking method(s) to use for locking the mbox files during reading.
To use multiple values, separate them with spaces.
Descriptions of the locking methods can be found at mbox locking.
mbox_very_dirty_syncs
Default | no |
---|---|
Value | boolean |
See Also |
If enabled, Dovecot performs the optimizations from
mbox_dirty_syncs
also for the IMAP SELECT, EXAMINE,
EXPUNGE, and CHECK commands.
If set, this option overrides mbox_dirty_syncs
.
mbox_write_locks
Default | dotlock fcntl |
---|---|
Value | string |
Allowed Values | dotlock dotlock_try fcntl flock lockf |
See Also |
Specify which locking method(s) to use for locking the mbox files during writing.
To use multiple values, separate them with spaces.
Descriptions of the locking methods can be found at mbox locking.
In many systems, the user's mails are by default stored in /var/mail/username
file. This file is called INBOX in IMAP world. Since IMAP supports multiple mailboxes, you'll need to have a directory for them as well. Usually ~/mail
is a good choice for this.
For an installation such as this, the mail location is specified with:
# %u is replaced with the username that logs in
mail_location = mbox:~/mail:INBOX=/var/mail/%u
It's in no way a requirement to have the INBOX in /var/mail/
directory. In fact, this often just brings problems because Dovecot might not be able to write dotlock files to the directory (see below). You can avoid this completely by just keeping everything in ~/mail/
:
# INBOX exists in ~/mail/inbox
mail_location = mbox:~/mail
mail_location
Keys For mbox, the default keys are:
Key | Default Value |
---|---|
FULLDIRNAME | <empty> (For mbox, this setting specifies the mailbox message file name) |
By default, index files are stored under an .imap/
directory.
See the mail_location setting for an explanation of how to change the index path. Example:
mail_location = mbox:~/mail:INBOX=/var/mail/%u:INDEX=/var/indexes/%u
Make sure that all software accessing the mboxes are using the same locking methods in the same order. The order is important to prevent deadlocking.
From Dovecot's side you can change these from mbox_read_locks
and mbox_write_locks
settings.
Often mbox write locks include dotlock, which means that Dovecot needs to create a new <mbox>.lock
file to the directory where the mbox file exists. If your INBOXes are in /var/mail/
directory, you may have to give Dovecot write access to the directory. There are two ways the /var/mail/
directory's permissions have traditionally been set up:
/tmp
). You can do this with chmod a+rwxt /var/mail
.You can give Dovecot access to mail group by setting:
mail_privileged_group = mail
NOTE: With LDA the mail_privileged_group
setting unfortunately doesn't work, so you'll have to use the sticky bit, disable dotlocking completely, or use LMTP server instead.
In some systems the /var/mail/$USER
files have 0660 mode permissions. This causes Dovecot to try to preserve the file's group, and if it doesn't have permissions to do so, it'll fail with an error like:
imap(user): Error: chown(/home/user/mail/.imap/INBOX, -1, 12(mail)) failed: Operation not permitted (egid=1000(user), group based on /var/mail/user)
There is rarely any real need for the files to have 0660 mode, so the best solution for this problem is to just change the mode to 0600:
$ chmod 0600 /var/mail/*
With POP3 it's been traditional that users have their mails only in the /var/mail/
directory. IMAP however supports having multiple mailboxes, so each user has to have a private directory where the mailboxes are stored. Dovecot also needs a directory for its index files unless you disable them completely.
If you really want to use Dovecot as a plain POP3 server without index files, you can work around not having a per-user directory:
/var/empty
mail_location
so that the mail root directory is also the empty directory and append :INDEX=MEMORY
to it. For example: mail_location = mbox:/var/empty:INBOX=/var/mail/%u:INDEX=MEMORY
/var/empty
as the directory containing other mailboxes than INBOX. If the directory is writable, all the users will have their mailboxes shared.By default Dovecot uses filesystem layout under mbox. This means that mail is stored in mbox files under hierarchical directories, for example:
| File | Description | | ~/mail/inbox
| mbox file containing mail for INBOX | | ~/mail/foo
| mbox file containing mail for mailbox "foo" | | ~/mail/bar/baz
| mbox file containing mail for mailbox "bar/baz" |
One upshot of this is that it is not normally possible to have mailboxes which are subfolders of mailboxes containing messages.
