Adding second M365 shared mailbox loses Email of first added

Infos:

  • Used Zammad version: 6.5.0
  • Used Zammad installation type: docker
  • Operating system: Ubuntu 24.04
  • Browser + version: brave-latest

Expected behavior:

Each added M365 (more than one!) shared mailbox should use their own Group and sending email address.

Actual behavior:

Adding the first mailbox everything is fine. Add, Autenticate, set group ok.
Adding the next mailbox causes the previous addes losing the email address (in the M365 Graph overview the Email address is empty

Steps to reproduce the behavior:

Just add a second M365 Graph shared mailbox

For explanation: I’ve done this before in a different zammad/docker instance. no problems at all. I have no idea whats causing this problem

Sounds like the accounts have the same mail address. We‘reat 6.5.2 btw.

Yes could be that the mail address has been moved, when it is about the same one. But without more information it’s difficult to say.

the added account do not have the same email addresss. I’ve moved to 6.5.2 to sort this out - same behaviour.

I tracked the database and found the following:
when adding a shared mailbox a new entry is added in table email_addresses. in the name field is a generic name (in my case it is “Ticket Reklamationen Support”.
When adding the next mailbox, the same name is used and the email address of the formerly added mailbox is overwritten.
When changeing the name by hand (in the database) i can add the next mailbox.
It seems that the process choosed the name field as key. But i cant change the name during import.
I don’t think this is a wanted behaviour???

Please don’t fiddle directly in your database. This may lead to data loss or inconsistent data.

If renaming the display name helps, then I’d wonder why the displayname is identical - is it the same mail account…? Zammad supports one logical account only (as in: you cannot add the same account more than once).

I wonder too - there a definitely different mail accounts. And changing the name in the database was the only way to solve this problem.
For me, this is a bug - where should I place this?

When you have clear reproduction steps on a fresh Zammad instance, you can create an issue on GitHub, but when it’s unclear to reproduce we can not do much.