After switching to LDAP, there’s no link between the user and the existing mailbox, and all admin roles are lost. Could this be due to a BASE DN misconfiguration? Is there a schema layout for a proper user transfer to LDAP?

Citadel -> LDAP -> User authentication works fine. The only thing I notice is the schema. Is there any documentation for this??

Funny you mention that, I just pushed a change into the code a few days ago that lets you specify an LDAP group whose membership conveys administrator privileges to Citadel users.  That'll be available soon.

However, the LDAP authentication system isn't really built to be able to switch over to it if you already have self-contained authentication.  It doesn't look for a user who has a matching name or email address.  It looks for a user with a matching uid (uid is set to -1 for self contained auth).  I suppose you could dump your database and edit the uid's and then reload it.  

Hello,

  How I change the timezone for individual users?  So far, I haven't found any way of changing it for users.  I found only one timezone option, but it seemed to be a fall-back option.  It didn't change anything for users.

Thanks,

Harlan...

Mo Jul 13 2026 09:12:19 EDT von IGnatius T Foobar Betreff: Re: Prepareing

 After switching to LDAP, there’s no link between the user and the existing mailbox, and all admin roles are lost. Could this be due to a BASE DN misconfiguration? Is there a schema layout for a proper user transfer to LDAP?

Citadel -> LDAP -> User authentication works fine. The only thing I notice is the schema. Is there any documentation for this??

Funny you mention that, I just pushed a change into the code a few days ago that lets you specify an LDAP group whose membership conveys administrator privileges to Citadel users.  That'll be available soon.

However, the LDAP authentication system isn't really built to be able to switch over to it if you already have self-contained authentication.  It doesn't look for a user who has a matching name or email address.  It looks for a user with a matching uid (uid is set to -1 for self contained auth).  I suppose you could dump your database and edit the uid's and then reload it.  

Interesting, maybe it’s faster to export the mailboxes, switch to LDAP, and import the mailboxes via Mailstore. That’s the approach I’ll test. Do you have a link to the LDAP group schema so I can try it with admin rights? ;)

Mi Jul 15 2026 09:23:21 EDT von IGnatius T Foobar Betreff: Re: Prepareing

Better yet, I'll give you a cheat code :)

Go to freeipa.org and check out their demo server. Point your Citadel LDAP to it and go !

I want use my own local LDAP. Its a very secure setup and productive server. I'll wait for a schema down build it inside my lldap. 

Dear All,

I am having the following problem with purge. During the processing after purge (where there are generated temporary files that can ocuppy more than 30GBytes. In my case until exaust my free disk space of 52GBytes) there happens several core dumps (each 2 minutes) that citadel recovers every time (until now). Can I infer that the data files are corrupt or are valid after the processing?

Thank you,

Luís Gonçalves.

I am having the following problem with purge. During the processing after purge (where there are generated temporary files that can ocuppy more than 30GBytes. In my case until exaust my free disk space of 52GBytes) there happens several core dumps (each 2 minutes) that citadel recovers every time (until now). Can I infer that the data files are corrupt or are valid after the processing?

Luis, this seems to happen to you a lot.  Are you sending large amounts of unusually large messages through your system?  If so, and you are limited on disk space, perhaps you should purge more than once per day, so the database can clean up its log files multiple times per day.

To do so, run the following command from your Citadel Server directory:

./sendcommand TDAP

Furthermore, in the next release of Citadel Server we will have support for LMDB, and you might want to convert your database from Berkeley DB to LMDB.  This is because LMDB doesn't write separate journal files and will manage your space better.