As an alternative, it is possible to configure Dovecot to store all mailboxes in a single directory with hierarchical levels separated by a dot. This can be configured by adding :LAYOUT=maildir++
to the mail location. There are, however, some further considerations when doing this; see child folders for some examples.
Under mbox format, Dovecot maintains the subscribed mailboxes list in a file .subscriptions
which by default is stored in the mail location root. So in the example configuration this would be at ~/mail/.subscriptions
.
If you want to put this somewhere else, you can change the directory in which the .subscriptions
file is kept by using the CONTROL
parameter. For example:
mail_location = mbox:~/mail:CONTROL=~/mail-control
would store the subscribed mailboxes list at ~/mail-control/.subscriptions
.
One practical application of the CONTROL
parameter is described at child folders.
By default, Dovecot stores messages for INBOX in an mbox file called "inbox", and messages for all other mailboxes in an mbox file whose relative path is equivalent to the name of the mailbox. Under this scheme, it is not possible to have mailboxes which contain both messages and child mailboxes.
However, the behaviour (for mailboxes other than INBOX) can be changed using the DIRNAME
parameter. If the DIRNAME
parameter is specified with a particular value, then Dovecot will store messages in a file with a name of that value, in a directory with a name equivalent to the mailbox name.
There are, however, some further considerations when doing this; see child folders for an example.
Under mbox, it is not normally possible to have a mail folder which contains both messages and sub-folders. This is because there would be a filesystem name collision between the name of the mbox file containing the messages and the name of the directory containing the sub-folders. For example:
~/mail/foo
.~/mail/foo/bar
, but this cannot happen because this relies on the existence of a directory ~/mail/foo/
which can't exist because there is already a file with that name.Under mbox, Dovecot normally stores mail folders in "filesystem" layout. In this layout, mail folders are stored in mbox files (potentially under subdirectories) with the same relative path as the mail folder path. For example:
File | Description |
---|---|
~/mail/foo | mbox file containing mail for mail folder "foo"; cannot create any mail sub-folders of "foo" |
~/mail/bar/baz | mbox file containing mail for mail folder "bar/baz"; cannot create any mail sub-folders of "bar/baz" |
~/mail/inbox | mbox file containing mail for INBOX |
If there is a requirement to be able to have a mail folder which contains both messages and sub-folders, then there are two ways to do it:
These approaches are described in more detail below.
Dovecot can be configured to keep mbox mail in a Maildir++-like layout. This makes Dovecot keep mail in mbox files where all the mailbox folder naming levels are separated with dots (with a leading dot). For example:
File | Description |
---|---|
~/mail/.foo | mbox file containing mail for mail folder "foo" |
~/mail/.foo.bar | mbox file containing mail for mail folder "foo/bar". We can now do this. |
~/mail/.bar.baz | mbox file containing mail for mail folder "bar/baz" |
~/mail/inbox | mbox file containing mail for INBOX |
This can be enabled by adding :LAYOUT=maildir++
to the mail location:
# Incomplete example. Do not use!
mail_location = mbox:~/mail:LAYOUT=maildir++
However, there is a problem. Under mbox, setting LAYOUT=maildir++
alone leaves Dovecot unable to place index files, which would likely result in performance issues. So when using LAYOUT=maildir++
with mbox, it is advisable to also configure INDEX
. Now, mail files (other than INBOX) all have names beginning with a dot, so if we like we can store other things in the ~/mail
directory by using names which do not begin with a dot. So we could think to use INDEX
to store indexes at ~/mail/index/
. Example:
# Incomplete example. Do not use!
mail_location = mbox:~/mail:LAYOUT=maildir++:INDEX=~/mail/index
If we do this, then indexes will be kept at ~/mail/index/
and this will not clash with any names used for mail folders. There is one more thing we may want to consider though. By default Dovecot will maintain a list of subscribed folders in a file .subscriptions
under the mail location root. In this case that means it would end up at ~/mail/.subscriptions
. This would then mean that it would be impossible to create a mail folder called "subscriptions". We can get around this by using the CONTROL
parameter to move the .subscriptions
file somewhere else, for example into the directory ~/mail/control
(again, choosing a name which doesn't begin with a dot so we don't collide with the names of mbox files storing mail folders). That gives us:
# Trick mbox configuration which allows a mail folder which contains both
# messages and sub-folders
mail_location = mbox:~/mail:LAYOUT=maildir++:INDEX=~/mail/index:CONTROL=~/mail/control
This then allows mail folders which contains both messages and sub-folders without possibility of naming collisions between mail folders and other data.
