TM1 Forum

Hi,

I am trying to set up a TM1 application that looks at a distinct server instance. Currently I have 3 applications set up on a server 'fin-plan', instance 'Model'. I have backed up last years complete model so that it resides on the server as 'Model2014'. I would now like to replicate one of the 3 applications, but on 'Model2014' so that we can see what we did last year.

I have managed to get performance modeller to log on to 'Model2014', and to create the application and add a websheet. The websheet looks for the server 'Model2014' and works OK in Perspectives.

The problem occurs when I try to validate the application in performance modeller. I get the error "The view 'IEM IE' is based on cube 'IEM IE' used by the application '{91ef9113-f170-4e99-bc35-8b9ba07a5559}'."
IEM IE is a cube in 'Model' and obviously in 'Model2014'.
I have tried restarting services but no joy. Is there some more fundamental issue in play here? I have read elsewhere that it is not possible to have 2 applications looking at the same cube. I am hoping that the fact that the 2 applications are looking at different server instances would mean that this is not a problem?

Any pointers gratefully received.
TM1 10.1.1

When you create an application it create a number of metadata objects on the tm1 server; these objects are the ones that are checked to make sure don't have any overlaps etc when you try to create a new application.
Presumably what has happened is you created the application THEN took a copy of the server; therefore taking these objects with it.

So although there is no application available to access the data on your new server instance - it still has all the relevant security objects etc as if you did have an application that could try and grab a hold of it at any point.

So what you are trying to do seems reasonable and is perfectly doable but you need to get rid of the objects that are causing it to clash.

Thanks Declan, very helpful.
Is there an easy (and safe) way to identify the objects I need to remove, obviously without screwing up the data?