RTCyrus3 - sendmail and cyrus-imap integration
Andrzej Adam Filip
anfi at xl.wp.pl
Fri Apr 20 12:52:51 EDT 2007
>> Andrzej Adam Filip wrote:
>> I would like to announce new version of RTCyrus recipes for Sendmail and
>> Cyrus IMAP integration.
>> RTCyrus3 has been designed for easy integration of sendmail with virtual
>> domains provided by Cyrus IMAP. It includes support for aliases.
> Hi Andrzej,
> It's great to see you still working on this.
> Currently we are using a murder setup, with virtual domains and
> socketmap for user checking. We are using version 2 of your
> sendmail/cyrus integration.
> I've setup this version 3 with slight modification. My virtual hosts are
> stored in LDAP, for example, so I've changed the define in sendmail.mc to:
> (with the appropriate LDAP entries and config in sendmail.mc).
> So I have version 3 working, and everything seems to be fine.
> I am wondering if there's any reason to switch from version 2 to version
> 3? Is there a benefit I'm not seeing?
There are two main improvements:
a) *full* support for aliases in cyrus virtual domains
user at cyrus.domain: ...
[there were problems with aliases for "no such mailbox"]
b) cyrus virtual domains are listed in *single* place/file
In RTCyrus2-mailertable the domains have to be listed in:
* access db
[HUPing sendmail daemon is required in case of changing the list]
> Also, is there no way of making the over quota message something other
> than "User unknown" or would that have to be handled in cyrus source?
Due to *sendmail.cf* limitations it would require changes in Cyrus-IMAP
implementation of socket map server.
In current implementation Cyrus-IMAP ignores "map name". Proper over
quota handling should do something like recipe below:
* return current results for map named "cyrus"
* return "in reply" mailbox status for map named cyrus2
mailbox OK => return "OK:mailbox" instead of current "mailbox"
mailbox over quota => return "ERR:OVER:mailbox" instead of current way
of "signaling problem" [sendmail.cf can't access "explanation string"]
I think it is a YASA ("yet another simple after") solution :-)
[pl>en: Andrew] Andrzej Adam Filip : anfi at priv.onet.pl : anfi at xl.wp.pl
Home site: http://anfi.homeunix.net/
More information about the Info-cyrus