There is one further wrinkle. Specifying :LAYOUT=maildir++
for mbox changes the default hierarchy separator from a slash to a dot. This should not be a problem for IMAP clients as the hierarchy separator is exposed through IMAP. However anything which expects to just "know" that the hierarchy separator is a slash may get confused. This can be worked around by configuring Namespaces to set the folder separator back to a slash.
In the default "filesystem" example from above, we can't create any sub-folders of "foo" because there is a file - foo
- in the way. So we could think to get rid of that file and put a directory there instead. But if we do that then we need somewhere to put the messages for folder "foo". We could think to put them in a specially-named file in the directory: foo/
. Then if we wanted to create a sub-folder of "foo" we would be fine because we could then do that. The rule would then be that messages go into the specially-named file in the directory corresponding to the mail folder name. We want to choose a special name which would be unlikely to collide with a folder name. We could think to use something like mBoX-MeSsAgEs
. Now, it turns out that you can configure Dovecot to do this using the DIRNAME
parameter:
# Incomplete example. Do not use!
mail_location = mbox:~/mail:DIRNAME=mBoX-MeSsAgEs
With that config, we would get a layout like this:
File | Description |
---|---|
~/mail/inbox | mbox file containing mail for INBOX |
~/mail/foo/mBoX-MeSsAgEs | mbox file containing mail for mail folder "foo" |
~/mail/foo/bar/mBoX-MeSsAgEs | mbox file containing mail for mail folder "foo/bar" |
However there is a problem. Under mbox, setting DIRNAME
alone leaves Dovecot unable to place index files, which would likely result in performance issues, or worse, if the index directory gets created first, this will obstruct the creation of the mbox file. So when using DIRNAME
with mbox, it is also necessary to configure INDEX
. The question then arises where to put index files.
Any directory under the ~/mail
directory could be considered as a mail folder. We could think to use a name beginning with a dot, for example ~/mail/.index
but that would then mean that it would not be possible to create a mail folder called ".index"; unlikely, but it would be nice to have as few implementation-specific restrictions as possible.
In addition, by default, Dovecot will create a file .subscriptions
at the mail location root to hold a list of mailbox subscriptions. This would make it impossible to create a mail folder called ".subscriptions". But we can move the .subscriptions
file to another directory by using the CONTROL
parameter. To get around these issues, we can add another directory layer which separates these purposes. For example:
# Trick mbox configuration which allows a mail folder which contains both
# messages and sub-folders
mail_location = mbox:~/mail/mailboxes:DIRNAME=mBoX-MeSsAgEs:INDEX=~/mail/index:CONTROL=~/mail/control
would result in the following layout:
File | Description |
---|---|
~/mail/mailboxes/foo/mBoX-MeSsAgEs | mbox file containing messages for mail folder "foo" |
~/mail/mailboxes/foo/bar/mBoX-MeSsAgEs | mbox file containing messages for mail folder "foo/bar" |
~/mail/mailboxes/inbox | mbox file containing messages for INBOX |
~/mail/control/.subscriptions | File containing list of subscribed mailboxes |
~/mail/index/INBOX/dovecot.index.* | Index files for INBOX |
~/mail/index/foo/dovecot.index.* | Index files for mail folder "foo" |
~/mail/index/foo/bar/dovecot.index.* | Index files for mail folder "foo/bar" |
~/mail/index/dovecot.mailbox.log | Other index files |
Restrictions on mail folder names are then minimised; we can't have mail folders with the names "mBoX-MeSsAgEs", "dovecot.index.*, or "dovecot.mailbox.log".
Unlike the Maildir++ layout approach above, because we are still using "filesystem" layout, the hierarchy separator remains as a slash.