Do you see it in GroupWise Administration if you are connected to the primary domain?

If no, then try a top down rebuild ...

it's a one-domain system, so yes - I am connected to the Primary Domain. Domain database verify returns all OK. I've even done a "Recover Database" - no errors there either. Can't figure out the correct curl command to delete it (I get unknown id as a result)...

My usual suggestion is to run a two domain GroupWise system without exception. That's the only way to repair a primary domain in case of troubles. However this does not help in your situation ...

So I have two other ideas:

• Open a case. Support will be able to repair and to clean up your down database.
• Or try to create a domain myDo and a post office myPo. After some minutes of successful operation, delete myPO again. Wait some minutes and then delete myDo.
  All these activities can happen on your existing server - but you have to use different ports 

@ Diethmar - would it help to deattach the primary domain and then create a GroupWise system from the detached domain?

George

Georg H. Weber I am not sure what you mean. If there is only one domain, then there is nothing to detach ....

@ Diethmar,

You have recommended to open a call and for the supporters it is a small operation to remove the orphand objects, I had this also once in the past. I can give the

I had a similar situation in the field and there was no support available. However, there is a fundamental prerequisite for this task, the path with the secondary or other domain files, including useren, must be present in the file system of the GW server. In principle the following steps:

gwadminutil ca -d /<path to primary> -g -f


gwadminservice -detached -home /<path to primary> -ip <new IP> -adminPort <9710 by default>

here previously a trick that binds a new IP as the first IP and the old IP as the second IP. The new IP then remains on the system. Important, the Cli command should be executed on the console of the GW server in an ssh session, because a Java app starts a terminal with the GWadmin service key.  At this point there is a new Certifcat and whatever else is attached to it. The second IP is removed afterwards, but check whether the second IP can still be found in the certificate later. There are also methods for removing it.

As far as I remember, the connection configuration is also deleted with the gwadminutilservice deateched. The domains and MTAs then no longer know anything about each other. I hope I remember this correctly.

With gwsc -i the config is entered into the gwha.conf again, starting from the Pri Dom. If the files are still available from the sec domain, a second domain is rebuilt and mounted. The connection configuration of the domains is also created again with this config. It is important to check all certificates for all objects of the GW system, I already had the case that there were still old certificates in the chain that had to be deleted.


As I said, this is just off the top of my head from your case in the past, what you read here must be checked for validity in the lab.

George

I was on a wrong path - I thought you want to do something similar like releasing a domain. But there is nothing to release ... 

One more thought.

I think this is a small environment (only one domain). So they should have the opportunity to open a case.

I know some large environments (i.e. universities) where no free cases are included.

Hi Diethmar, if only the Pri Dom is still there and there is a pending Sec Dom, you can act as if you were moving the system to a new server. As long as everything is still in the file system, it can be reassembled on foot if the GroupWise system has been deactivated beforehand, as far as I remember. With a small installation the admin may lack the experience and therefore the call you recommended is the better way.

As I said, my idea really needs to go back to the lab.

Translated with DeepL.com (free version)

Thank you both for your time/answers!!! 

I have opened a support ticket last week, as I also believe the wpdomain.db needs to be amended accordingly, but they still haven't asked for it.

My current plan is (if they don't fix the db in the next few days) to add a secondary domain and PO, move all users there, and then release the new domain and start a new system (this should result in a clean system with no pending operations).

There may be an easier way (e.g. through the REST API with curl), but I don't know the pending object ID:

" DELETE /system/pending/{id} "

Cheers